Google brings server-side purchase tracking to Shopify
A week on from Shopify's Everywhere Edition, the striking thing is how flat the response has been. The brands we work with have mostly shrugged it off, and when we dig into why, there appear to be a few simple reasons.
Some of the headline features, like the agentic dashboard, have been live for weeks or months already. Meanwhile, a lot of what is new is not really actionable. The dashboard is more or less the whole feature - there is nothing to actually go and do with it.
And there is a familiar fatigue to a release where a great deal gets announced and then you wait, sometimes for months, for it to be fully live. Campaign Autopilot, the AI marketing tool we covered last week, is the clearest example. In fairness it has only been a week, but if past releases have taught us anything it is that the gap between announcement and a feature that actually works can be months.
The lead story this week though has definitely got a lot more attention - Google is about to make a major change to how Shopify purchase data reaches GA4, and most stores will have it switched on without lifting a finger.
On to the news!
Google brings server-side purchase tracking to Shopify
Anyone who has heard me talk for more than five minutes knows that tracking is one of the biggest problems we run into working with Shopify stores, so anything that moves Shopify purchase data off the browser is on our radar (I wrote the long version up for Addingwell recently). On June 18th Google emailed Shopify store owners to say that from July, GA4 will start receiving purchase events directly from Shopify's servers for any store with the Google & YouTube app installed.
The data moves server to server, bypassing the browser entirely, and it is on by default, so if you have the app this will be live in your property with nothing to configure. Because it does not depend on a shopper accepting cookies or a tag firing in the browser, the purchase itself gets recorded more reliably than today.
I would not call tracking my specialism, so when I posted about this on LinkedIn I mostly wanted to hear from the people who do it full time. John Barnes, Dan Truman, Dan Bartow and Edward Upton, all of whom really know this area, weighed in, and the thread was the most useful thing I read on it all week. This is my synthesis of where they came out.
The consensus was that it is a real step up if you only run GA4 and maybe Google Ads, and not enough if you need more, because the native integration does not do what a dedicated server-side setup does: cookie restoration through the checkout, ITP handling, event monitoring and governance, the same gaps I have written about for Addingwell. The release also said nothing about server-side Google Ads, which is a big part of all this, and there is a fair question about whether this is really about helping merchants or about Google shoring up its own data. One small actionable point - because the measurement method changes partway through the year, any year-on-year comparison spanning the switch will look off, so add an annotation in GA4 and a note in any reports.
π Read the full threadβ
Google's advice for the AI search era is, helpfully, to keep doing good SEO
Google published a piece this month aimed at CMOs trying to work out what AI search changes about their marketing. The short answer it gives is reassuring - not much. Features like AI Mode and AI Overviews are built on the same core ranking systems as ordinary search and pull from the same index, so the work that earns visibility in AI results is the work that already earns visibility in search. Good SEO is good GEO, in its words, and there are no separate AI tricks to chase.
The useful part for a large catalogue is what it tells you to stop doing. It says do not write for bots, the systems read language the way a person does, so keyword-stuffed copy and content chopped into artificial little snippets are wasted effort. It says do not chase generic content, the thing that earns a place in an AI answer is the first-hand, specific material only you can write, the deep product knowledge a large catalogue actually holds.
We have seen this first-hand - sites leaning on generic AI-written copy have tended to perform poorly, while the ones with real, specific product detail hold up. And it says do not get pulled around by third-party AI-visibility scores, because Google does not evaluate those tools and they have no access to its internal metrics. It points to the new Search Console AI-features reporting and Merchant Center reporting as the measurement baseline instead.
One point is worth noting. Google says directly that special AI text files like llms.txt are not needed, at least for Google Search. We looked at this back in May when Shopify started adding those files to every store. The two are really different things - Google is talking about its own AI search while Shopify's files are aimed at agentic commerce, so one does not override the other. Itβs worth adding that there is still no real evidence the files are being used for much yet, on either side, including by Shopify itself. But if your only concern is Google, this is Google telling you not to spend time on it. As ever, the foundational work is where the return is.
π Think with Google on SEO in the AI search eraβ
Shopify Scripts stop running on 30 June
This has been a fixture of the last few newsletters, and now we are inside the final week. On 30 June the legacy Scripts engine is switched off for good, and Shopify has confirmed the date will not move again after two previous deadlines came and went. If you are on Plus and still run any Scripts, this is the one with a clock on it.
The reason to treat it as urgent is how it fails. Anything still running on Scripts, discount logic, shipping rules, payment customisation, stops executing with no storefront error and no alert in your admin. The discount simply stops applying, or the shipping rule drops out, and you find out from a margin gap or a customer query rather than a warning.
The fix is migrating that logic to Shopify Functions, and the sensible move is to have the Functions live and tested in a development store before the 30th. Stacking and ordering behave differently in Functions, so anything with complex discount logic usually needs more than one pass. If you inherited your store's checkout customisation from a previous agency and are not sure what is still riding on Scripts, this is the week to open the Script Editor and check.
π Shopify's guide to migrating from Scripts to Functionsβ
We've had Editions, now it's nearly time for DotDev
Shopify's DotDev conference runs in Toronto on 21 and 22 July, about a month out. It is partner-only and developer-first, and tickets have already sold out, so this is less a "go and register" note than a flag for what might be coming. DotDev tends to follow an Edition, and this year is no exception: the Everywhere Edition has been and gone, and the developer conference is next.
I had an interesting conversation with Fred Cohen at Quickfire Digital recently, and he made the very astute observation that this Edition was oddly thin on the developer side. A lot of it was already live, the headline AI features were existing tools reskinned, and there was little real substance for developers.
This Edition also came ahead of DotDev rather than alongside it, which raises an interesting question - is Shopify pulling the merchant-facing material forward on purpose and is saving the developer announcements for a stage where they get proper attention?
Either way, if you build on Shopify or rely on people who do, DotDev is worth keeping a closer eye on this year.
π DotDev 2026, Shopify's developer conferenceβ
A new standard for the knowledge you feed your tools
Google Cloud published the Open Knowledge Format this month, and it is worth a look for reasons that have nothing to do with storefronts or search. OKF is a way of holding the knowledge an organisation runs on, what a metric actually means, how two systems join, why an old process was dropped, as a set of plain markdown files with a little structure on top, so that both people and AI tools can read it without anything getting lost in translation. It came out of Google's data-analytics division, not its shopping products, and it is aimed at internal knowledge rather than anything customer-facing.
We are covering this it because it formalises something we have spent a lot of time on at Blink. Most of how we work now runs through Notion as a structured knowledge system rather than a pile of documents, and BlinkTank itself is built on it.
The part that tends to surprise people who see it is not that it stores things neatly but that it reads across everything we have captured and flags where two sources disagree, so a contradiction surfaces instead of sitting buried. What makes that work is the structure underneath, and OKF is an attempt to standardise exactly this. The interesting part is not Google's involvement but that the same pattern is turning up independently all over the place, in open-source tools built for it and in other agencies arriving at near-identical setups without comparing notes.
The takeaway for a store at your scale is not that you need to adopt OKF this week. It is that the way you keep your own knowledge, your product data rules, your reporting caveats, the reasons behind past decisions, is starting to behave like infrastructure.
Teams that hold that knowledge in a structured form an agent can read will get more out of every AI tool they touch, and the ones holding it in scattered documents and a few people's heads will keep paying to reassemble it. Structuring what you know for machines is becoming a real advantage, and OKF is a sign the rest of the industry is catching up to it.
π Google Cloud on the Open Knowledge Formatβ
Partner shout-out
This week it is Finishing Line, a UK fulfilment partner built for established DTC brands. They are a good fit for the kind of store this newsletter is written for - high order volumes, replenishment and repeat-purchase products, and real international ambitions.
We have referred clients to them and the feedback has been consistently good, which is the only reason they are here. If fulfilment is starting to hold your growth back, Josh Arrowsmith and the team are worth a conversation.
π Finishing Lineβ
As ever, thanks for reading.
If you found this useful, I would love it if you could share it with someone you think would find it interesting - whether that is someone managing a large-catalogue Shopify store or anyone else who follows this kind of thing.
See you next week.
Sam
Who am I and what does Blink do?
I'm Sam Wright, Managing Director and co-founder of Blink SEO. We're a search marketing agency that works exclusively with large-catalogue Shopify stores - the kind where the challenge isn't just visibility, but helping customers find the right product in the first place.
We specialise in SEO and PPC for merchants managing hundreds or thousands of products. Most of our work starts with taxonomy: how a store is structured and categorised, and what that means for organic performance, paid efficiency, and on-site discovery.
If you found this useful and aren't already subscribed, you can sign up at blinkseo.co.uk/blinktank.
If you run a large-catalogue Shopify store and want to talk about organic or paid performance, reply to this email or find me at blinkseo.co.uk.