Sam Wright

MD of @Blink_eCommerce, the specialist Shopify digital marketing for stores with large product catalogues.

[Blink - The expert Shopify marketing agency @Blink_eCommerce]

🚀 Our episode on the Conversion Tracking Playbook with @getelevar's Brad Redding just went live on YouTube!

We cover:

✅ SEO + LLMs
✅ Shopify Editions + AI
✅ Google Marketing Live takeaways
✅ Future of SEO/AI

Watch here → https://t.co/mIMWRURkkW

#Shopify #SEO #AI

[Blink - The expert Shopify marketing agency @Blink_eCommerce]

🚀 Shopify Editions just dropped with an AI block generator – but did you know you can enable it in ANY theme?

https://t.co/G1xxyy8Ini

[Sam Wright @hearpen149]

The ChatGPT + Shopify integration has cemented something we’ve been circling for a while:

The single biggest thing brands can do to prepare is start thinking like data engineers.

Why? Because Shopify isn’t just a storefront - it’s a product data platform. It’s already (or at least, should be) the central source of truth for your products. Taxonomy, variants, pricing, availability - all of this is already feeding into your Google and Meta ads. Soon, it’ll be powering ChatGPT too.

But this can go a step further. You can apply the same thinking to your creative messaging.

Product descriptions, collection copy, blog content - these are no longer just for human readers. With ChatGPT soon plugged directly into Shopify, your brand story becomes data. And that data needs to be structured, high quality, and machine-readable.

At Blink, we’re building creative systems the same way we build product feeds.

We break down brand storytelling into modular components - value props, differentiators, FAQs, benefits - and tag them consistently across surfaces: PDPs, collection pages, etc.

We use a shared glossary to define these blocks. Then we deploy them everywhere: prompts, schema, content.

This isn’t a new idea. Google’s Retail Playbook has been encouraging brands to surface their value props everywhere for years. What’s changed is who’s listening - and now, those systems are generative. They need structured signals to understand what makes your product different.

So, if you're wondering how to prepare for the new AI-powered commerce world:

– Start treating Shopify like a PIM, not just a storefront
– Turn your brand story into structured, reusable components
– Build these into your templates using metafields (yes, you might need to remap later – but structure now, adapt later)
– Push that data into every surface – PDPs, schema, feeds, internal search, and beyond

It’s a creative job - but it’s also a systems job. And the brands who figure that out early are going to win.

[Sam Wright @hearpen149]

For those that haven't heard the news, new code made public by OpenAI has suggested that it is working on allowing purchases from Shopify to complete directly inside ChatGPT.

SEO is the only marketing channel that depends on the structure and content of your own site - rather than working inside someone else’s platform like Meta or Google Ads. That’s why this potential ChatGPT + Shopify integration is such a big deal.

If this plays out, it could introduce a whole new product discovery journey - one based on conversation and exploration, rather than interruption and ad formats.

Right now, Meta dominates discovery. But it only really works for products with clear problem-solution messaging. There are thousands of brands out there selling great products that require more education, or that live within complex taxonomies that just don’t fit neatly into an ad.

This is where good SEO work - such as organising product data and taxonomy - suddenly matter a lot more. Clean tagging. Consistent attributes. Proper schema. If this becomes the backbone of discovery, brands with well-managed catalogues will win.

And that’s where Shopify is particularly well placed. Its standardised product model and centralised architecture mean the building blocks are already there.

If OpenAI + Shopify really deliver on conversational discovery, it’s not just a content or SEO opportunity - it’s a product data one. And this is exactly the kind of work we specialise in.

[Sam Wright @hearpen149]

Across our data, we're pretty sure sure that AI overviews are eating away at clicks for informational content.

Over the past 12 months across these 75 Shopify stores, blog impressions are up year on year- but clicks are consistently down. Importantly, there’s been no tangible impact on revenue.

Worth noting: April’s data isn’t complete yet, so we’re comparing a partial month this year to a full month last year, so you'll need to take that into account when looking at the chart.

[Sam Wright @hearpen149]

One of the more under-discussed aspects of AI (especially in ecommerce) is just how dependent everything still is on good data engineering.

There’s a lot of excitement around what AI can do in terms of modelling, pattern recognition, clustering and prediction. And that’s fair, some of it is genuinely amazing.

But what doesn’t get talked about enough is the invisible layer underneath: the pipelines, schemas, event tracking and ETL processes that make it all work. Without that, the outputs are unreliable, or worse, misleading.

In theory, AI reduces the need for complex queries or manual reporting. In practice, if the data isn’t well-structured or complete, you just end up getting the wrong answers more quickly.

That’s why data engineering feels increasingly like the moat around any serious AI implementation.

There's also this wave of interest in automation - tools like n8n, Airbyte, and others. And while they’re great in the right hands, there's plenty of people built quite complex pipelines on top of setups that were never that stable to begin with.

Without a solid grasp of how to model data - or even the principles behind it - these layers of automation can make things worse. You end up with systems that look impressive on the surface, but aren’t trusted internally because no one’s really sure how they work. Or whether the numbers are even accurate.

[Sam Wright @hearpen149]

As catalogues grow, the role of your homepage shifts...

Most stores start with a tight product range and build a homepage around that one flagship offer. Then the product range grows. Suddenly the homepage has to do everything - tell the brand story, explain new categories, and try to rank for high-intent terms all at once.

That’s usually when performance dips.

We saw it recently. A site that used to rank well for a competitive product term started slipping - not because anything changed with the product itself, but because the homepage no longer matched the query. Meanwhile, newer competitors were dedicating their entire homepage to a single product, stacked with proof points, educational content, and clear CTAs.

It’s a classic case of outgrowing your homepage. It was built for where the business was and not where it’s going.

At that point, you’ve got to pick a lane:

- Go brand-first: embrace the wider range and rework the homepage to sell the story, the proposition, and the long-term vision. It needs to be tight.

- Or go product-first: anchor everything around the flagship product and make it your search, conversion, and trust-building powerhouse.

But straddling both? Half brand, half product, not quite clear on either - that’s where rankings drop and customer journeys start to feel messy.

[Sam Wright @hearpen149]

Organic forecasting is a mess, but here’s how we model it anyway...

Most ecommerce brands can’t get a clean read on organic revenue.

* Tracking in GA4 is, broadly, terrible

* Other platforms aren't reliable ether

* Attribution and channel cannibalisation is a rabbit hole

But if you’re asking a client to invest £50k+ over 12 months, you need a model even if the data is incomplete.

Here’s the structure we use:

– Start with GA4 revenue, broken down by channel
– Reallocate a conservative % of Unassigned/Direct to Organic (we typically use 30%)
– Model growth from month 3 onward – we use 3-8% YoY as a baseline
– Build in costs: a front-loaded retainer (e.g. £5k for 3 months, £3k ongoing)
– Highlight the net revenue position by month, not just the total uplift
– Show the peak out-of-pocket point based on COGS – i.e. how far under water it goes before it turns profitable

This kind of model won’t be exact but it does make the opportunity legible.

It also allows for better pricing conversations, more structured performance terms, and a clearer business case for fixing tracking issues early.

Forecasting organic is never clean. But it doesn’t need to be perfect, it just needs to be transparent and commercially sound.

[Sam Wright @hearpen149]

Following on from my recent post about Facebook ads generating 20x ROAS while overall revenue stays flat (see link in comments) there’s something else worth digging into.

If a campaign is genuinely delivering a 20x return - not just on paper, but in terms of actual incremental revenue - that’s transformative.

It can take a brand from zero to £5-10 million a year. And when this happens, you’ll feel it fast.

You’ll see it in your bank account, not just your ad dashboard. The business becomes easier to scale. Things click.

That kind of growth usually comes from getting something fundamental right - a creative breakthrough, a shift in positioning, an insight that makes your product resonate in a way it hadn’t before.

But if Meta or Google Ads are showing a 20x return and your overall revenue is flat - or even shrinking - then something doesn’t add up.

You likely don’t have a performance breakthrough. You’ve (most likely) got a channel cannibalisation problem.

In other words - the platform is claiming credit for sales that would’ve come through anyway. Organic, email etc - they’re all getting eaten by your paid campaigns.

It still amazes me how many paid media freelancers and agencies miss this. They’ll report the headline ROAS and call it a win, without ever asking whether it’s actually moving the needle for the business.

[Sam Wright @hearpen149]

Shopify hasn’t completely removed complexity, it’s just changed the type.

One of the reasons Shopify became the default platform for so many merchants is because it made ecommerce feel simple.

But as a business scales, that simplicity gives way to a different kind of complexity.

This is something we see after a lot of migrations. For example, problems that would previously needed engineering resources can now be solved by configuration. This is great - simply, it means that more people can do them.

However, now you’re no longer dealing with hard development problems – you’re managing operational ones.

This is where a lot of migrations and redesigns fall short. The new site goes live, the project wraps up, and no one’s planning for the operational structure that follows.

We work with a lot of partners on re-platforming, and the very best ones bake all this into part of the process. If you'd like a recommendation on this drop me a message.

[Sam Wright @hearpen149]

A 20x ROAS sounds amazing, but it might be telling the wrong story.

We recently spoke to a Shopify brand seeing huge returns on Meta ads during their peak season - 15-20x ROAS on some campaigns.

This is impressive on paper, but it raise two important questions:

* Are you excluding existing customers from your campaigns?

If not, you’re paying to re-acquire people who were going to buy anyway. ROAS goes up but real profit doesn’t.

* Are other channels down while Meta is up?

If so, Meta may just be claiming sales that would’ve come through organic, email, direct etc.

When you have a correctly configured your paid campaigns (e.g. with proper exclusions, tighter attribution windows, deduplication across channels), your numbers might look lower - but they’re more accurate.

Instead of showing a 20x ROAS on Meta, the real number (after filtering out repeat customers or over-attribution) might be 6x.

Getting this stuff right helps you make better decisions, even if the dashboard looks less exciting.

[Sam Wright @hearpen149]

Most SEO agencies fail on hashtag#Shopify - not because they’re bad at SEO, but because they don’t understand how Shopify actually works.

They’ll build a keyword list, audit a few titles, and tweak the homepage copy.
But the real growth lives in the messy middle:

✅ Understanding how Shopify collections work (and where they fall short)
✅ Using tags and metafields to build scalable, structured taxonomy
✅ Creating schema based on the right product attributes
✅ Making template updates safely - without bloating the theme or breaking things
✅ Managing internal linking and breadcrumbs in a flat system
✅ Working at scale - hundreds (or even thousands) of updates each month, not just a handful of pages

None of that fits neatly into a 50-slide strategy deck but, if you've got a large catalogue at least, it’s where the value is.

The root problem is that most SEO teams are used to CMSs with folder hierarchies. Shopify isn’t built that way. So they try to force a tree structure onto a flat platform, and give up when it doesn't work.

Your SEO doesn’t need more audits. It needs the right team shape.

For large-catalogue Shopify stores, that usually looks like:

✅ SEO strategist
✅ Data engineer
✅ Software engineer
✅ Feed optimisation specialist

That’s the gap Macaroni Software is designed to fill.

Even without this though, the main point is that Shopify SEO is not “just SEO.” It’s it's own thing altogether.

[Sam Wright @hearpen149]

Many Shopify stores don’t realise they’re sitting on a data problem - not an SEO one.

If your store has more than a few hundred SKUs, you're no longer just managing a front-end experience. You're running a lightweight PIM (product information management) system - whether you want to or not.

But Shopify isn’t built like a traditional PIM. It lacks hierarchy. The default taxonomy is flat. And trying to scale with tags alone quickly becomes a mess.

That’s where most stores (and agencies) fall short. They treat product data like a set of labels, not a structured system.

But if you start thinking like a PIM - setting up parent/child categories, defining attributes clearly, tagging consistently, using metafields to their full extent - everything else improves:

* You can scale your collections cleanly

* You can build rules-based product tagging logic

* Your Google Shopping feeds become more accurate

* Schema becomes richer and easier to automate

* You unlock long-tail SEO opportunities at scale

We’ve seen this pattern over and over. This isn't about optimising for users or search engines, but building an internal structure that scales.

[Sam Wright @hearpen149]

Every year I say the same thing, and every year I hope this is the one where it lands.

And yes - I’m saying it again. Because I’m an eternal optimist. 😀

Right now is the time to start planning if you want your SEO performance to peak in Q4 - the most important quarter of the year for almost every ecommerce business.

Now, I know what you're thinking. Of course we’d say that. New business is great, right?

Absolutely. But let me explain why...

When we implement SEO changes we start to see meaningful movement in around 60 days. The chart below shows the impact of improved collection page content just two months after go-live.

That’s great news. It means if we hit publish today, we’d be seeing real results by early June.

The challenge is that it's never as simple as that.

To get to the point where we can ship those changes, we usually need 30 to 60 days of prep work. Keyword mapping, fixing template issues, generating content, aligning on priorities etc all takes time.

And once you factor in project sign-off approvals, stakeholder buy-in, and internal processes, we’re suddenly staring straight into Q4 itself.

We get it - ecommerce planning is always focused on the next quarter, not the one after. And I hate the usual SEO clichés as much as everyone - “it’s a marathon, not a sprint” etc.

But if you're even thinking about getting your site into shape for Q4, now is the time to act. And we’d be more than happy to talk through what that might look like.

[Sam Wright @hearpen149]

We’ve been seeing great results lately from consolidating blog content – especially where multiple articles are cannibalising a key commercial term.

In one recent example (screenshot below), we spotted three blog posts all targeting the same topic - and all competing with each other and the main category page.

We took the best bits from each post, combined them into one clear and valuable hero piece, and updated the internal linking to support it.

The result was that the main category page jumped from position #3 to #1 for the target search term, and we’ve estimated a solid revenue uplift as a result.

If you’re sitting on years of content, it’s worth checking if any of it is quietly holding back your most important pages.

[Sam Wright @hearpen149]

I've had a bunch of new connections recently, so it's a good time to go back to the basics!

So, what makes Blink - The only search marketing agency for Shopify stores with large product catalogues different from other agencies?

Our approach is built on three core principles:

1. eCommerce marketing for large catalogue stores is fundamentally different to any other kind of business

2. The things that make it different – like the importance of taxonomy and working at scale – are data-led challenges, not creative ones

3. Shopify’s consistent structure gives us the ability to implement repeatable, effective strategies across projects

These principles led us to build Macaroni Software – our own eCommerce SEO platform. We’ve also built our team to reflect this same thinking.

Instead of a traditional agency model built around content writing and link building, our team includes data specialists, software engineers, and SEO, PPC and email marketers who’ve worked extensively with high SKU counts.

It’s this blend of repeatability and deep technical capability that allows us to consistently deliver 100% channel revenue growth within 12 months.

If you're working with a large catalogue store on Shopify, this is what we do best.

[Sam Wright @hearpen149]

It’s absolutely wonderful to receive feedback like this.

We’ve been working with Gary Ingram and the team at The Diamond Store for some time now, focusing on their migration to hashtag#Shopify.

In terms of SEO migrations, this has been one of the most complex projects we’ve ever worked on - with a legacy platform, multiple data sources, and decades of technical debt to contend with.

We’re extremely pleased with the results so far, and it genuinely means a lot to receive this kind of feedback.

It’s also been a real pleasure working with The Diamond Store team, and a massive thanks to Elodie P. for leading the project from our side. As always, you’ve done an outstanding job.

Huge thanks as well to Myles Ejegi-Memeh and Jake Hardy. It’s been fantastic to work with partners with their level of expertise and enthusiasm.

Now that the replatforming is complete, we’re really excited about the next phase – shifting our focus to growth!

[Sam Wright @hearpen149]

#Shopify stores: I’m going to spill my secrets on how to increase your organic revenue right now.

No joke - make these changes and you’ll see an immediate uplift in organic revenue. Let’s go:

✅ Set up server-side tracking (we recommend using Elevar - https://t.co/Afn22p5Add)

✅ Make sure your tracking is correctly configured - pay particular attention to checkout extensibility and your cookie consent setup

✅ Check your event sequences - make sure they’re firing in the right order otherwise you'll get loads of Direct revenue

Next – head over to Google Ads:

Exclude brand terms from PMax

Separate prospecting and remarketing - create dedicated campaigns for new customers vs returning ones

Use audience exclusions - exclude existing customers, high-frequency purchasers or loyalty members using first-party data or CRM integrations

Check for declining AOV vs increased PMax spend - could be a sign of saturation

Enable “New Customer Acquisition” in PMax

In Meta:

Optimise for conversions - not impressions or reach

Exclude warm audiences - people who’ve recently visited your site or interacted with your brand

The eagle-eyed among you will spot one thing here. None of this has anything to do with SEO performance.

It’s all tracking and channel cannibalisation.

And yet, it has a huge impact on your organic revenue.

If you’re a Shopify store with more than 100 products, as a rule of thumb we’d expect organic revenue to account for around 20% of your overall mix.

But if you’re using GA4 reporting out of the box, most stores are seeing 6-10%.

That’s a huge difference when it comes to making business decisions, and it's one of the biggest challenges we face in the work we do.

I'm sure I've also missed a bunch of things here, so if anyone has any tips please share

[Sam Wright @hearpen149]

Here's a little preview of the reporting view from the Macaroni Software Shopify app.

It brings Search Console data directly into your store, and automatically segments it by page type.

Lots more to come here too!

[Sam Wright @hearpen149]

Making good progress with our custom tagging app!

We're building this to solve a very specific (and very common) problem – working with large catalogues on hashtag#Shopify. Think thousands of products and thousands of tags, spread across messy datasets with inconsistent structures.

Our app lets us combine metafields, tags, other attributes and custom logic to quickly and cleanly tag products – then export everything for bulk upload.

Clean product data is critical to what we do. This tool makes it faster, more accurate, and way less painful.