Katalaga Hakim — Marketing Notes

How to Use Claude to Run a Better Google Ads Account

Katalaga Hakim — Fri, 11 Sep 2026 00:00:00 GMT

Most people use AI for Google Ads like they use a slightly faster copywriter.

They ask for headlines. They ask for keywords. They ask for ad variations. Maybe they paste in a product description and ask for a landing page.

Then they look at the output and decide AI is either amazing or useless.

That misses the more interesting opportunity.

The real advantage comes when you stop treating Claude as a tool for generating individual pieces of an account and start giving it enough context to understand the account as a system.

Your products.

Your customers.

Your reviews.

Your competitors.

Your search terms.

Your ad performance.

Your product feed.

Your landing pages.

Your previous tests.

Your conversion data.

Once that information is sitting in one place, the AI can do something much more useful than writing another responsive search ad.

It can help you investigate the account.

The quality of the output starts before the first prompt

This is the part most people skip.

They open Claude, type something like "Create a Google Ads strategy for my ecommerce store," and then wonder why the answer sounds generic.

Of course it does.

The model doesn't know the business.

It doesn't know which products actually sell.

It doesn't know why customers buy.

It doesn't know what objections stop them.

It doesn't know which competitors are taking market share.

It doesn't know what happened in the account last month.

You can't expect useful strategy from an AI that has been given almost no strategic context.

A better approach is to build a project around the business first.

Give it the product pages, best sellers, reviews, objections, competitor websites, previous ads, search-term reports, offers, brand guidelines, landing pages and performance exports.

Then ask it to understand the business before asking it to recommend anything.

That last part matters.

Tell it to identify what it knows, what it doesn't know and what information is missing.

You might discover that the problem isn't the AI.

Your own account knowledge is incomplete.

Give the AI customer language, not just keywords

Keyword tools tell you what people search.

Customers tell you why they search.

Those are different things.

If you're selling a product that solves a particular problem, the language customers use when describing that problem can become incredibly useful for advertising.

Reviews can reveal objections.

Reddit discussions can reveal frustrations.

YouTube comments can reveal what people have already tried.

Customer support conversations can reveal why someone hesitated before buying.

That information can influence your ads, landing pages and even your product positioning.

An AI model is particularly useful here because it can process a large amount of customer language and organize it into patterns.

What are people frustrated about?

What solutions have they already tried?

What do they want instead?

What objections keep appearing?

Which phrases sound like genuine customers rather than marketing copy?

Now your ad copy has something real to work from.

That's much better than asking AI to "write compelling headlines."

Competitor research becomes more useful when you look for patterns

The same principle applies to competitors.

Don't just ask Claude to summarize three competitor websites.

Give it their ads, landing pages, offers, guarantees, product positioning and other publicly available marketing material, then ask it to compare them.

You want to know what everyone is saying.

More importantly, you want to know what nobody is saying.

If several competitors are making the same promise, that tells you something about the market.

It doesn't mean you should copy them word for word. It means the market may already understand and respond to that particular angle.

But if customer research keeps revealing an important concern that none of your competitors address, that's potentially more interesting.

That is where the combination of research and AI becomes useful.

The AI can organize the evidence.

You still decide what is worth testing.

Don't let keyword research become a spreadsheet exercise

There is another common AI mistake.

Ask for keyword research and you'll get a very impressive list.

Hundreds of keywords.

Categories.

Search intent.

Match types.

Potential ad groups.

Everything looks organized.

And then someone launches half of it.

That's how an account becomes a keyword dump.

The better question isn't:

"What keywords can we target?"

It's:

"Which searches are commercially important enough to deserve budget right now?"

For an ecommerce account, that distinction matters.

Someone searching for a specific product is different from someone researching a broad problem.

Someone comparing two solutions is different from someone searching for a particular brand.

Someone looking for a gift is different from someone ready to purchase the product today.

AI can help categorize those searches and turn a large keyword universe into a smaller launch plan.

That is far more useful than producing another enormous spreadsheet nobody will act on.

Every campaign should have a job

Once you've decided what deserves to be targeted, the next problem is account structure.

This is another place where AI can be useful, provided you make it explain itself.

Don't just ask for campaign names.

Ask why each campaign exists.

What signal is it supposed to generate?

What traffic is it supposed to capture?

What should determine whether it gets more budget, stays where it is or gets cut?

That forces the account to have a purpose.

A branded campaign has a different job from non-branded search.

Shopping has a different job from competitor search.

Remarketing has a different job from prospecting.

A testing campaign has a different job from an established campaign that is already producing profitable purchases.

If a campaign has no clear role, it becomes very easy to keep spending money simply because it exists.

Good account structure isn't about having more campaigns.

It's about knowing why each one is there.

AI can write the ads, but it shouldn't invent the message

This is where AI-generated copy often becomes painfully obvious.

"Discover premium quality."

"Transform your everyday routine."

"Experience the difference."

None of those statements tell us much.

The better copy comes from everything that happened earlier in the process.

Customer language gives you the problem.

Competitor research gives you the market context.

The product gives you the actual mechanism and benefits.

The offer gives you the reason to act.

The search query gives you the intent.

Now Claude has something to work with.

That also means the copy should change depending on where the customer is in the buying process.

Someone searching for a specific product doesn't need the same message as someone still trying to understand their problem.

The first customer may need reassurance and proof.

The second may need education.

Treating both searches as identical because they contain related keywords wastes a lot of the information Google is already giving you.

Your product feed deserves the same attention as your ads

For ecommerce, there is another piece of the system that gets overlooked: Merchant Center.

Your product feed is how Google understands what you're selling.

That means product titles, descriptions, categories, identifiers, attributes, variants, pricing, availability and images all matter.

AI can help audit that information at scale.

Give it a product feed and ask it to find missing attributes, weak titles, unclear descriptions and opportunities to make the product information more useful.

You can then rewrite titles around the actual search intent and product characteristics rather than leaving every product with a generic name.

This matters because Shopping isn't just an advertising problem.

Google needs to understand the product before it can decide when that product is relevant.

A great campaign cannot compensate for poor product information indefinitely.

The landing page has to continue the conversation

There is no point creating a highly relevant ad and then sending the customer to a generic page.

Imagine someone searches for a very specific product requirement.

The ad speaks directly to that requirement.

They click.

The product page says something vague like "Premium Quality. Better Living."

Now the customer has to figure out whether the product actually solves the problem they searched for.

You've just handed the customer another job.

The landing page should continue the conversation started by the search.

That means the headline, product positioning, proof, objections and call to action should make sense for the intent that brought the visitor there.

AI can help create different landing-page briefs for different search intents.

But again, the value isn't that Claude can write a headline.

The value is that it can take the research, customer language and search intent you've already collected and use those inputs to build a more relevant page.

Then the account starts feeding itself

This is where the workflow gets interesting.

Once the campaigns are running, you have another source of information.

Actual performance.

Instead of looking at Google Ads once a week and asking whether ROAS went up or down, export the data and give the AI a structured problem to investigate.

Compare the last seven days with the previous seven.

Look at spend, revenue, conversions, CPA, CPC, CTR, impression share and search terms.

Then ask it to flag meaningful changes.

Which campaigns are deteriorating?

Which products are spending without converting?

Where is irrelevant traffic appearing?

Which campaigns are limited by budget?

Which are limited by rank?

Is branded traffic leaking into campaigns that are supposed to acquire new customers?

What changed?

Most importantly:

What should we investigate next?

Now the AI isn't generating strategy from thin air.

It is analyzing what actually happened inside the account.

That is a completely different use case.

Don't scale because one week looked good

One of the easiest ways to damage a Google Ads account is to confuse noise with a trend.

A campaign has a great week.

Someone increases the budget.

Performance falls.

The budget gets reduced.

Another good week appears.

The process starts again.

You are effectively making decisions based on whatever happened most recently.

A better approach is to look across several time windows.

Seven days.

Fourteen days.

Thirty days.

If a campaign is performing well across multiple windows, the signal becomes more interesting.

If something looks amazing for three days and mediocre over the previous month, you probably don't have enough evidence yet.

AI is useful here because it can compare these periods quickly and force you to look at the broader pattern rather than the most exciting number in the dashboard.

The important part is the rule you give it:

Don't recommend scaling based on one good week.

The real advantage is the feedback loop

This is why I don't think the biggest opportunity with AI in Google Ads is simply "AI writes your ads."

That is the least interesting part.

The more valuable system looks something like this:

Customer research informs positioning.

Positioning informs keyword and campaign strategy.

Search intent informs ad copy.

Campaign data informs landing-page decisions.

Product data informs the Merchant Center feed.

Performance data reveals what worked.

Those results become new information for the next round of research and testing.

The system gets better because every stage feeds the next one.

That is very different from opening a new chat every Monday and asking for five ad headlines.

AI doesn't replace the operator

There is one part of this workflow that shouldn't be automated away.

Judgement.

AI can tell you that CPA increased.

It can identify which campaign caused the change.

It can compare the search terms.

It can suggest several possible explanations.

But somebody still needs to decide what the business should do.

Maybe the product margin doesn't support the current acquisition cost.

Maybe the offer is weak.

Maybe the landing page is the problem.

Maybe the campaign is actually fine and demand has changed.

Maybe a competitor launched a better offer.

Maybe the account is being optimized toward the wrong conversion.

Those aren't purely technical questions.

They require someone who understands the business.

AI is extremely good at processing information and reducing the amount of manual work required to get from data to a decision.

It is much less useful when you give it a broken business and ask it to make the business unbroken.

That's the distinction.

The Google Ads account becomes a system, not a collection of tasks

The interesting future of AI-assisted Google Ads isn't a person disappearing and Claude taking over the account.

It's one operator being able to manage much more of the account because the repetitive analytical work has become dramatically faster.

Research can happen faster.

Keyword analysis can happen faster.

Competitor analysis can happen faster.

Feed audits can happen faster.

Ad variations can happen faster.

Landing-page concepts can happen faster.

Performance audits can happen faster.

Testing ideas can happen faster.

But the strategy still depends on the quality of the information going into the system and the judgement of the person deciding what to do with the output.

That's why the strongest Google Ads workflows won't be built around AI alone.

They'll be built around AI + account data + customer research + advertising platforms + human judgement.

Give Claude no context and you'll get generic marketing advice.

Give it the business, the customers, the competitors, the product feed and the performance data, then make it explain the evidence behind its recommendations.

Now you have something much more interesting.

Not an AI that writes your Google Ads.

An AI that helps you operate the account.

5 Ways to Improve Your Shopify Retention

Katalaga Hakim — Wed, 09 Sep 2026 00:00:00 GMT

The ads are running. The product pages have been worked on. Email is collecting subscribers. Maybe there is even a loyalty program sitting inside the store.

Then the customer buys.

And disappears.

That gap between the first order and the second one is easy to overlook because nothing appears to be broken. The store is still getting sales. The campaigns are still spending. Revenue is still coming in. There isn't an obvious warning inside Shopify telling you that a large part of your customer base has no reason to return.

So the natural response is usually to go looking for more new customers.

That can become an expensive habit.

If you are continually paying to acquire another customer while the customers you already paid to acquire never come back, you can end up solving the wrong problem.

The more interesting question is not simply how to increase repeat purchases. It is why aren't these customers buying again in the first place?

That question changes what you do next.

Start with the gap between order one and order two

Before adding another retention app, look at the customers who made one purchase and never made a second.

How long has it been?

What did they buy?

How much did they spend?

Are some products much more likely to produce a second purchase than others?

Do certain customer groups return while others disappear?

Is there a normal repurchase cycle for the product?

This matters because not every customer is supposed to buy again next week.

Someone buying a consumable product may naturally need another order in 30 or 60 days. Someone buying a piece of furniture may not have another reason to purchase for years. Someone buying a gift may never have been a particularly strong repeat customer to begin with.

Treating all of these customers as one retention audience makes the numbers much less useful.

The first thing to understand is the expected buying behaviour.

Then you can see where the actual gap is.

Having an email flow doesn't mean email is working

This is another trap Shopify stores fall into.

The store has a welcome sequence.

There is a post-purchase email.

There is a winback campaign.

Maybe there are abandoned cart emails too.

So everyone looks at the setup and concludes that retention is covered.

But installing the machinery is not the same as producing the result.

A post-purchase sequence might send customers three emails and still give them no compelling reason to buy again.

The timing might be wrong. The product recommendation might make no sense. The message might sound like another promotion instead of something useful. The customer might have bought a product that does not need replacing yet.

A retention system has to follow the customer's actual relationship with the product.

If customers normally reorder after eight weeks, sending a heavy second-purchase offer three days after delivery isn't necessarily clever. It may simply be noise.

Timing is part of the strategy.

Discounts can make the numbers look better while making the business worse

When repeat purchases are low, the easiest idea is often another discount.

Give customers 15% off their next order.

It can work.

But there is a problem with making discounts the default retention strategy: you may be paying for the repeat purchase with margin you could have kept.

A customer who would have bought again anyway has now been trained to wait for the code.

That doesn't mean discounts should never be used. They can be useful when there is a clear reason for them and the economics work.

But the question should be bigger than, "What discount will get them back?"

It should be, "What would make this customer genuinely want to return?"

Sometimes that is a better product recommendation.

Sometimes it is replenishment timing.

Sometimes it is a complementary product.

Sometimes it is simply making the next purchase easier.

And sometimes the problem has nothing to do with marketing at all.

Loyalty programs don't create loyalty by themselves

A points program can look impressive inside a Shopify admin.

Customers earn points. They have a balance. There is a rewards page. Everything is technically working.

But customers don't owe you a second order because you installed a loyalty app.

The reward has to matter to them.

If someone has to spend hundreds of dollars to earn a reward they don't particularly care about, the program is not creating much motivation.

This is why I would be careful about looking at the existence of a loyalty program as evidence that retention is being handled.

The better question is whether customers understand the benefit and whether that benefit changes their behaviour.

The same applies to store credit.

Store credit can be interesting because it gives a customer a concrete reason to have value sitting inside the store. But it still needs to be tested against the economics of the business and the behaviour of the customer.

The mechanism matters less than whether it changes what happens between order one and order two.

The post-purchase period is more important than many stores treat it

The customer has just bought something.

At that point, the store already knows more about that customer than it did before the first order.

It knows what they purchased. It knows the price point. It may know the product category, variant, quantity, and other information that can help determine what they might need next.

That information shouldn't disappear into an order confirmation email.

The post-purchase experience can help create the next reason to buy.

A customer who bought a particular product may need something that works with it. Another customer may need a replacement after a predictable period. Someone else may be better suited to a higher-value version after using the first product.

Those are very different follow-up conversations.

Sending every customer the same "Come back and shop" email ignores the information the store already has.

Good retention starts becoming more specific.

Look at the product before blaming the customer

Sometimes the reason people don't return is sitting in the catalogue.

A store may have one product that generates lots of first-time purchases but very few second orders.

Another product might have a smaller number of customers but a much higher repeat rate.

That difference matters.

If the business treats every first-time customer as equally valuable, it can miss what the products themselves are telling it.

Some products naturally create repeat behaviour.

Others are one-off purchases.

Others become repeat purchases only when there is a strong complementary product, replenishment cycle, or reason to upgrade.

This is where retention becomes an ecommerce problem rather than simply an email-marketing problem.

You have to look at what was purchased, what happens afterward, and what could reasonably happen next.

Don't change five things at once

One of the more useful ideas in this whole discussion is also one of the least exciting.

Test one thing at a time.

If repeat purchases are weak, you could change the email flows, launch store credit, rebuild the loyalty program, introduce a new discount, change the post-purchase offer, and redesign the retention emails all in the same month.

If the numbers improve, you have no idea why.

If they get worse, you have no idea what caused it.

Start with the gap between the first and second order.

Find the customers who aren't returning.

Look for patterns by product, customer type, and time since purchase.

Then test a specific intervention.

Maybe it is better-timed email.

Maybe it is a product-specific follow-up.

Maybe it is a complementary offer.

Maybe it is store credit.

Maybe the existing loyalty program is fine and simply isn't being presented at the right point in the customer journey.

The point is to learn what changes behaviour rather than accumulating more retention software.

Retention starts with a reason to return

There is a tendency in ecommerce to treat retention as another marketing channel.

Run email.

Add loyalty.

Send a winback.

Offer a discount.

Measure the repeat purchase rate.

But the customer doesn't think about any of those things.

They think about whether they want something from the store again.

That is the real problem to solve.

If the first purchase was good, the product was useful, the experience was easy, and the store has something relevant to offer them next, the second purchase becomes much easier to earn.

If none of those things are true, another automation probably won't fix it.

So before adding another app to the Shopify store, I'd start with a much simpler question:

What happened between the customer's first order and the point where they decided not to buy again?

Once you can answer that, the right retention strategy usually becomes much easier to see.

And if you can't answer it yet, that's probably the first thing worth working on.

How to Sync Shopify With eBay Without Making Your Store More Complicated

Katalaga Hakim — Tue, 08 Sep 2026 00:00:00 GMT

Selling on Shopify and eBay at the same time sounds more complicated than it actually is.

The problem usually starts when you try to manage the two platforms separately. Products are updated in Shopify, then someone has to remember to update eBay. Inventory changes in one place but not the other. An order comes through eBay and has to be entered manually. Prices are different across the two platforms. Eventually, a simple attempt to sell on another marketplace turns into another administrative job.

That is exactly what marketplace integration is supposed to prevent.

For a Shopify store that already has its products, inventory and order management set up, connecting eBay through Shopify Marketplace Connect can give you a central place to manage the relationship between the two platforms. Instead of treating eBay as an entirely separate store, you can use Shopify as the operational base and push the relevant product information across to eBay.

But connecting the accounts is only the beginning. The important part is understanding what happens after the connection, what information eBay requires before a product can be listed, and how to price products when you are offering free shipping.

Start With the Marketplace Connect App

If Marketplace Connect is already installed on your Shopify store, you do not need to install another eBay integration just to get started.

The app can be found inside the Shopify admin under Settings → Apps and sales channels. From there, you can open Marketplace Connect and manage the marketplace connections associated with the store.

Once eBay is connected, the next stage is configuration.

This is where it helps to think about the integration as a system rather than simply a button that says "sync."

You have several separate things to configure: products, prices, inventory, listings, shipping, orders and eBay-specific information.

The connection gives the platforms permission to communicate. It does not automatically mean every product is ready to sell.

Connecting eBay Is Not the Same as Being Ready to Sell

One of the confusing parts of setting up a new eBay integration is that you can successfully connect an eBay account and still be unable to publish products.

The reason is that eBay distinguishes between having an account and being set up as a seller.

A new seller may need to complete additional registration and verification before eBay allows listings to go live. Depending on the account, that can involve identity verification, payment information and other seller requirements.

This is why you can sometimes connect Shopify and eBay successfully, send a product through Marketplace Connect, and then receive an error telling you that you need to create a seller's account.

The integration itself may be working perfectly.

The eBay account simply is not ready to accept the listing yet.

That distinction saves a lot of unnecessary troubleshooting.

If Marketplace Connect reports that the seller account has not been created, the first place to look is eBay itself rather than Shopify.

Once the seller registration is completed, you can return to Shopify and retry the listing.

Your First Products May Still Show as "Errored"

Another common point of confusion is seeing a product marked Errored inside Marketplace Connect after successfully connecting eBay.

That does not necessarily mean the Shopify-to-eBay connection has failed.

It means eBay rejected that particular listing attempt.

The useful thing to do is open the error and read the actual message instead of immediately changing random settings.

eBay listings can require information that Shopify does not necessarily need in the same format. Categories, item specifics, shipping information, return policies and other listing details may need to be mapped before a product can be published.

There can also be warnings that look serious but are not actually blocking the listing.

For example, an account that does not have an eBay Store subscription may receive a message relating to store categories. That does not mean the entire Shopify integration is broken. It may simply mean that the listing is trying to use a store-category feature that the account does not have.

The same principle applies to payment-related messages. If eBay uses Managed Payments, older payment settings may generate informational warnings that are no longer relevant to how the transaction is processed.

The important message is the one that actually prevents the listing from being published.

Set Up the eBay Listing Before You Push Everything Across

Once the seller account is ready, the next job is making sure the eBay listing information is configured properly.

This is where templates become useful.

Your shipping policy, return policy and other listing settings should be established before you start pushing a large catalogue to eBay. Otherwise, you can end up with dozens of products failing for the same reason.

It is much easier to fix the system once than to fix the same problem fifty times.

The exact information required will depend on the products you sell, but the general idea is straightforward: Shopify contains your product catalogue, while eBay has its own marketplace requirements that need to be satisfied before those products can become active listings.

Get the template and marketplace settings right first.

Then publish a small number of products as a test.

If one product goes through successfully, you have a working example that you can use when configuring the rest of the catalogue.

Inventory Sync Is One of the Most Important Parts

The biggest reason to connect Shopify and eBay in the first place is usually inventory management.

Imagine you have five units of a product in Shopify.

Someone buys two from your Shopify store.

Your available inventory becomes three.

If eBay still thinks you have five, you now have two systems reporting different stock levels.

That is how overselling happens.

With inventory synchronization enabled, changes made in Shopify can be reflected on the eBay side, reducing the need to manually maintain two separate inventories.

The same principle applies when an item sells through eBay. The order should flow back into the connected system so that your inventory remains consistent.

This is one of the strongest reasons to use an integration rather than manually copying listings between platforms.

But synchronization should still be monitored.

Automation reduces repetitive work. It does not eliminate the need to check that the automation is actually doing what you expect.

Price Synchronization Needs More Thought

Inventory synchronization is relatively straightforward.

Pricing is different.

You may not want your eBay price to be identical to your Shopify price because the economics of the two channels are different.

eBay charges selling fees, and if you decide to offer free shipping, you are also absorbing the shipping cost.

That means a product selling for $79.95 on Shopify may need to be priced higher on eBay if you want to protect the same underlying revenue.

This is where a pricing rule can be useful.

If your Shopify product is $79.95 and your shipping cost is approximately $8, simply listing it on eBay for $79.95 with free shipping means you are absorbing the shipping expense and eBay's selling fees from the same amount of revenue.

The marketplace price needs to account for those costs.

A simple working formula used in the setup process is:

eBay Price = (Shopify Price + $8) ÷ 0.85

The $8 represents the assumed shipping cost, while 0.85 leaves room for approximately 15% in marketplace fees and related costs.

So a $79.95 Shopify product becomes approximately $103.47 under that calculation.

The exact fee structure can vary by category, account and other circumstances, so this should be treated as a pricing framework rather than a universal eBay fee rule.

The important business principle is more useful than the exact percentage: do not copy your Shopify price to eBay without accounting for the economics of the second sales channel.

Free Shipping Changes the Calculation

If you are considering free shipping on eBay, shipping should be built into your marketplace pricing rather than treated as an afterthought.

Suppose your normal ground shipping charge is $7.99.

On Shopify, the customer may pay that separately.

On eBay, you may decide to advertise the same product with free shipping.

The customer sees a higher product price but no additional shipping charge.

That can make the offer easier to understand and may make the listing more attractive to buyers who prefer free shipping.

But the important part is that you are still paying the carrier.

Free shipping does not mean shipping is free for the seller.

It simply means the seller has incorporated that cost into the selling price.

That is why a product priced at $64.95 on Shopify cannot simply become $64.95 on eBay with free shipping and still have identical economics.

The shipping cost has to come from somewhere.

Do Not Reprice the Entire Catalogue Blindly

One mistake I would avoid is changing every eBay price before testing the system.

Take a few products first.

Calculate the marketplace price.

Publish them.

Then look at the actual numbers.

How much did eBay charge?

How much did shipping actually cost?

What did you receive after the marketplace fees?

Did the product remain competitive against similar listings?

This gives you real information about the channel.

Your $8 shipping assumption might be reasonable for one product but too low for another. A heavier product or a product going to a more expensive shipping zone may have a completely different fulfillment cost.

Likewise, different product categories can have different fee structures.

The goal is not simply to create a formula that produces a higher price.

The goal is to create a price that leaves the business with a sensible margin after the costs of selling through eBay.

Shopify Should Remain the Operational Source

Once everything is connected, the biggest advantage is not simply having products appear on another marketplace.

It is having a central system.

Your Shopify catalogue remains organized in one place. Product information can be pushed to eBay. Inventory can synchronize. Orders can come back into Shopify. Price adjustments can be handled through marketplace rules rather than manually editing every listing.

That reduces the amount of repetitive work required to maintain another sales channel.

But it also means your Shopify catalogue needs to be clean.

If your product titles are messy, your descriptions are incomplete, your variants are poorly organized or your inventory is inaccurate, the integration does not magically fix those problems.

It simply moves the information from one system to another.

Good automation depends on good inputs.

Start Small, Then Scale

There is rarely a good reason to connect hundreds of products and immediately push all of them to eBay without testing the workflow.

Start with a small group.

Make sure the seller account is active.

Connect eBay.

Configure your templates.

Enable the appropriate inventory and price synchronization.

Publish a few products.

Check the listings directly on eBay.

Place yourself in the position of a customer and look at the final listing. Check the title, photographs, description, price, shipping information and return policy.

Then check the financial side.

If the listing sells, what happens to the order?

Does the inventory update?

Does the order appear correctly in Shopify?

Does the final amount you receive still make the product profitable?

Once those pieces work, expand the catalogue.

This approach is much safer than pushing the entire store across and trying to figure out what went wrong afterward.

The Real Benefit Is Not Just More Listings

The reason to connect Shopify and eBay is not simply that you can now say, "We're on eBay."

The real benefit is operational.

You are taking an existing product catalogue and giving it another distribution channel without creating an entirely separate manual system to manage it.

That can be valuable for an ecommerce business.

But the integration itself is not the strategy.

The strategy is deciding which products belong on eBay, how they should be priced, what shipping offer makes sense, how inventory should be controlled and how the resulting orders fit into the rest of the business.

Once those decisions are made, the technology becomes much simpler.

Connect the accounts.

Configure the rules.

Fix the marketplace requirements.

Test a few products.

Watch the inventory.

Check the economics.

Then scale.

That is a much better way to approach marketplace expansion than simply installing an app and assuming the work is finished.

The best integrations are the ones you eventually stop thinking about because the routine work is happening in the background.

But getting to that point requires more than pressing "Connect."

It requires a clean catalogue, sensible pricing, accurate inventory and a clear understanding of what each sales channel costs the business.

Ecommerce SEO Is Simpler Than Most People Make It

Katalaga Hakim — Mon, 07 Sep 2026 00:00:00 GMT

Ecommerce SEO has developed a strange habit of making a relatively straightforward job sound like a specialised science.

There are endless discussions about advanced technical strategies, complicated content systems, elaborate keyword maps and whatever new tactic happens to be popular that month. None of those things are necessarily useless. The problem is that they can distract from the work that actually matters most on an ecommerce site.

If I were working on the SEO of a Shopify store, I would start with a much simpler question: which pages can actually make the business money?

Usually, the answer is sitting in plain sight.

The product pages and collection pages are the commercial centre of the site. These are the pages where someone can go from searching for something to actually buying it. So before publishing another fifty blog posts or chasing obscure keywords, those pages need to be as useful, relevant and well-structured as possible.

Start With the Pages That Make Money

The first job is to improve the pages you already have.

Product detail pages need to properly explain the products. The titles should be clear, the descriptions useful, the page structure logical and the information a customer needs to make a purchasing decision should be easy to find.

Collection pages need the same attention.

A collection should represent a real category or search intent rather than simply being a folder containing a group of products. If customers search for a particular type of product, and the catalogue supports that distinction, there should be a useful page that serves that intent.

This is where ecommerce SEO can be very different from the way people approach informational websites.

You do not necessarily need hundreds of pages.

You need the right pages.

If a collection is already attracting impressions but is poorly structured, improving that page can be far more valuable than creating another article that receives little commercial traffic.

The same applies to product pages. If a product is already relevant to a search but the page does a poor job explaining what the product is and why someone should buy it, creating more content elsewhere on the site does not solve the underlying problem.

Then Create the Pages You Are Missing

Once the existing money-making pages are in good shape, look at what the store does not have.

This is where keyword research becomes useful.

You are not simply collecting keywords because a tool says they have search volume. You are looking for meaningful gaps in the catalogue.

Perhaps customers are searching for a product category that does not have its own collection. Perhaps there is a specific use case that deserves a landing page. Perhaps the store has several products that belong together under a category that has never been properly built.

Those are the kinds of pages worth creating.

The objective is to gradually build a site that covers the commercial territory surrounding the products you sell.

If people are searching for something that your business can genuinely provide, but the website has no useful page for that intent, there is an obvious opportunity.

Create the page.

Then make it good.

Make the Important Pages Easy to Reach

Once you have the right pages, make sure they are not buried somewhere deep inside the website.

A useful rule is to keep important commercial pages within roughly two or three clicks of the homepage where the site's structure allows it.

That does not mean every product needs to be linked directly from the homepage. It means the site's architecture should make important pages easy to discover.

The homepage should lead naturally into major categories. Those collections should lead to relevant products and subcategories. Related pages should connect where there is a genuine relationship.

A customer should not have to wander through five layers of navigation to find a product category that is important to the business.

Neither should the site's internal structure make its most valuable pages look like afterthoughts.

Good site architecture is partly about SEO, but it is also simply good ecommerce.

If a page matters to the business, it should be easy for people to find.

Connect Similar Money Pages

This is where internal linking becomes especially useful.

Your important pages should not exist as isolated destinations.

A collection about winter jackets can link to relevant product pages. A related collection can point towards another category. A useful buying guide can send visitors to the collections it discusses. Product pages can connect customers with related products where that makes sense.

The point is not to add links simply because somebody told you every page needs a certain number of internal links.

The links should help people move through the site.

They also help search engines understand how the pages relate to one another. A site with clear relationships between its products, collections and supporting content is easier to understand than one where every page sits on its own.

Think of the internal linking structure as the network connecting the commercial parts of the store.

If you build useful pages but never connect them, you are making the search engine and the customer work harder than necessary.

Then Build Some Authority

Once the important pages exist and the internal structure makes sense, you still need to give the site a reason to be trusted.

That is where backlinks come in.

You do not necessarily need an enormous link-building campaign. For many ecommerce stores, a consistent number of relevant, worthwhile backlinks over time can be much more useful than chasing a huge number of low-quality links.

The important thing is that authority should support the site as a whole and, where appropriate, help strengthen the pages that matter commercially.

A useful article might attract a link naturally. The article can then link to a relevant collection. That collection can lead visitors towards products.

The homepage may earn links as the brand becomes known. Supporting pages can contribute to the overall authority of the domain.

Again, the goal is not to make the SEO report look impressive.

The goal is to strengthen the pages that matter.

That Is Most of the Work

This is why I think ecommerce SEO gets made to sound more complicated than it needs to be.

There are certainly technical problems that can become complicated. Large ecommerce sites can have indexing issues, faceted navigation problems, duplicate URLs, international SEO considerations, JavaScript rendering issues and all kinds of other things that require proper technical work.

But those situations do not change the basic job.

Build the pages that make money.

Create the important pages you are missing.

Make those pages easy to reach.

Connect related commercial pages with sensible internal links.

Build authority consistently.

Then keep improving the pages based on what customers and search engines are telling you.

That does not mean SEO is effortless. It means the fundamentals are easier to understand than the industry sometimes makes them appear.

A store does not rank because it has the most complicated SEO strategy.

It ranks because the search engine can understand what its pages are about, the pages satisfy the intent behind relevant searches, the site is structured properly and there is enough authority and trust behind the domain.

The advanced work has its place.

But if the product pages are weak, the collections are poorly structured and the important pages are buried three levels deep, I would fix those things before worrying about anything clever.

Ecommerce SEO is not about doing everything.

It is about doing the right things to the pages that matter, then building a structure that allows those pages to become stronger over time.

Google’s Recommended Budget Is Not Your Business Strategy

Katalaga Hakim — Sat, 05 Sep 2026 00:00:00 GMT

There is a particular notification in Google Ads that has probably convinced more advertisers to spend money than almost anything else in the platform.

“Limited by budget.”

Then comes the suggestion: increase your budget to capture more traffic.

It looks important. There is a warning attached to it. There are apparently more impressions available. More clicks are sitting there, just out of reach.

The natural reaction is to think, “Maybe we should increase the budget.”

And sometimes you should.

But the notification is not telling you that increasing your budget is good for the business. It is telling you that Google believes there is more auction volume available.

Those are very different things.

Google does not know what your profit margin looks like. It does not know your actual cost per acquisition target. It does not know whether the leads coming through your campaign are any good, whether your sales team can close them, whether customers refund their purchases, or whether one product makes twice as much money as another.

It is looking at the advertising system from the advertising system’s point of view.

There are searches available. Your campaign could enter more auctions. Your budget is restricting how much you participate.

So Google recommends spending more.

That is a perfectly reasonable recommendation from Google's perspective.

It is not necessarily a reasonable business decision.

The recommendation is about volume, not profitability

Imagine a campaign that is converting at 1%.

The average CPC is $2.

You spend $200 and receive 100 clicks. At a 1% conversion rate, that produces one sale.

Now Google tells you that the campaign could capture more traffic if you increase the budget.

So you double the budget to $400.

You can now buy roughly 200 clicks.

If nothing else changes, you might get two sales.

You spent twice as much and got twice as many sales.

That is not really optimization. You have simply purchased more of the same outcome.

And if the additional traffic is less qualified, the situation can become even worse.

This is where advertisers sometimes get confused. They see more spend, more clicks, and perhaps even more conversions in the account, and assume the campaign is growing.

But growth in conversion count is not the same thing as improvement in performance.

If the underlying conversion rate, customer quality, margins, and acquisition economics have not improved, increasing the budget has done very little beyond increasing the size of the machine.

Google cannot see the parts of the business that matter most

This becomes even more obvious when you look at what happens after the conversion.

Google may record a lead.

But it does not automatically know whether that lead was qualified.

It may record a purchase.

But it does not necessarily know whether that customer will refund the order.

It may see ten conversions from two campaigns.

Your business may know that eight of those conversions came from high-margin products while the other two barely made money.

Your sales team may know that one campaign produces leads that almost never close.

Your finance team may know that the acceptable acquisition cost is $80, while the advertising account is celebrating conversions at $140.

Those differences matter enormously to the business.

They are not always visible in the same way inside Google Ads.

That is why treating every platform recommendation as business advice is dangerous.

Google is very good at telling you what can happen inside Google.

You still have to decide whether it should happen.

More budget can make a bad account more expensive

I have seen the same pattern repeatedly.

An account is struggling.

Someone notices the “Limited by budget” message.

The daily budget goes up.

Spend rises almost immediately.

CPC may creep up. Click volume increases. CTR stays roughly where it was. Conversion rate remains flat, or sometimes gets worse.

Then comes the obvious question:

“Why didn't performance improve?”

Because the budget was never the problem.

The account was simply being given more money to operate with the same underlying problems.

Maybe broad search terms were bringing in people with little buying intent. Maybe negative keywords had not been maintained properly. Maybe the landing page was slow on mobile. Maybe the offer was unclear. Maybe conversion tracking was double-counting actions. Maybe offline sales were never being imported, so the account was optimizing toward conversions that did not actually represent business value.

In a Performance Max campaign, the problem might be weak signals, poor asset coverage, or simply a lack of useful information about what a valuable customer actually looks like.

Giving that account another $50 a day does not fix any of those things.

It just gives the inefficiency a larger budget.

The first thing I want to know is where the money is going

Before increasing spend, I would rather look at the account than look at Google's recommendation.

Start with the search terms.

Are people actually searching for what you sell?

A campaign can have excellent-looking keywords and still spend money on searches that are only loosely related to the business. If irrelevant traffic is getting through, increasing the budget simply gives those searches more opportunities to consume it.

Then look at the cost per qualified lead, rather than stopping at the number of conversions.

Ten leads sounds better than five until you discover that the five leads from the smaller campaign actually became customers and the other five did nothing.

Tracking deserves the same attention.

Is the conversion action firing correctly? Is the same customer being counted more than once? Are you measuring the action that actually matters, or simply the easiest action for Google to record?

For ecommerce accounts, revenue needs context too. A campaign generating $10,000 in revenue is not automatically better than one generating $7,000. Margins, product mix, returns, and acquisition costs can completely change the picture.

And then there is the landing page.

If someone clicks an ad and lands on a slow page with a vague offer and no obvious reason to trust the business, the answer is not always “send more people there.”

Sometimes the answer is to fix the page.

Small improvements can change the economics completely

Go back to that campaign converting at 1%.

At $2 per click, 100 clicks cost $200 and produce one sale.

Now suppose the targeting gets tighter, the search intent improves, the landing page communicates the offer more clearly, and the tracking is cleaned up.

The conversion rate moves from 1% to 2%.

The same 100 clicks can now produce two sales.

Nothing about the budget had to change.

You simply got more value from the traffic you were already paying for.

This is the part that gets lost when advertisers focus too heavily on impression share and available traffic.

There is always more traffic somewhere.

There are always more auctions you could enter.

There are always more clicks you could buy.

The harder question is whether buying them makes economic sense.

Scaling should come after the account earns the right to scale

There is nothing inherently wrong with increasing a Google Ads budget.

In fact, there are situations where it is exactly the right move.

If the campaign has clean targeting, reliable tracking, strong conversion data, healthy economics, and a clear path to additional profitable demand, then restricting the budget can absolutely limit growth.

But that is a different situation from seeing “Limited by budget” on a campaign and immediately treating it as an instruction.

Before scaling, I want to know that the structure is clean.

I want to know that the traffic is relevant.

I want to know that the conversions being reported are real and useful.

I want to know what happens after the conversion.

I want to know what the landing page is doing with the traffic.

And, most importantly, I want to know whether the economics still make sense when we spend more.

Google's recommendation answers one question:

Can this campaign technically spend more money?

Your business needs to answer another:

Should it?

Those are not the same question.

A recommendation for more traffic is not a recommendation for more profit.

So the next time Google tells you that your campaign is “Limited by budget,” don't panic and reach for the budget slider.

Look at the account first.

Fix the structure. Verify the data. Improve the conversion path. Make sure the traffic is worth buying.

Then, if the numbers support it, increase the budget.

That is scaling.

Everything else is just spending more.

An Ecommerce Manager Should Do More Than Manage Shopify

Katalaga Hakim — Sat, 05 Sep 2026 00:00:00 GMT

"Do you know Shopify?"

It is probably one of the easiest questions you can ask when hiring an Ecommerce Manager.

It is also one of the least useful.

Someone can know how to add products, edit a theme, create collections, install apps, update prices and change images without having much understanding of ecommerce at all.

They can be perfectly competent inside Shopify and still miss the reason sales dropped last week.

The problem is that Shopify is only one part of the system.

An ecommerce manager needs to understand what happens around it.

Traffic comes in. Tracking tells you what happened. Merchant Center tells Google what you sell. Ads create demand. The store has to convert that demand. Orders have to be fulfilled. Customer behaviour creates more data. Revenue eventually tells you whether all of those pieces worked together.

If the person managing the store only understands the Shopify admin, they are managing one screen.

They are not necessarily managing ecommerce.

The Shopify dashboard doesn't tell you everything

Imagine an online store suddenly sees a drop in reported purchases from Google Ads.

A task-oriented Shopify manager might start looking at products, discounts, checkout settings or the theme.

Those things could be responsible.

But the problem could also be somewhere else.

Maybe the purchase event stopped firing correctly.

Maybe Google Tag Manager was changed.

Maybe GA4 is recording fewer purchases than Shopify.

Maybe the Merchant Center feed has product errors.

Maybe a campaign started spending more on products with poor margins.

Maybe the ads are still generating clicks, but the landing pages aren't converting.

Maybe nothing is technically broken at all. The traffic mix simply changed.

You need someone who can investigate the entire chain before deciding what to change.

That is the difference between knowing Shopify and understanding the ecommerce system built around Shopify.

Tracking is part of the job

This is one of the biggest gaps I see in the way Ecommerce Manager roles are defined.

Tracking is often treated as something the analytics person or media buyer handles.

But if you're responsible for ecommerce performance, you need to understand enough about the data layer to know when the numbers don't make sense.

You don't necessarily need to be a specialist in every part of GA4 or Google Tag Manager.

You do need to be able to investigate questions such as:

Why did Shopify record 100 orders while GA4 recorded 72?

Why does Google Ads show a different amount of conversion value?

Did the purchase event actually fire?

Was something changed recently?

Is this a normal attribution difference or a genuine tracking problem?

Which number should the business use when making a decision?

Those are not theoretical questions.

If the data is wrong, everything downstream becomes harder.

The media buyer might optimize toward bad information. The Ecommerce Manager might report the wrong performance. The owner might increase a budget because a dashboard says revenue is growing when the underlying numbers don't support it.

You don't need an Ecommerce Manager who knows every implementation detail.

You need one who notices when the numbers stop making sense and knows where to look next.

Merchant Center is part of ecommerce management too

Google needs to understand what you're selling.

For a Shopify store running Shopping or Performance Max campaigns, the product feed is one of the most important connections between the store and Google.

Product titles matter.

Descriptions matter.

Product identifiers matter.

Categories and attributes matter.

Images matter.

Availability and pricing matter.

Feed errors matter.

A manager who understands this can look at an advertising problem and ask a better question.

Instead of immediately saying, "We need to increase the budget," they might ask whether Google is receiving the right product information in the first place.

That distinction can save a lot of wasted spend.

Your store may have excellent products and a perfectly reasonable advertising budget, but if Google doesn't understand the products correctly, you're making the advertising system work with bad information.

An Ecommerce Manager should understand that connection.

They should understand what happens after the click

This is another reason the role can't stop at Shopify administration.

Suppose Google Ads is generating traffic but sales are weak.

The obvious reaction is to blame the ads.

Sometimes the ads are the problem.

Sometimes they aren't.

Maybe the search terms are relevant and the traffic is qualified, but the product page is weak.

The product title might not immediately explain what the product is.

The images might not answer the questions customers have.

The description might be full of generic marketing language.

Reviews might be difficult to find.

Shipping information might be unclear.

The page might be slow.

The call to action might be buried.

The offer might simply be uncompetitive.

An Ecommerce Manager needs enough CRO knowledge to recognize these problems.

Otherwise, every conversion problem becomes an advertising problem.

And that can get expensive very quickly.

They should be able to read advertising data

You don't necessarily need your Ecommerce Manager to be the person running every Google Ads campaign.

But they should understand what the numbers mean.

They should know the difference between CPC, CTR, conversion rate, CPA, CAC, ROAS and conversion value.

More importantly, they should understand how those numbers connect.

If CPC rises, what happened to conversion rate?

If ROAS falls, which products or campaigns caused it?

If a campaign is generating sales, are those sales actually profitable?

If one product has a much higher ROAS than another, should it receive more budget?

If a campaign suddenly starts spending heavily, where is that money going?

If search terms are irrelevant, how much budget is being wasted?

An Ecommerce Manager doesn't need to live inside Google Ads all day.

But they should be able to sit down with the advertising data and understand what it is telling them.

Otherwise, they're dependent on somebody else to explain the business to them.

The best Ecommerce Managers connect the systems

This is what gets missed when companies write job descriptions around software.

They create a list:

Shopify.

GA4.

Google Ads.

Merchant Center.

GTM.

Email.

Maybe Klaviyo.

Maybe Meta Ads.

Then they look for someone who has touched all of those platforms.

But knowing the platforms isn't the real skill.

Understanding the connections between them is.

A customer clicks an ad.

That click needs to be tracked.

The customer lands on a product page.

The page needs to persuade them to buy.

The purchase needs to be recorded correctly.

The order needs to reach the fulfillment process.

Revenue needs to appear in the appropriate reporting systems.

That purchase then becomes information that can influence future advertising and marketing decisions.

One system feeds another.

An Ecommerce Manager should understand that chain well enough to identify where something has gone wrong.

This is why hiring tests should change

If you're hiring someone for an Ecommerce Manager position, asking "Do you know Shopify?" doesn't tell you very much.

Almost anyone with enough exposure to Shopify can learn the interface.

A better test is to give them a problem.

For example:

Your Shopify store recorded 150 purchases last month. Google Ads reports 121. GA4 reports 98.

What would you investigate first?

Or:

Google Ads spending increased by 30%, but revenue hasn't moved.

What would you look at?

Or:

A collection page has started receiving significantly more organic impressions but clicks haven't increased.

What could be happening?

Or:

Your Performance Max campaign is spending heavily on a group of products that have poor margins.

What would you do?

Those questions reveal much more than asking someone whether they know how to create a Shopify collection.

You're testing whether they can investigate.

Whether they understand relationships between systems.

Whether they can distinguish symptoms from causes.

And whether they know what should happen next.

That last part matters more than people think.

A manager shouldn't need to be told what to look for

There is a big difference between completing tasks and managing a business function.

A task-oriented person waits for instructions.

"Update these products."

"Change this banner."

"Create this collection."

"Upload these images."

"Check this order."

An Ecommerce Manager should be able to notice things without being asked.

Organic traffic is falling on an important collection.

A product feed has errors.

A campaign's cost per purchase is increasing.

A high-traffic product page has a poor conversion rate.

Tracking numbers don't reconcile.

A product is getting ad spend without producing enough revenue.

A previously strong campaign has changed behaviour.

These are not tasks someone should necessarily assign every morning.

They are things a manager should be looking for.

The role is partly about maintaining the store, but it is also about paying attention to the business.

The real question is whether they understand the machine

Shopify is important.

But Shopify is not the business.

It is one of the systems the business runs on.

The store sits between traffic and revenue. Advertising creates demand. SEO creates visibility. Merchant Center communicates product information to Google. Analytics and tracking tell you what happened. The product pages turn visitors into customers. Operations get the order out the door. Customer behaviour tells you what worked.

The Ecommerce Manager sits in the middle of all of that.

They don't have to personally perform every job.

They do need to understand enough about each part to know how the pieces interact.

That is what makes the role valuable.

So when you're hiring your next Ecommerce Manager, don't stop at:

"Do you know Shopify?"

Ask what happens when the numbers don't match.

Ask how they would investigate wasted ad spend.

Ask how they would diagnose a tracking problem.

Ask what they look for in a Merchant Center feed.

Ask how they decide whether a conversion problem is coming from the traffic or the store.

Ask what they would check if revenue suddenly dropped.

And most importantly, ask:

"You notice something is wrong, but nobody has told you what to fix. What do you do next?"

That question will tell you much more about an Ecommerce Manager than a list of Shopify tasks ever will.

Because you don't really need someone who knows how to use Shopify.

You need someone who understands what Shopify is connected to, what those connections are supposed to do, and how to find the problem when something breaks.

Otherwise, you've hired someone to complete Shopify tasks.

Not someone to manage ecommerce.

Why ChatGPT Can't Replace Semrush or Ahrefs for SEO

Katalaga Hakim — Sun, 09 Aug 2026 00:00:00 GMT

A lot of people who are not actually doing SEO seem to have discovered a shortcut.

They'll tell you that you don't need Semrush or Ahrefs anymore because ChatGPT, Claude, or some other AI tool can do everything those platforms do.

It sounds convincing if you don't spend much time inside SEO tools.

Ask someone who actually does SEO what they use Semrush or Ahrefs for, though, and the argument starts falling apart pretty quickly.

The problem is that people are confusing an AI assistant with a data platform.

ChatGPT can help you think through an SEO strategy. It can explain search intent, help organize keywords, suggest content structures, analyze information you give it, and make a lot of the work around SEO faster.

But that is very different from having a large, continuously updated database of search results, backlinks, keywords, competitors, rankings, and other SEO signals.

That distinction matters.

The backlink data alone is difficult to replace

One of the first things an SEO tool gives you is visibility into the link profile of a website.

You can enter a competitor's domain and start investigating who links to them, which pages attract links, what kind of links they have, which referring domains they have gained, and how their authority compares with other sites.

That changes the questions you can ask.

Instead of saying, "I think this competitor is ranking because they have good content," you can investigate what is actually happening.

Maybe their category pages have attracted links from industry publications. Maybe one particular guide has accumulated hundreds of referring domains. Maybe their strongest pages are receiving internal links from almost everywhere on the site.

You can then look at the sites linking to them and ask whether there are realistic opportunities for your own site.

ChatGPT cannot simply know all of that because you asked it to.

You need the underlying data.

Semrush and Ahrefs exist partly to collect, process, and organize that data so an SEO can investigate it.

You need to know what is actually ranking

Keyword research is not just coming up with a list of phrases that sound relevant to a business.

The important question is what Google is actually returning for those searches.

Search results tell you a lot about intent.

If you search for a keyword and the first page is dominated by category pages, product pages, comparison pages, or informational guides, that gives you clues about what Google believes the searcher wants.

SEO tools make it much easier to investigate those patterns at scale.

You can see which keywords a competitor ranks for, which pages are bringing them traffic, where they rank, how those rankings have changed, and which terms they may be close to winning.

That is much more useful than asking an AI model, "Give me 100 keywords for my ecommerce store."

The latter gives you ideas.

The former gives you evidence.

Those are not the same thing.

Keyword research needs data

This is another area where people misunderstand what SEO keyword research actually involves.

You don't want keywords simply because they are related to your product.

You want to understand demand, competition, intent, variations, existing rankings, and the relationship between different searches.

Suppose you sell running shoes.

An AI model can easily produce hundreds of phrases related to running shoes. That's useful as a starting point.

But an SEO doesn't stop there.

Which of those searches have meaningful demand? Which ones are already dominated by enormous sites? Which ones have commercial intent? Which variations are essentially the same search? Which keywords does a competitor rank for that you don't? Which terms are driving traffic to their product and category pages?

That is where a database becomes valuable.

The tool isn't valuable because it can magically tell you the perfect keyword.

It is valuable because it gives you enough data to make a better decision about which opportunities are worth pursuing.

The real value is finding opportunities

This is probably the biggest misunderstanding.

Good SEO is not about collecting as many keywords as possible.

It is about finding opportunities.

A competitor might be ranking on page two for hundreds of keywords with a single page. That could tell you there is an opportunity to build a better page.

You might discover that competitors have strong content around a topic but weak commercial pages behind it. That could reveal an opportunity to build a collection or product page that better matches the search.

You might find that a competitor has links pointing to an outdated article while you have a substantially better resource. That changes the way you approach link building.

You might discover that several competitors rank for the same group of related searches while your site has nothing targeting them.

Those are actionable findings.

They come from comparing actual search and competitive data.

AI can help you interpret those findings. It can help you organize them, prioritize them, turn them into briefs, write outlines, and even help you decide what to investigate next.

But it still needs something to work with.

This is where the "AI replaces SEO tools" argument gets confused

There is a perfectly reasonable argument that AI can replace some of the manual work SEO professionals used to do.

That's true.

If you previously spent an hour manually grouping keywords, summarizing competitors, creating content briefs, or turning research into a plan, an AI assistant can make parts of that process considerably faster.

That's a productivity improvement.

It doesn't mean the underlying data is no longer necessary.

A calculator didn't make mathematics useless. Spreadsheets didn't make accountants unnecessary. And ChatGPT didn't make SEO data disappear.

The tools solve different problems.

Semrush and Ahrefs are useful because they provide access to SEO data that you can investigate.

ChatGPT and Claude are useful because they can help you reason about information, generate ideas, organize research, explain concepts, and speed up execution.

Put them together and they become much more useful than either one on its own.

You can use AI without pretending it is an SEO database

This is the part I wish more people understood.

If you're doing SEO for a real business, you don't need to choose between AI and traditional SEO tools.

Use the SEO tools to collect the evidence.

Then use AI to help you work with that evidence.

Export keyword data and have AI help cluster it by intent.

Give it competitor findings and ask it to identify patterns.

Give it a list of ranking pages and ask it to help build a content brief.

Use it to compare opportunities and organize your priorities.

Use it to turn research into something your team can actually execute.

That is a much more useful application of AI than pretending you can throw away the data and simply ask a chatbot what to do.

Because SEO is still an evidence-based discipline.

You need to know what people are searching for. You need to understand what is ranking. You need to understand what your competitors are doing. You need to understand where links are coming from. You need to identify gaps and opportunities.

An AI model can help you make sense of that information.

It doesn't remove the need for the information itself.

So when someone tells you that ChatGPT or Claude has completely replaced Semrush or Ahrefs, ask them a simple question:

Where is your SEO data coming from?

If the answer is essentially, "I asked ChatGPT," that's not a new SEO workflow.

It's a different way of guessing.

And there is a very big difference between using AI to make SEO faster and using AI as a substitute for the evidence SEO decisions are supposed to be based on.

Installing GA4 Doesn't Mean Your Tracking Is Working

Katalaga Hakim — Thu, 30 Jul 2026 00:00:00 GMT

Setting up Google Analytics on a website looks easier than it actually is.

You create the property, create a web data stream, get a Measurement ID, paste the Google tag into the website, and expect the numbers to start appearing.

Then you open Analytics and see a message telling you that no data has been received.

That is where people usually start wondering whether they did something wrong.

Sometimes they did. Sometimes the tag is working perfectly and Google Analytics simply has not processed everything yet. The important part is knowing the difference.

Take a new ecommerce site such as PheromoneParfums.com. The first step is not installing some mysterious piece of tracking code. It is creating the right GA4 property and connecting the website to it properly.

Inside Google Analytics, you create a new property under the correct Analytics account. Give it a name that makes sense, choose the appropriate reporting time zone and currency, and complete the business information Google asks for.

For an ecommerce business, the property should be configured around what you actually need to measure. Page visits are useful, but eventually you will want to know which products people viewed, what they added to their cart, what they purchased, how much revenue those purchases generated, and where those customers came from.

After creating the property, you create a Web data stream. This is the connection between the website and the GA4 property.

Google gives that stream a Measurement ID beginning with G-. That ID is important, but it is not the tracking itself. It is the identifier that tells Google Analytics which property should receive the data.

This is where a common misunderstanding begins.

Copying the Measurement ID is not the same thing as installing Google Analytics.

The website has to actually load the Google tag when someone visits it.

If you are working with a website where you have access to the code, Google provides a Google tag that can be installed in the site's <head> section. If the website runs on a platform such as Shopify or WordPress, there may be a better platform-specific method through an integration, plugin, app, or tag management setup.

The method matters because putting the code in the wrong place can leave you with a perfectly configured GA4 property that receives nothing.

And that is why the next step should always be verification.

The Google Analytics screen can be misleading

One of the easiest mistakes to make is looking at the main Analytics dashboard immediately after installing the tag and assuming that the numbers should already be there.

GA4 has different reporting areas, and they do not necessarily update at exactly the same time.

Realtime reporting is particularly useful when testing a new installation. Open the website in a separate browser window and visit a few pages. Then open the Realtime report in GA4.

If your visit appears there, you have strong evidence that the tag is communicating with the property.

This is much more useful than simply staring at the main dashboard waiting for numbers to change.

A new installation can also produce a confusing situation where the Analytics interface still displays a message such as "No data received" while Realtime already shows an active user.

That does not automatically mean the installation has failed.

The first thing to check is whether the website is actually sending data.

Test the website, not just the Analytics dashboard

There is a simple way to approach this.

Open the website and then open GA4's Realtime report. Browse a few pages on the site. If the visit appears, the basic connection is working.

You can also inspect the website itself.

Open the page source and search for the Measurement ID. If the ID is present in the page's code, you know the tag has at least been added to the page. If it is completely absent, the problem is probably with the installation rather than GA4 itself.

Browser extensions and privacy settings can complicate testing as well. Ad blockers and other privacy tools can interfere with analytics requests, so a clean browser session can be useful when diagnosing a new setup.

The point is to test each part of the chain rather than assuming that one screen tells you everything.

Website → Google tag → GA4 property → Realtime data → standard reports.

If the chain breaks somewhere, you need to find the broken part.

Installing the tag is only the beginning for ecommerce

There is another important distinction for ecommerce websites.

Getting pageview data into GA4 does not mean you have finished your analytics setup.

A perfume store, for example, needs more than a record saying that somebody visited the homepage.

You eventually want to understand the commercial journey.

Someone views a product. They add it to the cart. They begin checkout. They purchase.

Those actions can be represented through ecommerce events such as view_item, add_to_cart, and purchase.

Once those events are implemented correctly, GA4 becomes much more useful because you can begin connecting website behaviour with actual commercial outcomes.

You can start asking better questions.

Which products are getting attention but not purchases?

Where are people abandoning the buying process?

Which traffic sources are bringing customers rather than just visitors?

Which products generate the most revenue?

Those are the questions that make analytics valuable.

Installing the basic Google tag only gives you the foundation.

Where Google Tag Manager fits

For a small website with basic requirements, directly installing the Google tag may be enough.

But once you start tracking more things, the setup can become messy.

You might want to track form submissions, important button clicks, ecommerce events, advertising platforms, remarketing tags, and other marketing tools. Adding each piece directly into the website's code can turn tracking into a maintenance problem.

That is where Google Tag Manager can become useful.

Instead of repeatedly changing the website whenever you want to add or modify a tracking tag, you can manage many of those tags through one central system.

That does not mean every website needs a complicated tag-management setup. It means the tracking architecture should match what the business actually needs.

A five-page brochure website does not need the same measurement system as a growing ecommerce store.

The real mistake is stopping after installation

The biggest problem with analytics setup is not usually creating the property.

It is assuming the job is finished once the code has been pasted.

A proper installation has several stages.

You create the correct property. You create the web stream. You install the tag using the appropriate method. You verify that the website is sending data. You check Realtime. Then you make sure the events and ecommerce tracking you actually care about are being recorded correctly.

Only then do you have something worth relying on.

And this matters because bad analytics can be worse than no analytics.

If you believe Google Ads generated 120 purchases when the tracking is actually recording something incorrectly, you may make budget decisions based on numbers that were never accurate in the first place.

The same applies to SEO, email, social traffic and every other channel you are trying to evaluate.

Analytics should not simply produce numbers. It should give you confidence that the numbers represent what is actually happening on the website.

So when you install GA4, don't stop when the tag is pasted.

Make sure the website is sending data, make sure GA4 is receiving it, and then make sure it is recording the events that matter to the business.

That is the difference between having Google Analytics installed and actually having useful analytics.

Claude Can Do a Lot for SEO. It Still Needs the Data.

Katalaga Hakim — Sat, 11 Jul 2026 00:00:00 GMT

There is a strange argument happening in SEO right now.

People are discovering that ChatGPT and Claude can do an impressive amount of SEO work, and some have jumped from that observation to a much bigger conclusion: you don't need Ahrefs anymore.

I don't think that's the lesson.

The interesting part isn't that Claude can replace pieces of an SEO workflow. It absolutely can. The interesting part is what happens when you give the AI access to the data that an SEO tool has spent years collecting.

That is where things start getting useful.

Pair Ahrefs with Claude through its MCP, then bring in your own Google Search Console data, and you have something much more powerful than either tool operating alone.

Ahrefs gives you information about the search landscape. Search Console tells you what is actually happening on your site. Claude can sit between the two and help you make sense of it.

That's a very different proposition from asking ChatGPT to "do SEO."

Your website already has a lot of SEO data

Google Search Console is one of the most useful sources of information an ecommerce business already has.

It can show you the queries people are using to find your site, the pages receiving impressions, clicks, average positions and the differences between what Google is showing and what users are actually clicking.

Export that data and give it to Claude.

Now the conversation changes.

Instead of asking an AI to guess which keywords you should target, you can ask it to look at the queries your site is already appearing for.

You can find pages sitting just outside the strongest positions. You can identify queries generating impressions but very few clicks. You can spot pages ranking for terms that aren't properly reflected in their titles or content.

That is much more useful than generating a list of 100 hypothetical keywords.

The data is coming from your site.

Then bring in Ahrefs

This is where the workflow gets even more interesting.

Ahrefs gives you a different view.

You can investigate keyword difficulty, competing pages, backlinks, referring domains, competitor rankings and the broader search landscape around a topic.

So now you have two datasets answering different questions.

Search Console can tell you:

What is happening on my website?

Ahrefs can help answer:

What is happening around the search terms, competitors and pages I'm trying to compete with?

Claude can help connect those answers.

That connection is the valuable part.

Imagine you have a collection page that is already getting thousands of impressions but isn't receiving many clicks.

Search Console gives you the actual query and performance data.

Ahrefs can help you understand the competitive environment around those terms.

Claude can then help you work through the problem: Is the page targeting the wrong intent? Is the title weak? Is another page on the site a better candidate? Are competitors covering something that your page doesn't? Does the page need better internal links or more authority?

You're no longer asking AI to invent an SEO strategy from scratch.

You're giving it evidence and asking it to reason through the evidence.

That's a much better use of AI.

AI is the analyst. The tools are the evidence.

This is the distinction that gets lost in the "ChatGPT replaced Ahrefs" conversation.

An AI model is very good at working with information.

It can compare datasets, find patterns, organize messy exports, explain why something might be happening, create content briefs, identify opportunities and turn a pile of SEO data into something a human can actually work with.

But it doesn't magically know your site's search performance.

It doesn't automatically know every keyword your competitors rank for.

It doesn't have a live, complete picture of your backlink profile simply because you asked it to analyze your SEO.

Those things require data.

That's what tools such as Ahrefs are for.

And once you have that data, AI becomes considerably more useful.

The best workflow isn't Ahrefs versus Claude.

It is Ahrefs + Search Console + Claude.

This changes how keyword research should be done

Traditional keyword research often starts with a large keyword list.

You enter a seed term, collect hundreds or thousands of related terms, filter them by volume and difficulty, then start deciding what to target.

There is nothing inherently wrong with that.

But your own Search Console data gives you something a generic keyword database cannot: evidence that Google already considers your website relevant to certain searches.

That is valuable.

Suppose your store sells a particular type of product and you discover that Google is already giving one of your collection pages impressions for several related searches.

Those searches might not have been in your original keyword plan.

Instead of ignoring them because they weren't part of the initial research, you can investigate them.

Are they commercially relevant?

Does the collection page satisfy the intent?

Is there another page that should be ranking?

Could the page be improved?

Could internal links strengthen it?

Is there a separate collection or product category hiding inside that demand?

Now keyword research becomes less about finding random opportunities and more about understanding the opportunities already appearing around your business.

Ahrefs can then help you determine whether there is a larger opportunity beyond what your own site is currently seeing.

That's a much more informed process.

The same applies to content

This is also where AI can save a lot of time without turning SEO into content spam.

You don't need Claude to produce another generic article because a keyword tool says there is search volume for a phrase.

First, look at what your site is already appearing for.

Then look at the competing pages.

Then decide whether there is actually a content gap worth filling.

Maybe you discover that your product pages are getting impressions for questions customers have about choosing between two products.

That could be a useful buying guide.

Maybe a collection page is appearing for a broader informational query but doesn't answer an important part of the search.

That might tell you what supporting content the site needs.

The data helps you decide what deserves to be created.

Claude can then help you research, structure and draft it.

That is very different from publishing 50 AI articles because a spreadsheet told you that 50 keywords exist.

And this is where ecommerce SEO gets particularly interesting

For ecommerce stores, SEO decisions shouldn't be made around traffic alone.

You want to know which pages are close to generating commercial value.

Search Console can show you which product and collection pages are already getting visibility.

Ahrefs can help you understand the authority and competition around the searches those pages are targeting.

Your analytics and store data can tell you what happens after the click.

Now you can start connecting the pieces.

A page might have relatively modest traffic but attract highly commercial searches.

Another page might receive considerably more organic traffic but almost no meaningful business.

Which one deserves more attention?

That is not a question that keyword volume can answer by itself.

And it certainly isn't a question that an AI model should answer by guessing.

You need the underlying data.

The real advantage is the combination

This is why I don't buy the idea that using ChatGPT or Claude means you no longer need SEO tools.

The better way to think about it is that AI makes those tools more useful.

Before, you might spend hours exporting data, filtering spreadsheets, comparing competitors and trying to work out what matters.

Now you can give the data to an AI system and have it help you investigate it much faster.

The human still needs to make the important decisions.

Is this keyword actually relevant to the business?

Should this be a collection page, product page or informational article?

Is the competitor genuinely better, or just more authoritative?

Is this ranking opportunity worth pursuing?

Will improving this page actually help the business?

AI can help you get to those questions faster.

It shouldn't remove the need to ask them.

Don't replace the tools. Change how you use them.

The irony of the "AI replaced Ahrefs" argument is that it often misunderstands what makes AI powerful in the first place.

Claude becomes much more useful when it has access to real information.

Give it your Search Console export.

Give it your Ahrefs data.

Give it your site's structure.

Give it information about your products and customers.

Now you're not asking a language model to pretend to be an SEO tool.

You're giving an AI system the raw material required to help you think like an SEO.

That's the workflow I'd rather build.

Use Search Console to understand your own search performance.

Use Ahrefs to understand keywords, competitors and authority.

Use Claude or ChatGPT to analyze, compare, organize and turn that information into decisions.

Then actually make the changes.

That's how AI fits into SEO.

Not by throwing away the tools that provide the evidence, but by using AI to get more value from the evidence you already have.

An Ecommerce Manager Should Be Looking at the Whole Business

Katalaga Hakim — Sun, 24 May 2026 00:00:00 GMT

There is a strange way some businesses define an Ecommerce Manager.

They give someone access to Shopify, ask them to upload products, change a few banners, update collections, and occasionally check whether the orders are coming in.

Technically, the store is being managed.

But nobody is really managing the ecommerce business.

A growing store is a chain of things that depend on one another. Traffic comes in, people land on the store, tracking tells you what they did, products get pushed into Merchant Center, ads send more traffic, product pages turn some of that traffic into customers, and eventually all of it shows up as revenue.

If something breaks somewhere in that chain, the problem doesn't always announce itself.

You might have plenty of traffic but a terrible conversion rate. Google Ads might report a healthy ROAS while the products being sold have terrible margins. Merchant Center might say the feed is approved while half the product titles are poorly structured. A product page might be getting organic traffic every day and still be doing almost nothing for the business.

This is why an Ecommerce Manager needs to look beyond Shopify.

They should be checking the whole path:

Traffic → Store → Tracking → Feed → Ads → Conversion → Revenue.

And not once when they start the job.

Regularly.

That means looking at Shopify's conversion data and understanding where the money is actually coming from. It means checking GA4 and GTM to make sure the numbers being reported are worth trusting. It means looking through Google Ads rather than simply checking whether the account has a nice-looking ROAS.

It means looking at Merchant Center and the product feed.

It means knowing what is happening inside Shopping and Performance Max, checking search terms, finding wasted spend, and understanding which products are actually getting the budget.

Then you go back to the store.

Because sometimes the problem you are trying to solve in Google Ads isn't a Google Ads problem at all.

Maybe people are clicking the ad and landing on a product page that doesn't explain the product properly.

Maybe the collection structure is confusing.

Maybe the important information is buried.

Maybe the mobile experience is irritating enough that people simply leave.

Maybe the product has a healthy click-through rate and plenty of traffic but almost nobody buys it.

You won't find that by staring at the Google Ads dashboard.

You have to follow the customer.

And this is where ecommerce management becomes much more interesting.

The numbers start talking to each other.

Suppose Google Ads tells you a campaign has a 6x ROAS. That sounds great until you look at Shopify and discover that most of those sales are coming from products with very little margin.

The campaign isn't necessarily "bad."

But the business isn't necessarily making as much money as the dashboard suggests either.

Or perhaps Merchant Center tells you everything is approved. Green checks everywhere. Nothing appears broken.

That doesn't mean the feed is good.

Approval only tells you that the products have cleared Google's requirements. It doesn't mean your titles are well structured, your product information is useful, or that you've given Google the clearest possible understanding of what you're selling.

The same thing happens with SEO.

A product page can rank.

It can receive impressions.

It can get clicks.

And still be a terrible sales page.

There is nothing contradictory about that. Google can send someone to a page because it is relevant to their search. Whether that page persuades them to buy is a different question.

This is why I don't think an Ecommerce Manager should be judged by how many products they uploaded or how quickly they can change something in Shopify.

Those are tasks.

The real job is figuring out what is happening to the business and what needs attention next.

Where are we losing money?

What's working?

What isn't?

Which products deserve more attention?

Where are customers dropping off?

What can we scale?

And perhaps the most important question:

What should we fix first?

Because there is always something.

You can spend the entire day fixing small things on a store and still avoid the problem that is costing the business the most money.

Maybe the tracking is wrong.

Maybe the feed is holding Shopping back.

Maybe you're paying for clicks that have almost no chance of converting.

Maybe your best product isn't getting enough exposure.

Maybe the traffic is fine and the product page is the problem.

Maybe the store converts well but the economics of the business don't work.

These are very different problems, and they require very different solutions.

That's why ecommerce management is much bigger than keeping Shopify tidy.

A good Ecommerce Manager needs to understand how the different parts of the business connect, follow the numbers from the first click all the way to the actual sale, and keep looking for the point where things are leaking.

Sometimes that means fixing the store.

Sometimes it means fixing the tracking.

Sometimes it means cleaning up the feed.

Sometimes it means cutting wasted ad spend.

Sometimes it means pushing a product that everyone else has overlooked.

And sometimes it means admitting that the campaign everyone thought was doing brilliantly isn't actually making the business much money.

The job isn't to keep everything busy.

It's to keep the business moving in the right direction.

That's a very different responsibility.

Your Developer Should Never Control Your Website

Katalaga Hakim — Thu, 30 Apr 2026 00:00:00 GMT

A few days ago, an auto shop contacted me because they wanted their WordPress website redesigned.

That part isn't unusual. Businesses redesign websites all the time. The interesting part came when I asked them for access.

They didn't have it.

Not the WordPress login. Not the hosting account. Not the domain account.

Nothing.

Their previous developer had been managing everything for them, including the Namecheap hosting and domain, and when the business asked for access, the developer simply refused to hand it over.

So now they had a website they had paid for, a domain they had paid for, hosting they had paid for, and a developer sitting between them and all of it.

They weren't happy.

And honestly, I understand why.

The reason they originally came to me was already frustrating enough. Their website had become unreliable. Sometimes it worked. Sometimes it didn't. They would send messages about problems and wait days for a response. When the developer eventually replied, the reported issue might still be there, or the "fix" would somehow create a different problem.

Then the developer would disappear again.

Eventually, they got tired of the whole arrangement and wanted someone else to take over. They wanted the site redesigned, properly maintained, and some of the features they'd been asking for added.

That should have been a fairly straightforward project.

Instead, before we could even start, we had to deal with the fact that they couldn't access their own website.

And this is something I think more business owners need to understand before they hire anyone to build or manage their website.

Your developer should have access to your website. They should not own the keys to it.

There is an important difference.

If I'm building a WordPress website for a client, I need access to WordPress. I may need access to the hosting account. I may need access to DNS, the domain, email settings, Cloudflare or whatever else the project requires.

But none of that means those accounts need to belong to me.

In fact, I prefer the opposite.

I tell clients to create the hosting account themselves.

If they need a domain, they buy the domain themselves.

If they need hosting, they open the hosting account themselves.

If they need a particular service, they register it in their own name and with their own payment details.

Then they invite me.

That might mean adding me as another user, giving me administrator access to WordPress, or providing the appropriate developer-level access to the hosting platform.

I can get in, do the work, configure everything, fix the problems, build the site and maintain it.

But the client still owns the house.

I'm just the person they've given a key to.

This sounds like a small distinction when you're sitting down to hire a developer and all you want is a website.

It isn't.

Imagine paying a contractor to renovate your physical shop and then discovering that the contractor kept the title to the building, the keys to the front door and the only copy of the alarm code.

You'd probably have a few questions.

Websites are easier to treat this way because they don't feel physical. A domain is just a name on a screen. Hosting is just another monthly bill. WordPress is something you log into through a browser.

But these things are still business assets.

Your domain is an asset.

Your hosting account is an asset.

Your website is an asset.

Your email accounts can be business-critical assets.

And the account that controls them should ultimately remain under your control.

A developer can leave.

A developer can become unresponsive.

A developer can close their business.

You can decide that you don't want to work with them anymore.

You can find another developer.

You can have a disagreement.

None of those things should leave you standing outside your own website waiting for somebody else to let you in.

That is the part I find most concerning about arrangements like this.

The client often doesn't realise there is a problem until they need access.

Everything works fine while the relationship is good.

The developer handles the hosting. They renew the domain. They manage WordPress. They make changes when asked.

The client doesn't think much about it because, well, why would they?

Then something goes wrong.

The developer stops replying.

The website goes down.

The business wants to move to somebody else.

And suddenly the person they need to contact is also the person standing between them and their own accounts.

Now the redesign isn't the first problem anymore.

Getting control back is.

And sometimes that can become a much bigger mess than it ever needed to be.

This is why I think businesses should get into the habit of separating ownership from access.

You own the account.

Your developer gets access to the account.

Those are two completely different things.

If you're hiring someone to build a WordPress website, create the hosting account yourself. Buy the domain yourself. Keep the recovery email under your control. Keep the billing information under your control. Turn on the security measures yourself.

Then give your developer what they actually need to do the job.

When the project ends, you can remove their access.

If you decide to hire someone else, you can invite the new developer.

Nothing has to be transferred from one developer's personal account to another developer's personal account because the business never lost control in the first place.

And there is another advantage to doing it this way: it makes the relationship healthier.

The developer doesn't have to hold your website hostage to prove that you still need them.

You don't have to worry about what happens if they disappear.

They get paid to do the work.

You keep ownership of the thing you're paying them to build.

That's how it should be.

There are plenty of good developers who will have no problem with this. In fact, a professional developer should generally be comfortable working within a client's accounts and being given the access necessary to do the job.

You shouldn't have to surrender ownership simply because someone knows more about WordPress than you do.

You wouldn't give your accountant ownership of your bank account because they know more about accounting.

You wouldn't give your mechanic ownership of your car because they know more about engines.

Your website deserves the same common sense.

So if you're a business owner about to hire a developer, ask one very simple question before anything gets built:

"Who owns the accounts?"

If the answer is you, good.

If the answer is "I'll set everything up for you," ask a second question:

"Can you set it up under my account and then give yourself the access you need?"

That small decision can save you an enormous headache later.

Because the best developer-client relationship isn't one where the developer controls everything.

It's one where the developer knows exactly what they're responsible for, the client knows exactly what they own, and neither person has to depend on the other for access to something that was paid for by the business.

Your developer should have the keys.

They just shouldn't own the building.

Manual CPC vs. Max Conversions: Which Google Ads Strategy Is Better for Ecommerce?

Katalaga Hakim — Wed, 22 Apr 2026 00:00:00 GMT

One of the most important decisions you can make in a Google Ads account is also one that gets surprisingly little attention: how you want Google to bid.

For ecommerce stores, that usually comes down to a question of control versus automation.

Do you set your bids manually with Manual CPC, or do you give Google more control with Max Conversions?

There isn't one answer that works for every store. I've seen both approaches work well. The mistake is assuming that switching to automated bidding automatically means better performance.

Sometimes it does.

Sometimes you simply end up paying more for the same customers.

Start With What the Store Can Already Do

If a campaign is already generating purchases consistently and Manual CPC is producing a strong conversion rate at a reasonable cost, I wouldn't rush to change it just because Google recommends an automated bidding strategy.

That recommendation isn't a business case.

The question is whether Max Conversions can acquire more customers profitably than the strategy you're already using.

This is particularly important for ecommerce because CPC itself isn't the goal.

You can have a campaign with a higher CPC and still make more money.

You can also have a campaign with an impressive conversion rate that loses money because the clicks are too expensive or the products have thin margins.

The numbers need to be looked at together.

A Higher CPC Doesn't Automatically Mean Better Performance

This is where automated bidding can become expensive without looking obviously broken.

Suppose Manual CPC is generating purchases at a healthy rate, while Max Conversions produces a similar purchase rate but pushes the average CPC significantly higher.

Google may be finding more opportunities to convert, but if those additional signals aren't translating into enough additional revenue or profit, you're simply paying more to get roughly the same result.

For example, imagine two campaigns bringing in customers at roughly the same purchase rate.

One pays $1.50 per click.

The other pays $2.25.

If the conversion rates and average order values are broadly similar, the second campaign has a problem. You're spending more to acquire essentially the same outcome.

The bidding strategy isn't inherently bad.

It just isn't giving the store enough additional value to justify the extra cost.

Ecommerce Changes the Equation

Lead generation often revolves around cost per lead.

Ecommerce has another layer.

You need to care about the economics of the order.

A useful way to think about the relationship is:

CPC → conversion rate → cost per purchase → average order value → margin → profitability

That's why obsessing over CPC alone can be misleading.

A $3 click isn't necessarily bad if it consistently produces profitable $150 orders.

A $0.50 click isn't necessarily good if the traffic rarely buys.

The job is not to get the cheapest click.

The job is to acquire profitable customers.

When Manual CPC Can Make Sense

Manual CPC can be useful when you already understand the account and have enough control over where your traffic is coming from.

You can control bids at the keyword or product level, be more selective about what you're willing to pay for certain searches, and avoid automatically allowing the system to push bids higher simply because it sees a potential conversion.

That control can be especially valuable when margins are tight.

If you know that a particular product cannot support a certain acquisition cost, you don't necessarily want the bidding system making aggressive decisions without considering the economics of the product.

Manual bidding can also make sense when you're still building your data and want to understand how the account behaves before handing more control to automation.

But it isn't universally better.

There are markets where competition is intense enough that being too conservative with bids simply means losing valuable auctions.

That's where testing becomes important.

When Max Conversions Can Make Sense

Max Conversions is useful when Google has enough conversion data to identify patterns and enough flexibility in the campaign to act on those signals.

The system can adjust bids based on signals you aren't manually evaluating for every auction.

Device.

Location.

Search context.

Audience signals.

Time.

And other auction-level information.

That can be difficult to reproduce manually.

But automation works best when you give it a good foundation.

If your conversion tracking is unreliable, your product pages are weak, your campaign structure is messy, or you're feeding Google poor product information, switching to Max Conversions isn't going to fix the underlying problem.

It can simply make the system spend money faster.

Competitive Markets Are Different

This is where there is no universal rule.

In highly competitive markets, aggressive bidding may be necessary simply to compete for valuable searches.

A conservative Manual CPC strategy might look great in a spreadsheet because the CPC is low, while quietly losing the auctions that matter most.

But that doesn't mean you should automatically switch everything to Max Conversions.

Test it.

You can run one approach against the other under controlled conditions, or gradually introduce automated bidding and compare what happens to the metrics that actually matter.

Look at purchases, cost per purchase, conversion value, average order value and return on ad spend.

Don't declare a winner because one strategy produced a lower CPC.

The Metric That Matters Is What the Click Produces

This is the part I would keep coming back to.

Google Ads can make CPC look like the main event because it is one of the easiest numbers to see.

It isn't.

For ecommerce, the question is what happens after the click.

If Manual CPC generates a similar purchase rate to Max Conversions but at a lower CPC, that difference can flow directly into your acquisition economics.

If Max Conversions produces more purchases, higher conversion value and a better return on ad spend despite the higher CPC, then the higher CPC may be completely justified.

That's why bidding strategies should be judged by business outcomes rather than isolated advertising metrics.

Don't Change Bidding Before Fixing the Foundation

Before deciding that Manual CPC or Max Conversions is the problem, check everything around it.

Is purchase tracking working correctly?

Are your product titles and descriptions giving Google enough information?

Is the Merchant Center feed clean?

Are products segmented properly?

Are you sending traffic to the right product pages?

Are your prices competitive?

Are shipping costs creating friction?

Are your best-selling products actually receiving enough budget?

If the answer to those questions is no, bidding strategy is probably not your biggest problem.

A bidding system can only work with the information and conversion signals it receives.

Give Google a weak foundation and automation won't magically turn it into a good account.

Test the Strategy Against the Economics

I don't think the right question is:

"Should I use Manual CPC or Max Conversions?"

The better question is:

"Which bidding strategy acquires profitable customers more efficiently for this store?"

That answer can change as the account grows.

A strategy that works when a store has limited conversion data may not be the strategy you want six months later.

A campaign that performs well with Manual CPC in one market might need more automation in another.

And a bidding strategy that looks expensive on a CPC report might turn out to be the most profitable option once you look at revenue and margin.

That's why I wouldn't treat bidding strategies as permanent account settings.

Test them.

Give each strategy enough data to be evaluated properly.

Then judge the result based on purchases and economics, not on which number looks nicest in the Google Ads interface.

For ecommerce, the cheapest click has never been the goal.

The profitable customer is.

Collection Pages are the most important part of an eccomerce site

Katalaga Hakim — Sat, 07 Mar 2026 00:00:00 GMT

I have seen plenty of Shopify stores with dozens, sometimes hundreds, of products and almost nothing useful for Google to understand.

The homepage looks fine.

The product pages exist.

There is a blog with a handful of articles.

And then you open the collection pages.

A heading.

A product grid.

Maybe a sentence or two.

Then the footer.

That is a strange way to build an ecommerce site, because collection pages are often some of the most commercially valuable pages on the entire store.

They sit between broad category searches and individual products. They can target people who already know roughly what they want, while still giving them enough information to decide which product to buy.

Yet many Shopify stores treat collections as little more than folders for products.

That is where the problem starts.

A collection should answer a question

Suppose you sell supplements.

You create a collection called “Magnesium Supplements.”

You could put 20 products on the page and leave it at that.

But what does Google learn from the page?

And more importantly, what does the customer learn?

A better collection page explains what the category actually is.

What is magnesium used for?

What is the difference between magnesium glycinate and magnesium citrate?

Who might prefer one form over another?

What are the common reasons people buy these products?

When might someone take them?

The point is not to write 400 words because some SEO checklist told you that 400 is the magic number.

There is no magic word count.

The point is to give the page enough useful context to satisfy the searcher.

If 300–400 words helps you explain the category properly, great.

If the topic needs less, use less.

If it needs more, use more.

The copy should exist because the page has something worth explaining, not because someone wanted to hit a word target.

The same applies outside supplements.

A “Modern Wall Art” collection can explain what makes a piece modern, which rooms it tends to work in, how to think about sizing, and the difference between framed prints and canvas.

Now the collection is doing more than displaying products.

It is helping someone make a purchase.

That is what useful SEO content should do.

The product grid is not the whole page

One of the easiest ways to spot an underdeveloped ecommerce collection is to look at the page above the fold.

You see the collection title.

Then products.

Then more products.

It feels like a warehouse catalogue.

A strong collection can be structured more deliberately.

There might be a short introduction explaining what the category contains, followed by the product grid.

Below that, useful supporting information can answer common questions, link to related collections, provide sizing or comparison information, and deal with objections that might otherwise send the customer back to Google.

The exact layout will depend on the store.

But the principle is simple.

The collection should feel curated rather than dumped.

A customer should understand what they are looking at and why these products belong together.

Sometimes the collection you need does not exist yet

This is where search intent becomes interesting.

Imagine a store with one large collection called “Running Shoes.”

It might contain road shoes, trail shoes, marathon shoes and wide-fit options.

From the store's perspective, that seems perfectly logical.

From a search perspective, it is much less useful.

Someone searching for trail running shoes has a different problem from someone looking for running shoes for wide feet.

Someone training for a marathon has different requirements again.

Those are different buying intentions.

So instead of forcing everything into one broad collection, you can create useful category pages such as:

Running Shoes for Wide Feet.

Trail Running Shoes.

Marathon Running Shoes.

Running Shoes for Flat Feet.

Now each collection has a clearer purpose.

You have created another doorway into the store.

This does not mean creating hundreds of nearly identical collections just to capture keywords. That usually produces a mess.

The collection needs to represent a real distinction in the products and the customer's intent.

If people genuinely search for it and your store genuinely serves that need, it probably deserves a page.

Then connect the pages

Creating useful collections is only half the job.

They also need to make sense together.

Take a Trail Running Shoes collection.

Somewhere on that page, it might make sense to mention that customers looking for road running shoes can browse the Road Running Shoes collection.

A Protein Powder collection might naturally connect to Shakers, Pre-Workout or Mass Gainers.

These links are useful to customers because they help them move around the store.

They are also useful to search engines because they establish relationships between related pages.

Over time, you start building a topical structure rather than a collection of isolated URLs.

This is where internal linking becomes much more interesting than simply sprinkling links into random blog posts.

Your collections should understand each other.

Keep important pages close to the surface

I like a simple rule for ecommerce architecture: important pages should not be buried.

If a valuable collection requires five clicks and several awkward navigation steps before someone can reach it, the site is telling you something.

It is not very important.

You do not need to put every collection into the main navigation, either.

That creates its own problem.

A header containing 40 categories is not a sophisticated information architecture. It is a list that nobody wants to use.

Use the homepage to feature important collections.

Use breadcrumbs.

Use contextual links.

Use the footer for useful category navigation.

Use related collections where they make sense.

The goal is not to make every page one click away.

The goal is to make important pages easy to discover.

For many stores, keeping key collections within roughly two or three clicks of the homepage is a sensible target.

FAQs can reveal what the collection is missing

A good FAQ section is not there because someone told you every page needs one.

It is useful when customers genuinely have questions.

Consider a collection for waterproof hiking boots.

A shopper might want to know:

Are they suitable for snow?

How waterproof are they?

Are they true to size?

Can they handle rocky terrain?

How long do they typically last?

Those questions tell you something about the searcher's concerns.

They also give the page an opportunity to answer long-tail queries naturally.

And sometimes the questions reveal that the collection itself needs better information.

If five customers keep asking about sizing, perhaps the page needs a sizing guide.

If people keep asking whether the boots work in snow, perhaps that information belongs much higher on the page.

SEO and CRO often end up pointing at the same problem.

The customer is telling you what the page is missing.

Add information that helps someone choose

There are plenty of small elements that can make a collection more useful without turning it into an enormous piece of content.

A comparison table can help someone understand the differences between products.

A quick sizing guide can remove uncertainty.

Shipping information can answer a common objection.

Trust signals can reassure a first-time buyer.

A short buying guide can explain which type of product is suitable for which situation.

None of these exists purely for Google.

They exist because people are trying to make decisions.

The fact that they can also make a page more useful to search engines is a nice side effect.

The technical work still matters

All the useful content in the world will not compensate for a technically poor page.

Your collection should have a clear H1.

The title should describe the category properly.

The URL should be clean and understandable.

Images should be appropriately compressed.

The page should work properly on mobile.

The layout should not become painfully slow because someone uploaded enormous images for every product.

These are not exciting tasks.

They are also not optional.

A collection page that is difficult to crawl, slow to load or confusing to navigate is not going to become a great SEO asset simply because you added another 400 words underneath the products.

Authority has to reach the collections too

Another common ecommerce SEO mistake is building links almost exclusively to the homepage or random blog posts.

The homepage matters.

But if your goal is to make important collection pages rank, those pages need authority as well.

Relevant supplier mentions can help.

Industry publications can help.

Niche roundups and comparison articles can help.

Useful community participation can sometimes create legitimate references.

Relevant directories can make sense in the right industry.

The important word is relevant.

I would rather have a small number of genuinely useful links pointing toward important commercial pages than a pile of meaningless links created simply to make a monthly SEO report look busy.

Authority is not valuable because a spreadsheet says you built ten links this month.

It is valuable when it helps the pages that matter become more competitive.

Your blog should support your collections

This is where ecommerce content marketing becomes much more useful.

Suppose you sell hiking boots.

You publish an article called “How to Choose the Right Hiking Boot.”

The article can explain the different types of boots, terrain considerations, waterproofing, weight and fit.

Then it can naturally point readers toward your Waterproof Hiking Boots collection, Lightweight Hiking Boots collection and Winter Hiking Boots collection.

Now the blog is doing something useful for the commercial pages.

The same principle works across almost any category.

The blog answers broader questions.

The collection captures the buying intent.

The product page closes the transaction.

That is a much better relationship between content and commerce than publishing 50 articles that never link to anything you actually sell.

Collections should not be finished forever

One of the biggest misunderstandings about SEO is that you optimize a page once and move on.

Search data gives you new information all the time.

Google Search Console might show that a collection is appearing for queries you never deliberately targeted.

Customers might repeatedly ask questions you have not answered.

A new product might create an obvious internal-linking opportunity.

A collection might be ranking but attracting the wrong kind of visitor.

The title might need refining.

The copy might be outdated.

The products might have changed.

That gives you reasons to revisit the page.

Every 30–60 days is a reasonable review cycle for important collections, although the right frequency depends on the store and how quickly things change.

Look at the queries.

Look at the clicks.

Look at the rankings.

Look at the products.

Look at the internal links.

Then make improvements based on what you actually see.

SEO is much more useful when treated as an ongoing process than as a launch-day checklist.

The collection page is where several parts of ecommerce meet

This is why I think Shopify stores often have less of a traffic problem than they think.

They have an architecture problem.

The store has products, but the categories are weak.

It has content, but the content does not support the commercial pages.

It has backlinks, but they are not reaching the pages that need authority.

It has traffic, but the collection pages do little to help visitors decide what to buy.

Fixing that does not necessarily require more traffic.

Sometimes it requires making better use of the traffic you already have.

A well-built collection can rank for valuable searches, introduce a customer to the right group of products, link them deeper into the store, answer their questions and move them toward a purchase.

That is a lot of work for a page that many Shopify stores treat as nothing more than a product grid.

The best ecommerce SEO is not about creating more pages for the sake of having more pages.

It is about giving every important page a job.

Your product pages should sell the product.

Your collection pages should organize and capture buying intent.

Your content should answer questions and support those commercial pages.

Your internal links should connect the whole thing.

And your authority should reach the pages that actually matter.

Traffic is useful.

But traffic arriving at a poorly structured store is like pouring water into a bucket with a hole in it.

Before you spend another month trying to get more visitors, look at the pages those visitors are supposed to land on.

You may find that the opportunity was already there.

What You Actually Need to Learn to Become a Shopify Developer

Katalaga Hakim — Tue, 03 Feb 2026 00:00:00 GMT

There is a particular kind of job post that makes Shopify development sound much harder than it is.

"Full-stack Shopify developer."

Then comes a list of Liquid, JavaScript, React, Node.js, APIs, webhooks, databases, app development and about fifteen other things that make someone who is just getting started wonder whether they have accidentally applied for a job at NASA.

The reality of agency work is usually much less dramatic.

If your goal is to learn Shopify development and eventually get hired by an agency, you don't need to start by trying to become a software engineer who happens to know Shopify. You need to become useful inside a Shopify store.

That distinction changes what you should learn first.

Start with Liquid and theme development

A large part of the day-to-day work in Shopify agencies is not building some enormous application from scratch.

It is taking a store that already exists and making it better.

A client wants a different homepage layout. A product page needs another section. The collection page needs better filtering. The cart needs different logic. A trust section needs to be added. A design from Figma needs to become an actual Shopify page.

That is where Liquid, HTML, CSS and JavaScript become useful.

Liquid is particularly important because it is the language that connects the Shopify store's data with the theme.

You need to be comfortable editing templates, creating custom sections, working with blocks and schema, and controlling what merchants can change through the Shopify theme editor.

You should be able to look at a design and think about how you would actually build it inside Shopify rather than simply knowing what the design is supposed to look like.

That is a much more valuable skill than memorizing another framework.

The best way to learn is to build stores

This is where a lot of people get stuck.

They spend weeks watching Shopify tutorials.

Then another course.

Then another video about Liquid.

Then they discover a video called "Complete Shopify Development Roadmap 2026" and start that one too.

Eventually they know a lot about Shopify development and have built absolutely nothing.

Build instead.

Take a fictional brand and create the store yourself.

Give it a real niche. Give it a real product catalog. Make decisions about the homepage. Build custom sections. Create a product page that actually makes sense. Think about mobile navigation, collections, the cart and the buying journey.

Then build another one.

You don't need ten unfinished tutorial projects sitting in a folder. Two or three well-built demo stores can teach you far more if you actually make decisions instead of following someone else's clicks.

And those stores become something you can show an agency.

That matters because agencies aren't particularly interested in hearing that you completed a twelve-hour course.

They want to know whether you can take a design, understand what the merchant needs and turn it into a clean working store.

Your portfolio should make that obvious.

A good Shopify developer understands more than code

This is another part that gets overlooked.

A Shopify store is not just a collection of templates.

Someone has to decide where the information goes.

Should the homepage lead with the product or the brand story?

Should the product page show the reviews above or below the description?

Does the collection page make it easy to compare products?

Is the add-to-cart button obvious on mobile?

Does the customer understand shipping before reaching checkout?

Are there unnecessary sections slowing the page down?

These are development decisions, but they are also ecommerce decisions.

A developer who can build a beautiful section that makes the buying experience worse isn't particularly useful.

The better developer understands why the section exists in the first place.

That is why your demo stores should not just look good in a screenshot. They should demonstrate judgment.

Show custom sections. Show thoughtful layouts. Show improvements to the customer journey. Show that you cared about speed. Show that you understand how the store is supposed to work.

The code matters.

So does the decision behind the code.

Then learn the backend side

Once you're comfortable working with themes, the backend side of Shopify becomes much easier to put into perspective.

You don't need to imagine yourself building the next Facebook.

A lot of practical Shopify backend work is about connecting Shopify to other systems.

An order needs to be sent to a CRM.

A webhook needs to trigger a workflow.

Customer or product data needs to be retrieved through the Admin API.

A subscription system needs to communicate with Shopify.

A shipping platform needs order information.

A CRM, automation platform or another business system needs to stay synchronized with the store.

That is where Shopify APIs, webhooks and something like Node.js become useful.

You're often wiring systems together.

The interesting part is that these tasks are much easier to understand once you've spent time inside an actual store. You start seeing the connections because you understand what the business is trying to accomplish.

What "full-stack Shopify developer" actually means

The phrase sounds intimidating because "full-stack" has become one of those job titles that can mean almost anything.

In Shopify, though, it can be much more practical.

On the frontend, you might be working with Liquid, JavaScript, HTML and CSS, with some React depending on what you're building.

On the backend, you might work with Shopify APIs, webhooks and Node.js while connecting external services.

You don't necessarily need to spend every morning writing React applications.

If you're strong with Liquid and JavaScript, you can already do a substantial amount of useful Shopify work.

Then you add the backend knowledge as the problems you encounter require it.

That is a much more sensible learning path than trying to master everything simultaneously.

What agencies actually care about

There is a difference between knowing a technology and being useful with it.

Someone can list Liquid, JavaScript, React, Node.js and Shopify APIs on their resume and still struggle to build a decent product page.

Someone else can have a smaller technical stack but take a messy Shopify store, understand what is wrong with it and fix the problem properly.

The second person is usually much more interesting to an agency.

Because agency work is full of problems.

A client doesn't normally arrive saying, "Please demonstrate your knowledge of Liquid."

They say, "The homepage doesn't look right on mobile."

Or, "We need this section added."

Or, "The product page isn't converting."

Or, "The cart needs to work differently."

Or, "Can you make this design look like the Figma?"

Your job is to understand the problem, work out the cleanest solution and ship it without creating three new problems somewhere else.

That is development.

Don't build your career around certificates

Certificates can show that you completed something.

They don't show that you can build.

A working store does.

If I were reviewing two junior Shopify developers and one showed me five certificates while the other showed me three well-built stores with custom sections, sensible UX decisions, good performance and an explanation of what they changed, I'd be far more interested in the second portfolio.

The stores give me something to inspect.

I can see the work.

I can ask why a particular decision was made.

I can see whether the developer understands the difference between making something technically work and making it work well.

That's the standard you should aim for.

Don't just build something and put a link in your portfolio.

Explain what you changed and why.

Maybe the original product page buried important information, so you reorganized the layout.

Maybe the theme had too many unnecessary sections, so you simplified it.

Maybe the mobile experience was awkward, so you redesigned the navigation.

Maybe the store was visually inconsistent, so you created reusable sections and standardized the design.

Now you're not just showing a website.

You're showing how you think.

And that is much harder to fake.

The roadmap is simpler than it looks

If I were starting from zero with the specific goal of becoming useful to a Shopify agency, I wouldn't try to learn everything at once.

I'd start with Shopify fundamentals and theme architecture.

Then I'd spend serious time with Liquid.

Then HTML and CSS.

Then JavaScript.

Then I'd build several stores from scratch instead of following tutorials forever.

After that, I'd get comfortable with Shopify's APIs and webhooks and learn enough Node.js to build practical integrations.

React can come into the picture when the type of Shopify work you're doing actually requires it.

And throughout the process, I'd keep asking the same question:

Can I make this store better?

Because that is ultimately what an agency is paying you to do.

Not to recite a technology stack.

Not to collect certificates.

Not to call yourself full-stack.

To take a real ecommerce problem and solve it cleanly.

If you can take a Figma design and turn it into a working Shopify page, improve a product page without destroying the buying experience, build reusable theme sections, understand how Shopify data moves through APIs and connect the store to the other systems a business relies on, you're already much closer to being employable than someone who has simply watched another hundred hours of tutorials.

The stack will keep changing.

The ability to build, diagnose and improve a store is the part that stays useful.

Are Liquid Themes Still Relevant? Or Is Headless Shopify the Move?

Katalaga Hakim — Sun, 18 Jan 2026 00:00:00 GMT

A few years ago, telling someone you were building a Shopify store headlessly had a certain effect.

It sounded serious.

The frontend was custom. The architecture was more complicated. There was usually a modern JavaScript framework somewhere in the stack, a separate hosting setup, APIs connecting different systems and enough technical vocabulary to make a normal Shopify theme sound like something from another era.

For some brands, especially larger Shopify Plus businesses, going headless made genuine sense. There were situations where the flexibility of a custom frontend justified the additional engineering work.

But something has changed.

As ecommerce teams have matured, the conversation has become less about what architecture looks impressive and more about what architecture allows the business to do efficiently.

That changes the question.

The question is no longer whether headless Shopify is more technically powerful than a Liquid theme. It usually is. The more useful question is whether that additional power solves a problem the business actually has.

For many Shopify stores, the answer is no.

Headless Became a Symbol of Progress

There was a period when headless felt like the obvious next step for ambitious ecommerce brands.

A custom frontend meant almost complete control over the customer experience. Developers could build highly interactive interfaces, unusual animations and experiences that were difficult to reproduce inside a traditional theme architecture.

It also carried a certain perception.

Headless sounded like enterprise ecommerce. It suggested that the brand had moved beyond the limitations of a normal storefront and was now operating on a more sophisticated technical foundation.

There is nothing inherently wrong with that.

The problem is when the architecture becomes the achievement instead of the business outcome.

A more complicated system is not automatically a better system. It is simply a system with more moving parts.

And once a store is operating at scale, those moving parts have a cost.

The Real Cost of Going Headless

A headless Shopify store is not simply a Shopify theme with a different frontend.

You are introducing another application into the business.

Instead of Shopify providing the storefront presentation through Liquid, you have a separate frontend application, potentially built with technologies such as Next.js or Remix. That frontend needs hosting, deployment processes, monitoring, maintenance and coordination with Shopify's APIs and the rest of the technology stack.

The complexity does not stop once the initial build is finished.

A developer has to maintain the frontend. API integrations need to remain functional. Deployments need to be managed. Changes need to be tested. Frontend and backend systems need to remain compatible. The team needs to understand what happens when something changes on one side of the architecture.

You have moved from maintaining a Shopify storefront to maintaining a software product built around Shopify.

For the right business, that can be completely justified.

For many businesses, it is simply more infrastructure than they need.

Most Stores Are Not Limited by Liquid

Think about what the majority of ecommerce stores are actually trying to accomplish.

They need product pages that explain the product properly. They need collection pages that make the catalogue easy to navigate. They need fast load times, effective conversion elements, clean integrations, useful landing pages and a checkout experience that does not get in the customer's way.

None of those requirements automatically demands a headless frontend.

Modern Shopify themes are considerably more capable than the old image of a basic Liquid template suggests. With Liquid, sections, metafields and metaobjects, developers can build highly customised storefronts while keeping the underlying Shopify theme architecture.

You can create different templates for different products. You can build dynamic landing pages. You can structure content around specific product attributes. You can create reusable sections and components that allow merchants to change parts of the storefront without asking a developer to rebuild the page.

There is a lot that can be done before a business reaches the point where rebuilding the frontend becomes necessary.

The mistake is assuming that the existence of a technical limitation somewhere means the entire storefront architecture needs to be replaced.

The Underrated Advantage Is Merchant Control

There is another benefit of staying with a well-built Liquid theme that does not receive nearly as much attention as technical flexibility.

The merchant can actually use it.

A Shopify store is constantly changing. Campaigns are launched. Offers change. Products come and go. Collection structures evolve. Landing pages need to be created. Seasonal promotions appear. A merchandising team might want to rearrange a page because customers are responding differently than expected.

With a properly structured Liquid theme, much of that work can happen inside Shopify's existing content and theme management tools.

A marketer can adjust a section. A merchant can update content. A new template can be created for a campaign. A different product layout can be applied without rebuilding the entire frontend.

That matters because ecommerce is iterative.

The business does not finish the website and move on. The website is part of the business's daily operation.

If every relatively small change requires a developer, a deployment and a QA process, the technical flexibility of the architecture can start creating operational friction.

And friction is expensive.

Headless Does Not Automatically Mean Faster

Performance is often one of the strongest arguments for headless.

And there is truth behind it. A well-engineered headless storefront can be extremely fast and gives developers a great deal of control over how the frontend is rendered, cached and delivered.

But the comparison is not really between "headless" and "Liquid."

It is between a well-built implementation and a poorly built one.

A Liquid theme loaded with unnecessary apps, oversized images, inefficient code and third-party scripts can be slow. A headless storefront can also be poorly maintained, overloaded with scripts or built in a way that introduces its own performance problems.

In many stores, the biggest performance problems are not caused by Liquid itself. They come from what has been added around the theme.

Too many apps. Unoptimised media. Heavy tracking scripts. Poorly written custom code. Unnecessary third-party resources.

A clean, carefully developed Liquid theme can perform extremely well.

The fact that another architecture gives developers more theoretical control does not mean changing architectures will automatically solve the actual performance problem.

Sometimes the problem is simply the implementation.

So When Does Headless Actually Make Sense?

There are businesses where headless is absolutely the right choice.

If the storefront needs to behave more like an application than a conventional ecommerce website, the additional control can be valuable. A business might have a highly interactive customer experience, complex data coming from multiple systems, sophisticated personalisation requirements or frontend logic that is difficult to implement efficiently within a standard Shopify theme.

There can also be organisational reasons for choosing it. A large engineering team may already have the infrastructure and expertise required to maintain a custom frontend, making the operational cost much lower than it would be for a smaller team.

In those situations, headless is not a vanity project. It is solving a real architectural problem.

The issue is when a store goes headless because the team wants a more impressive technology stack.

Nicer animations are not an architectural requirement.

Neither is wanting the store to look more "enterprise."

The architecture should exist because the business needs what it provides.

The Team Matters More Than the Architecture Diagram

This is probably one of the most overlooked parts of the decision.

Headless development requires a different level of technical ownership.

You need people who understand frontend engineering. You may need backend and API expertise. Depending on the setup, you may also need someone comfortable with hosting, deployments, monitoring and the rest of the infrastructure surrounding the application.

That does not mean every headless store needs a dedicated DevOps department.

It does mean somebody needs to own those responsibilities.

If the business does not have that capability internally or through reliable partners, the architecture can become technical debt very quickly.

A Liquid theme keeps more of that responsibility inside the Shopify ecosystem.

For a smaller ecommerce team, that can be a significant advantage.

The simpler system is often easier to understand, easier to maintain and easier to hand over when somebody on the team changes.

Ecommerce Teams Need Speed More Than They Need Prestige

This is where the conversation eventually comes back to business.

An ecommerce team is constantly testing things. They change offers. They adjust product pages. They create campaigns. They test layouts. They add merchandising content. They respond to seasonal demand and customer behaviour.

The architecture needs to support that pace.

If a marketer sees an opportunity on Monday and needs a developer, a deployment and several rounds of QA before the change can go live, the business has introduced friction into its own growth process.

Sometimes that friction is worth accepting because the custom experience produces enough additional value.

Often it is not.

For many Shopify stores, the ability to make changes quickly inside the platform is more valuable than having complete frontend control that rarely gets used.

That is one reason Liquid remains relevant.

It is not because Shopify themes are somehow more sophisticated than a custom application. They are not.

It is because most businesses do not need a custom application to run a good ecommerce store.

Start With the Business Problem

Before choosing headless, I would ask a much simpler question:

What exactly are we unable to do with the current Shopify architecture?

If the answer is that the store needs a highly specialised experience that genuinely cannot be implemented effectively in a theme, that is worth investigating.

If the answer is that the team wants more flexibility, I would first find out what kind of flexibility is actually missing.

If the answer is better performance, I would audit the current implementation before assuming the architecture is responsible.

If the answer is more impressive animations, I would probably save the engineering budget.

There is a tendency in technology to treat architectural decisions as permanent markers of sophistication. In ecommerce, they are much more practical than that.

The best architecture is the one that allows the business to operate effectively without introducing unnecessary complexity.

For most stores, that means getting the existing Shopify foundation right before replacing it.

Liquid Is Still Relevant

The idea that Liquid is an outdated technology that ambitious Shopify brands eventually need to escape does not hold up very well.

A well-built Liquid theme can support sophisticated product pages, dynamic content, structured merchandising, custom landing pages, strong performance and a highly tailored customer experience.

More importantly, it keeps the storefront closely connected to the tools the ecommerce team is already using.

That matters.

The goal of an ecommerce website is not to demonstrate how advanced the development team is. It is to help the business sell products efficiently while giving the people running the store enough control to keep improving it.

Headless is a powerful option when a business genuinely needs it. But it should be treated as an architectural solution to a specific problem, not as the default definition of a serious Shopify store.

For most brands, the better investment may be much less glamorous: clean theme architecture, fast pages, better product templates, stronger collection structures, better merchandising, useful content and a storefront the marketing team can actually control.

Architecture should support the business.

It should not become the business.

Your Shopify Product Page Is Where the Sale Is Won or Lost

Katalaga Hakim — Tue, 30 Dec 2025 00:00:00 GMT

A Shopify store can have a beautiful homepage, a good-looking logo, a fast theme and perfectly organized collections, and still struggle to make sales.

Sometimes the problem is sitting one click away.

The product page.

This is one of the strange things about ecommerce. A business can spend days getting the homepage to look right because that is the page everyone sees when the store is being built. Then the product pages get filled in almost as an afterthought.

A product title is copied from the supplier. The description is pasted underneath it. Four product images are uploaded. The price is added. Maybe a review app is installed. Then someone wonders why the traffic coming from Google Ads or Meta isn't turning into enough orders.

The customer, meanwhile, has arrived on the page asking a very simple question:

"Why should I buy this?"

Your product page has to answer that question without making the customer work for it.

Start with the product, not the template

A common mistake when building Shopify stores is designing the product page before understanding what is actually being sold.

The page should change depending on the product.

A $20 phone accessory does not need the same presentation as a $300 piece of professional kitchen equipment. A skincare product has different questions around ingredients and usage. Apparel needs sizing information. A technical product may need specifications, diagrams and demonstrations.

The template is the container. The product is the thing that determines what belongs inside it.

That sounds obvious, but a surprising number of stores do the opposite. They take the default Shopify product template and try to squeeze every product into it.

The result is a page that technically works but does very little selling.

Before touching the design, figure out who the buyer is, what they care about, what objections they are likely to have and what information they need before they are comfortable purchasing.

That becomes the foundation of the page.

The title should tell people what they are looking at

Product titles are often treated as an SEO exercise, and that is where things go wrong.

A product title should identify the product clearly. It should not look like something copied from a supplier catalog.

There is a big difference between:

"2024 NEW HOT TRENDY WOMEN PUFFER JACKET WINTER BEST SELLER"

and:

"Marshmallow Crop Puffer Jacket"

The second one tells the customer what the product is. The rest of the information can be handled elsewhere.

The title also needs to make sense when someone encounters the product outside your store. It might appear in Google Shopping, a collection page, an email, a search result or a social post.

Clarity matters more than stuffing every possible keyword into the title.

The first screen has a job to do

When someone lands on a product page, you should not make them scroll through half the page before they understand what they are buying.

The important information needs to be visible immediately.

Product name.

Product rating, if genuine reviews exist.

Price.

A clear indication of the offer, if there is one.

A concise explanation of the product's main benefit.

The purchase option.

Basic shipping or returns information.

And, of course, strong product imagery.

This is not about cramming everything above the fold. It is about removing unnecessary detective work.

If someone has to scroll around looking for the price, figure out whether shipping is available to them, or work out what the product actually does, the page is making the sale harder than it needs to be.

The primary action should also be obvious. Add to Cart should not disappear underneath a paragraph of text or get lost among five competing buttons.

On mobile, this becomes even more important. A large percentage of ecommerce traffic arrives on a phone, where there is considerably less space to work with.

Product photography does more work than most descriptions

There is a point where writing more copy stops helping.

Better images can solve the problem instead.

The customer cannot pick up the product. They cannot turn it around, feel the material, check the size with their hands or see how it looks in their own environment.

Photography has to compensate for some of that.

A useful product image set normally includes a clear primary image, additional angles, close-up details and lifestyle imagery where it makes sense.

For apparel, that might mean seeing the product being worn. For furniture, it might mean seeing it in a real room. For professional equipment, it might mean showing the product being used.

A product photograph should answer questions before the customer has to ask them.

What does it actually look like?

How big is it?

What does the material look like?

How does it work?

What does it look like in use?

If the product has something customers regularly misunderstand, show it.

A size chart can be more useful than another decorative photograph. A short demonstration video can be more useful than another paragraph describing the same feature.

The goal is not to hit an arbitrary number of images. The goal is to give the buyer enough visual information to make a confident decision.

The description should sell the product without sounding like an advertisement

This is where many Shopify stores become strangely difficult to read.

The product description says something like:

"Experience ultimate quality and premium performance with our revolutionary, high-quality product designed to transform your lifestyle."

It sounds impressive until you realize it hasn't actually told the customer anything.

Good product copy is usually much more concrete.

Start with the reason the product exists.

What problem does it solve?

What does it make easier?

Who is it useful for?

Then explain the important benefits and features.

After that, give the technical information someone may need before purchasing.

The distinction matters.

A feature tells me what the product has.

A benefit tells me why I should care.

For example, "AISI 304 food-grade stainless steel" is a feature.

"Built for repeated use in professional kitchens and easy to clean after service" explains why that feature matters.

The technical specification still belongs on the page. But it should not be the entire sales argument.

Give people the information they need to remove their objections

Most customers don't arrive on a product page with only one question.

They may be wondering whether the product is the right size, whether it will arrive in their country, whether they can return it, whether it is durable, whether it works with something they already own, or whether the product is actually as good as the photographs suggest.

The best product pages anticipate these questions.

That is why a useful product page often needs sections for specifications, dimensions, materials, usage instructions, compatibility, shipping, returns and frequently asked questions.

Think of the page as a conversation.

The customer asks a question.

The page answers it.

The customer thinks of another objection.

The page answers that too.

Eventually there are fewer reasons left not to buy.

That is a much better way to think about product-page content than simply trying to make the description "long enough."

Trust has to appear where the decision is being made

Trust cannot live exclusively on the About page.

A customer deciding whether to buy a product needs reassurance on the product page itself.

That might come from genuine customer reviews, verified ratings, customer photographs, testimonials, clear shipping information, a straightforward return policy or a visible guarantee.

The important word is genuine.

Adding fake reviews or invented testimonials may make a page look more complete, but it creates a much bigger problem when customers realize the proof isn't real.

The strongest stores don't need to hide behind vague claims.

They show who bought the product, what those customers thought, how the product is used and what happens if something goes wrong.

Even simple things matter.

A real business email.

Clear shipping times.

A visible returns policy.

A physical business location where appropriate.

A company story that actually explains who is behind the brand.

These details don't look exciting during development. They become important when someone is deciding whether to hand over their money.

Then you can start thinking about conversion

Once the page answers the basic questions and establishes trust, conversion elements can help.

A free-shipping threshold can encourage a larger basket.

Bundles can make the offer more attractive.

Frequently bought together products can help customers complete their purchase.

Quantity discounts can make sense for products people buy in multiples.

A sticky add-to-cart button can reduce friction on mobile.

Email capture can give the business another opportunity to reach visitors who aren't ready to buy immediately.

These things can be useful, but they are not substitutes for a good product page.

A countdown timer will not rescue a product nobody understands.

A popup will not fix bad photography.

A "10% OFF" banner will not create trust where there is none.

Conversion elements work best when they sit on top of a solid buying experience.

Be careful with urgency

There is also a temptation to put urgency everywhere.

"ONLY 2 LEFT!"

"SALE ENDS IN 03:42!"

"BUY NOW!"

Sometimes genuine scarcity is useful. If inventory really is low, telling the customer makes sense.

But manufactured urgency can make a store feel like a dropshipping template rather than a serious business.

The same applies to discounts.

If every product is permanently "50% OFF," customers eventually stop believing the original price means anything.

The page should make the purchase easier, not noisier.

A good product page is part of the whole store

The product page does not operate in isolation.

A visitor might arrive from a Google Shopping ad without ever seeing the homepage.

Another customer might discover the brand on Instagram, click directly into a product and never read the About page.

Someone else might land on a collection page first, compare three products and then make the decision on the product page.

That means the product page has to carry enough of the brand's trust and product information to stand on its own.

The customer journey is not always:

Homepage → Collection → Product → Cart.

It can be:

Google → Product → Cart.

Or:

Instagram → Product → Exit.

Or:

Google Shopping → Product → Back to Google → Competitor.

The product page is often the moment where all the work done in advertising, SEO, social media and merchandising either turns into a sale or disappears.

That is why I treat product-page QA as one of the most important parts of a Shopify build.

The final test is surprisingly simple

Before launching a product page, forget for a moment that you built it.

Open it as a stranger.

Can you tell what the product is immediately?

Can you understand who it is for?

Can you see what it costs?

Can you understand the main reason to buy it?

Can you see the product properly?

Can you find the important specifications?

Can you understand shipping and returns?

Is there enough real proof to make the business believable?

Can you add the product to your cart without thinking about how the website works?

And perhaps the most useful question:

If I arrived here from an ad with no previous knowledge of this brand, would I have enough information to feel comfortable buying?

If the answer is no, don't start adding more apps.

Fix the page.

Because a Shopify store doesn't become conversion-ready when every section has been filled in. It becomes conversion-ready when the customer can move from "What is this?" to "I understand it" to "I trust it" to "I'll buy it" without having to fight the website along the way.

Your Google Ads Performance Has a Trust Problem

Katalaga Hakim — Tue, 02 Dec 2025 00:00:00 GMT

There is a point in every ecommerce journey where the advertising stops.

It happens immediately after the click.

Someone searches for a product. They see your ad. The targeting is good, the keyword is relevant and the product looks like exactly what they wanted.

They click.

Now they are on your product page.

This is where a different kind of marketing begins.

They look at the product photos. They check the price. They look for shipping information. Then, almost without thinking about it, they look at the reviews.

How many people have bought this?

What did they say?

Are the reviews recent?

Do the customers look like real people?

Does anyone actually seem happy with this product?

You can have an excellent Google Ads account and still lose the sale at this point.

The ad got the customer to your store.

The product page has to give them a reason to trust it.

The Ad Is Not the Whole Customer Journey

It is easy to look at Google Ads as though performance begins and ends inside the advertising account.

You can have strong targeting.

Clean keyword structures.

Logical campaign segmentation.

Good bidding.

Well-written ads.

And still have mediocre results.

Why?

Because Google does not complete the transaction for you.

The customer does.

And customers do not behave like spreadsheets.

They have doubts.

They compare.

They hesitate.

They look for reassurance before spending money with a company they may have never heard of before.

Reviews are one of the easiest ways for them to find that reassurance.

A product page with three reviews tells a very different story from one with hundreds of genuine customer reviews.

A review from last week feels different from one posted eight months ago.

A written review is useful.

A customer photograph can be even more convincing.

A brand that responds thoughtfully to questions and feedback sends another signal that there are actual people behind the store.

None of these things guarantee a sale.

But their absence can make a customer hesitate.

The Quiet Conversion Problem

Imagine two stores selling similar products.

The first has a good advertisement and a clean product page.

There are three reviews.

The newest one is months old.

There are no customer photos.

There is little information about returns.

The second store has hundreds of reviews, recent customer feedback, photographs from buyers and clear information about shipping and returns.

The customer may never consciously sit down and compare these things.

They simply feel different.

One store feels established.

The other feels uncertain.

That difference matters because the click has already cost money.

If a customer arrives through Google Ads and leaves because they do not trust the store, the advertising campaign still paid for that visit.

The problem was not necessarily the targeting.

The problem happened after the targeting did its job.

Trust Changes the Economics of the Click

This is where reviews become more than a nice feature on a product page.

Suppose you pay to bring a customer to your store.

If ten visitors produce one sale, the economics look very different from a situation where ten visitors produce two.

The advertising did not necessarily become cheaper.

The page became better at converting the traffic.

That means the same click can become more valuable.

Your acquisition cost can improve.

Revenue per visitor can increase.

And the conversion data being sent back to Google becomes stronger.

This matters when you are using automated bidding.

Smart Bidding learns from conversion signals. If the site consistently turns qualified visitors into customers, Google has better evidence about which auctions and users are valuable.

If the site receives good traffic but repeatedly fails to convert it, the account has a different problem to solve.

The important distinction is that Google Ads cannot manufacture trust that the website has not earned.

Look at the Product Page Through a Customer's Eyes

When auditing an ecommerce store, I would not stop at the ad.

Follow the entire journey.

Search.

Ad.

Product page.

Checkout.

Then ask what the customer is likely thinking at each stage.

On the product page, reviews are only one part of the picture.

Is the star rating visible?

Is the number of reviews believable?

Are they recent?

Are there customer photographs?

Can shoppers see questions and answers?

Is the return policy easy to find?

Is shipping clearly explained?

Does the product page explain what makes the product worth buying?

Are the claims supported by actual information?

Does the page look like a real business selling a real product?

These details work together.

Trust is rarely created by one badge or one five-star review.

It is created by the accumulation of small signals.

Review Volume Matters, But Recency Matters Too

There is a difference between having reviews and looking like a store that customers are still buying from.

A product with hundreds of reviews collected over many years may have strong social proof.

But if the latest review is ancient, a customer can still wonder whether the product is actively sold or whether those reviews are relevant to the current version.

Recent feedback helps create a sense that people are still buying the product.

That does not mean asking customers for reviews every five minutes.

It means having a system for generating genuine feedback after purchases and making sure useful reviews actually appear where shoppers can see them.

The important word is genuine.

The goal is not to manufacture five-star ratings.

It is to make the real customer experience visible.

Reviews Are Only One Layer of Trust

The same principle applies beyond product reviews.

Store-level ratings can help.

So can reputable third-party reviews.

Clear guarantees can help remove uncertainty.

Straightforward shipping information can prevent surprises.

A properly explained return policy can make the purchase feel less risky.

Customer-generated photographs can show what the product looks like outside the carefully controlled studio environment.

Even something as simple as a brand responding to customer questions can make a store feel more legitimate.

None of these elements needs to scream for attention.

In fact, the best trust signals often do the opposite.

They quietly answer the questions a cautious customer would otherwise have to ask.

What happens if this doesn't fit?

How long will it take to arrive?

What do other customers think?

Can I return it?

Is this business still active?

Will someone help me if there is a problem?

Every unanswered question creates a little friction.

Enough unanswered questions and the customer leaves.

The Competitor With More Trust Has an Advantage

This becomes particularly obvious when you compare two stores competing for the same search.

Both may have similar products.

Both may be running Google Ads.

Both may have reasonable prices.

Both may be targeting the same commercial keywords.

But one product page has strong, recent social proof and the other barely has any.

Now the competition is not simply happening inside Google Ads.

It is happening on the product page.

The customer clicked both stores.

Now they are deciding which one feels safer.

That is why I would be careful about blaming every weak ROAS number on the advertising account.

Sometimes the campaign is bringing the right people.

The website is simply giving them too many reasons to hesitate.

This Is Where Ads and Ecommerce Operations Meet

Reviews also reveal something broader about ecommerce management.

Someone has to make sure the review system works.

Someone has to make sure post-purchase emails are encouraging customers to leave genuine feedback.

Someone has to surface useful customer-generated content.

Someone has to notice when an important product has almost no reviews.

Someone has to identify product pages where trust is clearly weaker than it is on competing products.

That work can easily become part of a strong Shopify or ecommerce VA's responsibilities.

Not simply uploading products.

Not simply checking orders.

But watching the relationship between the store, the customer and the advertising.

If one product has excellent traffic but consistently poor conversion, investigate it.

Maybe the price is wrong.

Maybe the offer is weak.

Maybe the page is confusing.

Or perhaps the product has three reviews while the competitor has 700.

That is useful information.

Don't Buy More Traffic Before Fixing the Leak

This is one of the easiest ecommerce mistakes to make.

ROAS looks disappointing.

The immediate reaction is to increase traffic.

Increase the budget.

Launch another campaign.

Add more keywords.

Try another audience.

But if the existing visitors are already arriving at the right pages and failing to trust what they see, more traffic simply creates more opportunities to lose people.

Before spending more, look at the product pages.

Look at the reviews.

Look at the customer photos.

Look at shipping and returns.

Look at the offer.

Look at how clearly the page explains the product.

Look at the difference between your page and the pages competing for the same customer.

Sometimes the advertising account needs fixing.

Sometimes the bottleneck is sitting three inches below the product image.

Trust Is Part of Conversion Rate Optimization

This is why reviews belong in the same conversation as CRO.

You are not adding reviews because they look nice.

You are reducing uncertainty.

A customer who understands the product, sees credible evidence from other buyers, knows what delivery looks like and understands what happens if they need to return it has fewer unanswered questions.

Fewer unanswered questions can mean less hesitation.

Less hesitation can mean more purchases.

And more purchases from the same amount of qualified traffic changes the economics of your entire acquisition system.

That is the part people miss.

Google Ads does not operate in isolation.

The ad creates the opportunity.

The product page handles the opportunity.

The checkout completes it.

The review system helps make the next customer more confident.

Each part affects the next.

Audit the Page Before Blaming the Campaign

If your Google Ads performance feels stuck, I would not immediately start changing everything inside the account.

Look at the customer journey.

Are the reviews strong enough?

Are they recent?

Are customer photographs visible?

Is the review count credible?

Is review markup implemented correctly where appropriate?

Are shipping and returns obvious?

Does the landing page match the search and the ad?

Does the product page answer the customer's obvious objections?

And, perhaps most importantly, how does your page compare with the competitors receiving the same traffic?

The answer may have very little to do with your bids.

You can optimize the campaign endlessly, but eventually the customer has to make a decision.

They are not thinking about your Quality Score.

They are not thinking about your campaign structure.

They are wondering whether they should trust you with their money.

That is a different problem.

Ads bring the customer to the door.

The product page is where you convince them to walk inside.

And reviews are one of the clearest signs that other people have already done it.

For Clothing Brands, Email Is More Than a Sales Channel

Katalaga Hakim — Tue, 11 Nov 2025 00:00:00 GMT

A clothing store can have plenty of traffic and still have a retention problem.

Someone sees an Instagram post, clicks an ad, finds a hoodie they like, buys it, and disappears.

Then the store goes looking for another customer.

And another.

The cycle gets expensive quickly.

This is one reason email is particularly valuable for fashion and apparel brands. Clothing is not a category where people necessarily make one purchase and disappear forever. People change their wardrobes. Seasons change. Trends change. Sizes change. New collections arrive. A customer who bought a blazer six months ago may now be the perfect customer for a new pair of trousers.

Fashion also has something that many other ecommerce categories do not have to the same degree.

It is emotional.

People buy clothes because they need them, but they also buy them because of how they want to look, how they want to feel, and how they want other people to see them.

That gives an apparel brand many more reasons to stay in contact with a customer than simply sending another 10% discount.

The problem is that many stores barely use any of them.

The foundation comes before the campaigns

Before worrying about clever subject lines or elaborate seasonal campaigns, the basic email infrastructure needs to work.

A clothing store should have the obvious flows in place: welcome, abandoned cart, abandoned checkout, post-purchase and win-back.

These are not exciting campaigns.

They are the plumbing.

If someone joins the email list and receives nothing useful, the brand has wasted the moment when interest is highest. If someone adds a product to their cart and disappears, there should be a sensible follow-up. If someone buys a jacket today, the relationship should not end at the order confirmation.

And if a customer has not purchased in a long time, there should be a reason to remind them that the brand still exists.

This matters even more when the business is spending heavily on acquisition.

You have already paid to get the customer.

Email is where you get another chance to earn from the relationship.

A fashion welcome series should introduce a world

A clothing brand has an advantage here because it can sell much more than the individual products sitting in Shopify.

Imagine a streetwear brand.

The first email does not necessarily need to say, “Here are our products. Buy them.”

It can introduce the brand, its aesthetic and the reason it exists.

The next email might show how customers style one of the best-selling cargo pants. Another could feature customer-generated content. Another could introduce a first-order incentive.

The sequence starts to feel less like a discount funnel and more like an introduction.

That matters because people who buy fashion are often buying into an aesthetic.

They are not only deciding whether they like the trousers.

They are deciding whether they like the world surrounding the trousers.

A good welcome flow gives them a chance to see it.

Restocks are different from ordinary promotions

There is something interesting about a restock email.

A normal promotional email has to create the desire to buy.

A restock email often doesn't.

The customer has already done that part.

They wanted the product. They may have visited the page. They may have chosen a size or colour. Then it disappeared because their preferred option was sold out.

When that size comes back, the email can be incredibly simple.

Your size is back.

That is enough.

The intent already exists. The scarcity was real. The customer has already formed an attachment to the product.

This is why back-in-stock flows can be so valuable for fashion brands. They respond to an existing demand instead of trying to manufacture one.

And they should be automated.

If someone has specifically asked to be notified when a particular size returns, there is little reason to make a staff member remember to send the email three weeks later.

Sometimes you should sell the outfit instead of the product

Another opportunity sits in the way people actually shop for clothes.

Customers do not always think in individual SKUs.

They think in outfits.

A person buying a blazer may also need trousers. Someone buying a hoodie might want matching joggers. A customer looking at a dress may also be interested in the shoes or accessories that complete the look.

That gives the brand an opportunity to build emails around the entire outfit.

“Complete the look” is often more useful than simply sending another product recommendation.

It also has an obvious commercial benefit.

If the customer was already going to spend $70, helping them see the rest of the outfit may turn the order into $120 without requiring another customer acquisition.

That is one of the quiet advantages of good email marketing.

You can increase the value of existing demand.

Fashion gives you a calendar full of reasons to talk to customers

Clothing is seasonal by nature.

New collections arrive. Weather changes. Holidays arrive. Certain colours become popular. Certain products become relevant again.

A brand does not have to invent an excuse to email its customers every week.

The calendar is already giving it material.

A winter campaign can be built around coats, layering and cold-weather accessories. A summer campaign can introduce lighter fabrics and warm-weather pieces. A holiday edit can group products around occasions rather than product categories.

The smarter part is deciding who should receive what.

Someone who bought winter coats last year may be a much better audience for early access to the new winter collection than someone who has only ever bought summer clothing.

That is where segmentation starts becoming useful.

The more the email reflects what the customer has actually shown an interest in, the less it feels like advertising being sent to a database.

It feels like a recommendation.

And recommendations tend to be easier to act on than generic promotions.

Abandoned carts have their own objections

The abandoned cart email is another place where clothing brands can do better than “You forgot something.”

The customer may not have forgotten anything.

They may be unsure about the size.

They may be wondering how the fabric feels.

They may not know whether the item fits the way they expect.

They may be worried about returning it if it doesn't work.

Those are different problems from simply forgetting to complete checkout.

So the email can address them.

A fit guide can answer the sizing question. A clear returns policy can reduce risk. Customer reviews can show how the item looks on actual people. Useful product imagery can answer questions the original product page failed to answer.

The email is not just a reminder.

It is another opportunity to remove friction.

The purchase should be the beginning of the relationship

This is probably one of the most underused parts of apparel email marketing.

A customer has just bought something.

Most stores send an order confirmation, shipping updates and then eventually another promotion.

There is a much better opportunity in between.

If someone bought a denim jacket, why not show them how to style it?

Why not introduce customer photos featuring the same product?

Why not explain how to care for the fabric?

And after enough time has passed, why not recommend products that actually complement what they already purchased?

The customer has given the brand information about what they like.

Use it.

The post-purchase period is when the customer is already thinking about the product. It is a natural time to help them get more value from it and introduce them to the rest of the brand.

That can turn a single transaction into the beginning of a buying cycle.

Drops are where fashion can use exclusivity properly

Fashion brands also have something that many ecommerce businesses spend a lot of time trying to manufacture.

Exclusivity.

A new collection can be announced to existing customers before the public gets access.

Repeat buyers can receive early access. High-value customers can get a private preview. Recent purchasers can be given an opportunity to see the new collection first.

The message does not have to scream about a discount.

Sometimes the value is simply being early.

“VIP early access” works because it gives the customer a reason to care about the relationship itself.

They are not just another email address receiving a promotion.

They get something first.

Browse abandonment is particularly natural for fashion

Clothing is highly visual, which makes browsing behaviour unusually informative.

Someone might visit the same dress three times without buying it.

That tells you something.

Instead of sending a generic reminder, the brand can bring the product back into view alongside useful information: reviews, styling ideas, related pieces or, where it is genuinely true, low-stock information.

The goal is not to manufacture pressure.

It is to continue the conversation the customer already started.

That distinction becomes important because fake scarcity can damage trust very quickly.

If something is low in stock, say so.

If it isn't, don't pretend it is.

Fashion brands have enough genuine urgency without inventing it.

Win-back needs to show what changed

The weakest win-back email is often the most obvious one.

“We miss you.”

Fine.

But the customer probably needs a reason to come back.

What has changed since they last purchased?

New arrivals. New colours. New collections. Different styles. Products related to what they previously bought.

The message should answer a simple question:

Why should I care now when I didn't buy recently?

If the brand has genuinely changed or added something interesting, the email has something to say.

If nothing has changed, sending “we miss you” five times will not magically create demand.

The design matters because the product is visual

There is another reason apparel email deserves more attention than a plain text-heavy approach.

Fashion is visual.

The customer needs to see the clothes.

Good photography, clean layouts, sensible product hierarchy and mobile-first design all matter because a large portion of the shopping experience happens on a phone.

The email should not look like a spreadsheet with product names attached.

But it also should not become an elaborate design exercise where the customer has to hunt for the actual products.

The best fashion emails usually make the product obvious.

Strong image.

Clear context.

Useful reason to click.

Then get out of the way.

The numbers should tell you what is actually working

Email reporting can become unnecessarily complicated.

For a clothing brand, I would rather know whether revenue per recipient is improving, whether customers are buying again, whether bundles are increasing average order value and how long customers typically take to purchase again.

Those numbers tell you more about the health of the retention system than simply celebrating a high open rate.

A campaign can have an impressive open rate and generate very little money.

Another campaign can look less exciting in the dashboard and quietly produce significant revenue.

The point of the email is not to win an engagement trophy.

It is to move the customer relationship forward.

This is where someone needs to own the system

None of this works particularly well if email is treated as something that gets checked once every few weeks.

Someone has to notice that a popular product has been restocked and make sure the trigger works.

Someone has to build the new bundle.

Someone has to replace an outdated seasonal block.

Someone has to refresh the product imagery.

Someone has to review campaign performance and ask why one segment is responding while another isn't.

Someone has to keep testing.

That is another area where a capable Shopify or eCommerce VA can become much more valuable than the title suggests.

They are not simply “sending newsletters.”

They can build and maintain the flows, segment customers based on buying behaviour, create product and bundle logic, coordinate campaigns with the Shopify catalog, keep the visual presentation consistent with the brand and review performance regularly.

The work is operational, but the result is commercial.

Traffic gets expensive when retention is ignored

Imagine a clothing brand already doing $80,000 a month.

The traffic is there.

People are buying.

But there is no restock flow. The post-purchase sequence is weak. Products are never bundled intelligently. VIP customers receive the same treatment as everyone else. Seasonal campaigns are sent to the entire list without much thought.

The store may still grow.

It is just making growth harder than it needs to be.

Every customer has to be acquired again.

Every sale has to start from zero.

Email changes that equation.

A customer who already knows the brand is easier to reach than a stranger who has never heard of it. A person who bought a hoodie can be shown the matching joggers. Someone who wanted a sold-out dress can be told when it returns. A repeat buyer can get early access to the next collection.

You are not constantly trying to find another person.

You are giving existing customers more reasons to come back.

That is why email is particularly powerful for clothing brands.

Fashion already runs on cycles of desire, discovery, identity, seasonality and novelty.

Good email simply gives those cycles somewhere to go.

The best fashion brands do not use email only to shout, “10% off.”

They use it to introduce the brand, show customers how to wear the products, announce what is new, bring back what sold out, recommend what fits with what they already own and give loyal customers a reason to return.

Paid traffic may bring someone through the door.

Email gives you a reason to keep the relationship alive after they arrive.

And for a clothing brand, that relationship can be worth far more than the first order.

When Google Ads Stop Working, Don't Increase the Budget

Katalaga Hakim — Thu, 06 Nov 2025 00:00:00 GMT

There is a familiar moment in Google Ads.

The campaigns aren't performing. Clicks are coming in, the cost per click looks uncomfortable, conversions are thin, and someone eventually says the thing that seems easiest to try:

"Let's increase the budget."

It sounds reasonable. If the campaign isn't generating enough sales, perhaps it simply needs more traffic.

But a bigger budget does not fix a campaign that is sending the wrong people to the wrong message on the wrong page.

It just gives the problem more money to work with.

A better approach is to start much earlier in the process, before touching the bids.

The first problem is often the ad itself

One account we worked through had the usual symptoms. The instinct could easily have been to start adjusting bids, testing automated strategies and looking for some magical keyword combination that would suddenly make everything work.

Instead, the first thing to look at was the ad copy.

The ads needed to do more than mention the service.

They needed to match what the person had actually searched for, address the problem behind that search and make it clear who the offer was for.

That last part matters more than people think.

A generic ad can attract a lot of clicks because it doesn't really tell anyone anything. It leaves the door open to almost everyone.

A specific ad does the opposite.

It tells the right person, "Yes, this is what you're looking for."

And it gives the wrong person a reason to keep scrolling.

That is useful.

The goal of an ad isn't simply to get the highest possible number of clicks. It is to get the right clicks.

Then we looked at everything around the ad

Ad extensions are often treated as something you fill in after the campaign is already running.

They shouldn't be.

Structured snippets can reinforce the services being offered. Callouts can answer common objections. Sitelinks can take people directly to pages that matter instead of forcing everyone through the same homepage.

None of this is particularly complicated.

But collectively, these elements give the searcher more information before they click.

They also make the ad more useful.

If someone is searching for a specific service, showing related services through structured snippets or sending them directly to a relevant page through a sitelink is considerably better than giving them a generic "Learn More" button and hoping they figure it out.

Before we touched the bidding, the account already had a better chance of attracting the right visitor.

That was the point.

Then came the search terms

This is where many campaigns quietly lose money.

The keyword might look perfectly relevant when you add it to the campaign.

Then actual searches start coming in.

Some are useful.

Some are loosely related.

Some are completely wrong.

And some are the sort of searches that make you wonder how Google decided they were relevant in the first place.

So we cleaned them up.

Broad, irrelevant queries were removed. Match types were tightened where necessary. Negative keywords were added. Keywords were grouped according to search intent rather than simply being thrown together because they happened to contain similar words.

That distinction is important.

Two keywords can look almost identical while representing completely different intentions.

One person may be ready to hire someone.

Another may simply be looking for information.

Another may be researching prices.

Another may be looking for a job.

Another may be looking for something completely unrelated that happens to contain the same words.

Google sees the query.

The advertiser has to understand the intent behind it.

That is why the search-term report is one of the most useful places to spend time inside an account.

Only then did we touch bidding

This is probably the part people tend to get backwards.

Bidding is visible. It feels like an obvious lever to pull.

If performance is bad, change the bid.

If impressions are low, increase the bid.

If clicks are expensive, change the strategy.

But bidding cannot compensate for poor inputs.

So rather than immediately trying to scale, the campaign was moved to a simpler bidding approach for a short period. The purpose wasn't to suddenly explode the account.

It was to collect cleaner information.

We wanted to see what kind of traffic the revised campaign was actually attracting and how people were engaging with it.

This is an important distinction.

A bidding strategy is only as useful as the information feeding it.

If your targeting is messy, your search terms are polluted, your conversion tracking is unreliable and your ads are attracting the wrong people, automation doesn't magically turn that into good data.

It simply automates decisions based on bad data.

Then we followed the click

There is another place where Google Ads campaigns quietly fall apart.

The landing page.

You can write a highly relevant ad, target a good search and get the click at a reasonable cost.

Then the visitor lands on a page that has almost nothing to do with what they searched for.

That creates a strange disconnect.

Someone searches because they have a particular problem.

The ad speaks to that problem.

Then the landing page talks about something else entirely.

The customer has to start the journey again.

Instead, the page should continue the conversation started by the search.

If the search suggests urgency, the page should make the urgent solution obvious.

If the search is comparison-driven, the page should provide the information and proof needed to compare.

If the search indicates that someone is still trying to understand their problem, the page may need to educate before asking for the sale.

The landing page doesn't need to look spectacular.

It needs to make sense.

Tracking can make the whole thing even worse

There is one more problem that can make an account look broken even when parts of the campaign are working.

Bad conversion tracking.

If the account isn't recording conversions properly, you don't really know what is working.

You might cut a campaign that is producing valuable customers.

You might increase spending on traffic that looks good in the interface but produces nothing useful.

You might tell Google to optimize toward an event that isn't actually measuring the business outcome you care about.

At that point, the numbers become increasingly difficult to trust.

Before making major decisions, make sure the tracking reflects what you actually want the campaign to achieve.

Otherwise you're optimizing a dashboard rather than a business.

The interesting part was what happened after the cleanup

The budget didn't need to change.

The product didn't suddenly become better.

There wasn't some secret keyword that had been hiding in the account.

The structure simply became more coherent.

The ads were more closely aligned with the searches.

The search terms were cleaner.

The extensions supported the offer.

The landing page matched the intent.

The conversion tracking gave the campaign usable information.

And the account started behaving differently.

CTR improved.

CPC became more manageable.

Quality Score improved.

Conversions began tracking properly.

Most importantly, the spend started doing useful work instead of simply accelerating waste.

That is the part worth remembering.

The same budget can produce very different results depending on the structure underneath it.

How to Run Google Ads Profitably on a $50-a-Day Budget

Katalaga Hakim — Wed, 06 Aug 2025 00:00:00 GMT

A $50-a-day Google Ads budget doesn't give you much room for mistakes.

That's $1,500 a month, assuming you spend it every day. You can't throw money at five different campaign types, target everyone who might possibly be interested in your product, and wait around for Google to figure out what you're doing.

There simply isn't enough money for that.

With a small budget, the question isn't really, "How can I get the most traffic?"

It's, "Where can I put the next dollar that has the best chance of becoming a customer?"

That changes the way I would structure the account.

You're looking for people who already have some intention to buy. You don't have the budget to spend months educating completely cold audiences or funding a collection of experiments just because they look interesting inside the Google Ads interface.

At $50 a day, I'd rather have a small amount of highly relevant traffic than a large amount of people who clicked an ad because they were vaguely curious.

The first $10 goes toward branded search.

Keep it simple. Target your brand name on exact match and don't turn the campaign into a giant collection of variations that don't need to be there.

There is a reason for giving brand search its own little corner of the budget. It captures people who already know you and, more importantly, gives Google clean conversion data from people who have demonstrated a clear relationship with the business.

But there is an important distinction here.

Brand traffic isn't prospecting.

Someone searching for your company already knows who you are. You didn't necessarily create that demand with the ad. They may have seen your product somewhere else, heard about you from someone, visited the site before, or simply remembered the name.

So you don't want to look at those conversions and convince yourself that your prospecting campaign is suddenly brilliant.

The next $30 goes toward finding new customers.

For ecommerce, that means non-branded Shopping focused around your strongest three to five products rather than putting the entire catalogue in front of Google and hoping it finds something interesting.

Start with the products that already have a reason to be there. Products people want, products that convert, products with sensible margins, and products you actually want to sell more of.

For a service business, the structure is different. Use non-branded Search around the core services and keywords that indicate someone is actually looking for what you sell. Exact and phrase match can keep the initial targeting focused rather than opening the account up to every vaguely related search Google can find.

And there is one thing I would do immediately here: exclude your branded terms.

This sounds like a small detail until you look at what happens when you don't do it.

Brand searches are often easier to convert. Google knows that. If you leave the door open, the system has a very obvious incentive to find those easy conversions.

Now you can look at your campaign and see a nice conversion rate and a healthy return while acquiring very few new customers.

The account appears to be working.

It just isn't doing the job you thought you hired it to do.

The final $10 goes toward remarketing through Performance Max.

I'd keep this simple as well. Feed-only, focused on your bestsellers and aimed at people who have already interacted with the business in some meaningful way.

Someone visited the site and left.

Someone looked at a product.

Someone added something to their cart and didn't finish.

Someone engaged with the business and disappeared.

These people are different from someone who has never heard of you.

They've already crossed part of the distance.

That's why the small remarketing budget can make sense. You're not asking cold traffic to suddenly become interested in a product they've never seen. You're giving people who have already shown some interest another opportunity to come back.

So the account ends up looking almost boring:

$10/day → branded Search

$30/day → non-branded prospecting

$10/day → remarketing

And honestly, boring is fine.

At this budget, complexity is often just another word for spreading the money too thin.

You don't need a campaign for every audience imaginable. You don't need to launch every new Google Ads feature. You don't need fifteen different experiments running simultaneously so you can feel like you're doing sophisticated marketing.

You need enough data for Google to learn from, enough control to understand where the money is going, and enough focus that the budget isn't being wasted on traffic that was never particularly likely to buy.

The structure also gives you somewhere to go later.

As the account grows, you can add more products, expand your keyword coverage, introduce additional campaign types, test different approaches and eventually increase the amount you're spending.

But the underlying logic doesn't need to disappear just because the budget gets bigger.

You're still trying to separate people who already know the business from people you're trying to acquire.

You're still trying to protect prospecting campaigns from easy branded conversions.

You're still trying to understand which products, searches and audiences actually produce profitable customers.

The numbers simply get larger.

That's the part people sometimes miss when they talk about scaling.

Scaling isn't necessarily about throwing more campaigns into the account because you have more money.

It's about finding something that works, understanding why it works, and then giving it more room to work without destroying the economics that made it profitable in the first place.

A $50-a-day account forces you to learn that discipline early.

You have to decide what matters.

You have to decide which products deserve the budget.

You have to separate acquisition from existing demand.

You have to watch the numbers closely enough to notice when you're paying for traffic that isn't doing much for the business.

And you have to resist the temptation to make the account look complicated simply because complicated accounts look impressive.

If you can make a $50-a-day account behave sensibly, you have a much better foundation for a $500-a-day account.

And eventually a $5,000-a-day account.

The budget changes.

The basic question doesn't:

Where should the next dollar go?

That is the question worth building the entire account around.