#BlinkTank: what Shopify Editions means for large catalogue stores
We are a day late with this week's #BlinkTank. Shopify Editions launched late yesterday afternoon UK time, and we wanted time to go through it properly.
So this is a bumper edition. Shopify Editions is obviously the main story, with our early opinion on what actually matters for a large catalogue. But there is a lot of other ground to cover too - the EU cancel button that becomes mandatory on Thursday and affects UK stores selling into Europe, a Gulf ceasefire that could (hopefully) ease landed costs into Q4, and the Amazon and Perplexity hearing. Plus a shout-out to Blend Commerce, who have just updated a very useful set of CRO benchmarks.
On to the news!
What stands out in Shopify Editions
Shopify's Spring '26 Editions launched yesterday afternoon, 150+ updates across the platform, and we will go deeper as we work through it. Two things stand out first. One is that a good deal of what is being presented as the agentic headline is not new. The edition leads with the Agentic section in the admin, the hub for managing your AI channels, seeing what sells through them and watching how your products show up. That is the same thread we have tracked across recent issues: the Agentic dashboard, the agents.mdfiles, and UCP turning on by default on every store. The pieces have been arriving for months, and what Editions does is name them and bundle them.
The other, and the more interesting one, is what the bundling is for. The message Shopify is ramming home is that it wants to be your single point of truth, the one place you both manage your products and market them, across the store, AI channels, ads and retail. They are calling it the "Everywhere Edition", but the promise underneath that word is simplicity: one platform, on by default, no separate feed, no setup.
The thing to watch is how much that promise is doing. Everything in one place, one source of truth, easier. But - and this is a very important point to make - consolidation is not simplicity.
Pulling product data, attribution, marketing and agentic channels under one roof stacks a great deal of complexity rather than removing it, and the more the platform centralises, the more it matters that someone knows how to run it well. Coming from an agency that specialises in this kind of work, that is exactly what you would expect us to say. But it is also true - as Shopify absorbs more of the marketing stack, the expertise to use it properly is more important than ever, not less.
🔗 Shopify Editions: Spring '26
Shopify Catalog, and why your product data still decides the outcome
The headline of the agentic section is Shopify Catalog: Shopify takes the product data you already manage, structures and enriches it, maps it to the Standard Product Taxonomy, infers attributes like colour, material and size, and exposes that record to AI channels like ChatGPT and Copilot through the Universal Commerce Protocol. It is the data layer sitting under that Agentic section. Shopify says data syndicated this way drives 2x conversion in AI chats, with no separate feed to build. Most of this is not a new capability though. The attribute inference is the same Shopify Magic machinery that has been suggesting categories and filling in attributes for some time, and the taxonomy, UCP and agentic dashboard are all already live.
The part worth attention is what we have seen using it - the inference is only as good as the product description it reads from. A thin or half-finished description gives you wrong or empty attributes, and no amount of Shopify enrichment fixes a record that never had the detail in it. Shopify all pretty much says this in its own FAQ, which says it infers attributes only when the data is clear enough and that your own product data quality is the most useful lever you have.
This is the thing to be aware of as the agentic announcements keep coming. Once the structure is there, AI can automate an enormous amount on top of it, but it cannot will detail into existence that was never entered. Most stores will not be automatically opted into good AI discovery by this. The product-data work still has to be done, and Catalog raises the reward for having done it rather than removing the need.
🔗 Shopify: what Shopify Catalog is and how it works
Shopify is now doing attribution, and a lot of brands will prefer it to GA4
The other piece worth flagging is attribution. There is now a standalone Attribution view under Growth, showing sessions and sales by channel, conversion rate, AOV, a new-versus-returning split, and a selector to switch between attribution models, with ad cost and impression data now flowing in from supported marketing apps. As with Catalog, the capability is not brand new: Shopify has offered first click, last click and last non-direct click attribution since 2023. What Spring '26 does is promote it into a prominent view and pull in ad cost. The thing to know is that all three models are single-touch and click-based, with no multi-touch, data-driven or cross-device modelling, which is exactly why tools like Triple Whale and Northbeam exist.
The underlying data is the catch. It runs on Shopify's own session and referrer data, and Shopify does not store UTM or click IDs in the order by default, so the source is brittle across sessions. The classic issue is a customer who clicks a paid ad, returns days later by typing your URL directly, and buys, at which point the original source is dropped and the sale books as Direct. That is how 30 to 50% of revenue routinely ends up labelled Direct, the same misattribution we keep flagging for organic in GA4.
A lot of brands will prefer this to GA4 and Triple Whale anyway, precisely because it is easy and sits right there in the admin, and convenience wins on adoption. That is fair, but the risk is that an easy number you check daily is the kind you stop interrogating. With GA4 you at least have levers to clean things up, server-side tracking, enhanced conversions, a data-driven model. Here the levers are mostly upstream hygiene: clean UTM tagging on every paid and email link, persisting that source into the cart and order so it survives a multi-session journey, and checking your tracking fires at purchase. You can feed it cleaner inputs, but you cannot make a single-touch model see a multi-touch journey. Treat it as one model's directional view, not the answer.
🔗 Shopify: how attribution models work in marketing reports
Campaign Autopilot
Campaign Autopilot, in early access, promises to generate and run marketing for you from your store data, email through Shopify Messaging, Shop Campaigns and Meta ads, with you setting guardrails and approving or rejecting each action.
The concept of marketing on autopilot is an appealing one. The truth is we do not yet know how it behaves in the real world, and the track record of tools that promise to run your campaigns for you is that the gap between the pitch and what actually happens tends to be wide. We would want to see it working on real accounts before trusting it with anything that matters.
🔗 Shopify: Campaign Autopilot
Sidekick now reaches into your other apps
Sidekick became an operator in the last edition, building apps and Flow automations rather than just answering questions. What Spring '26 adds is App Extensions: Sidekick can now pull data from and take action inside third-party apps, launching with 15-plus partners including Klaviyo, Loop, Smile, Judge.me, Matrixify and Yotpo. Ask it about campaign performance or loyalty stats and it answers from the partner app alongside your store data, or it stages a change in that app for you to confirm. It also keeps working in the background across multiple chats. Shopify says weekly active shops using Sidekick were up 4x year on year in Q1, which is presumably why it is being positioned as the place you run the store from.
The same point applies as everywhere else in this edition. Staging a change for confirmation is the right design, and it keeps a human in the loop, but it also means the value depends on you actually reading what it has staged rather than waving it through. An operator that touches your live apps is more useful and higher-stakes than an assistant that just talks, and worth treating that way until you trust it.
🔗 Shopify: Sidekick App Extensions
Other updates
Three smaller changes that matter more for a large catalogue than their billing suggests.
Native A/B testing for themes and checkout, through Rollouts, now sits in the admin with scheduled changes and automatic winner selection, which displaces a category of paid testing tools.
🔗 Shopify: Rollouts → https://help.shopify.com/en/manual/markets-new/rollouts
Variant-level publishing lets you control which variants are available by channel and market, useful when your range differs across regions.
🔗 Shopify: publishing and unpublishing product variants → https://help.shopify.com/en/manual/products/variants/publish-variants
Customer accounts now keep shoppers signed in for up to 365 days, which should help returning-customer attribution and repeat purchase friction.
🔗 Shopify Editions: Spring '26 → https://www.shopify.com/editions/spring2026
And while we are on Shopify platform changes: Scripts
Not part of Editions, but the firm deadline in the same window. Shopify Scripts stop executing on 30 June. If anything on your store still runs on a Script - tiered or volume discounts, shipping-rate logic, payment-method gating at checkout - it stops working at the end of the month unless it has moved to Shopify Functions or a public app. For a large catalogue this is worth a deliberate check, because Scripts often do work at checkout that is easy to forget is there until it breaks.
🔗 Shopify: Functions and the Scripts migration
The EU withdrawal button becomes mandatory on 19 June
From Thursday 19 June, any store selling to consumers in the EU has to provide an electronic withdrawal function - a clearly visible way for a customer to cancel an eligible order online, without having to log in, call, or email. This is EU Directive 2023/2673, which adds a new article to the Consumer Rights Directive on a simple principle: cancelling a purchase should be no harder than making it. The part that catches people is who it applies to. It is not limited to EU-based businesses, so if you are a UK store selling to EU consumers you are covered, the same way GDPR reaches beyond EU borders, and it is not something most large-catalogue merchants can file under "not us".
What compliance requires is a prominent withdrawal button or link, a two-step confirmation where the buyer enters their details and then confirms, and an automatic confirmation email, available throughout the 14-day cooling-off period. This is a separate, new obligation from the model withdrawal form that already exists, the button is the new bit. The cost of ignoring it is real: Shopify's own guidance notes that without a compliant function after 19 June, the 14-day window can be extended to 12 months and 14 days, on top of fines that vary by member state. An order someone could cancel for a year is a much worse position than a 14-day window.
On the practical side, Shopify is releasing cancellation rules and a self-serve cancellation flow today, 17 June, letting buyers request cancellations on unfulfilled items from your admin. That helps, but be careful before assuming it closes the obligation: the directive is about a withdrawal declaration with a durable confirmation and a record you can produce later, and Shopify's flow is built around cancellations and returns rather than that full declaration-and-evidence chain. So waiting for Shopify is reasonable, but check what the native flow actually covers, and if there is a gap you can build a compliant button yourself or use one of the App Store apps that handle the full declaration, confirmation and audit trail for a few euros a month. It has specific exemptions by product type, so it is one to run past someone who knows your jurisdiction.
🔗 Shopify: EU right of withdrawal compliance
The Gulf ceasefire, and what a reopening could mean for landed costs
Zooming out from the Shopify admin for a moment, because this story is further from the day to day, but is obviously quite important. The US and Iran reached an initial agreement this week to extend their ceasefire and reopen the Strait of Hormuz, the waterway Iran has effectively controlled since the war began on 28 February and through which around a fifth of the world's oil normally passes.
The two sides signed a memorandum of understanding this week, with Iran gradually reopening the strait and the US lifting its blockade of Iranian ports. It is a framework rather than a settled peace, the ceasefire runs for 60 days, the nuclear questions are still unresolved, and even on the most optimistic reading it will take months for the energy disruption to fully unwind. But it is the first real sign of the pressure easing.
You can already see it at the operational edges: Gulf Marine Services, an offshore-energy support firm, confirmed this week that the last of four vessels evacuated from the Gulf has returned to its contract. One company's ships are not a trend, but they are the kind of small signal that shows up before the macro numbers move. And here is why a search and ecommerce newsletter is bothering with it: for anyone importing stock from Asia, the past two years of Red Sea and Hormuz disruption have meant the long route round the Cape, ten to fifteen extra days in transit, and Asia to Europe container rates well above pre-crisis levels. A real reopening reverses that, shorter voyages and capacity freed back into a fleet that is already oversupplied, and downward pressure on freight rates, though a sudden return can bunch ships into European ports and spike rates briefly before the trend takes over.
So the important question: could we see lower landed costs in the back half of the year heading into Q4? Possibly, and a quieter freight market going into peak would take some pressure off margins where large-catalogue stores feel it most. It remains about as unpredictable as these things get, and a framework signed on Friday can look very different a fortnight later. But after two years of arrows pointing the wrong way, it is a glint of something positive, and worth a thought as you plan Q4 buying.
🔗 Reuters via Axios: US and Iran agree to extend ceasefire and reopen Hormuz
A Google Ads bidding change to note
Briefly, for whoever runs your paid accounts. Google has made a few bidding and budgeting changes, the one worth noting is a backend change from 17 August to how bidding targets are optimised for budget-limited campaigns. Google expects a short calibration period where some advertisers see minor performance swings, with account notifications starting 6 July.
The practical point is to be cautious of turbulence in late August, and to check your CPA or ROAS targets still reflect the business going in. The other two changes, an expansion of Smart Bidding Exploration and a new Promotion mode for scheduling ROAS tolerance and budget around peak periods, are worth a look if you lean on automated bidding.
🔗 Google: bidding and budgeting updates
Amazon and Perplexity argue over who controls agent access
The case we flagged last week was heard at the Ninth Circuit on 11 June, in Seattle, before a three-judge panel. No ruling yet, and appellate decisions take weeks or months, so this is a marker rather than a resolution. The question remains whether an AI agent acting on a user's explicit instruction counts as an authorised visitor to a logged-in site, or whether the retailer can block it under a hacking statute written in the 1980s.
One exchange from the hearing is worth repeating, because it gets to the heart of the agentic-commerce question. Amazon's counsel argued that Perplexity's access is different in kind from its peers, pointing out that Brave and Microsoft Edge have the same technical capability to shop on Amazon on a user's behalf but decline to, on security grounds. The counter is the one Perplexity made in its brief, that a user directing an agent to act is authorisation. However the panel lands, the opt-in, protocol-led model, where a merchant chooses to expose itself to agents, looks more durable than agents acting without permission.
🔗 Background on the Amazon and Perplexity dispute
A faster way to request collaborator access to client stores
This one is for the agency, developer and support people on the list rather than the merchants, but if it applies to you it is a nice time-saver. Shopify Collab Request is a new Chrome extension from Micheal Concannon, aimed at anyone who sends collaborator access requests fairly often.
It lets you save access profiles, store URLs and collaborator codes, and keep track of pending requests, so you are not rebuilding the same request and digging through notes for codes every time you onboard a store. As Micheal puts it, this is the part of Shopify work nobody writes poems about, but if you do it a lot, the saved clicks start to add up.
🔗 Shopify Collab Request on the Chrome Web Store
Partner shout-out
This week it is Blend Commerce, a Shopify CRO agency that many in the industry will already be more than aware of. Adam Pearce and the team do CRO the way it should be done, and their thinking is squarely the kind we like. They have just updated their 2026 ecommerce conversion rate benchmarks, and it is worth looking at if you are trying to work out whether your store is actually performing or just feels like it is.
What makes it useful is that it refuses to hand you a single number. It pulls the headline figures together - IRP Commerce around 1.7%, Dynamic Yield's global benchmark near 2.7%, Littledata's Shopify median at 1.4% with the top 20% of stores above 3.2% and the top 10% above 4.7% - and then makes the point that matters: the average hides more than it reveals once you account for price point, device mix and traffic source. Their advice is not to chase the blended conversion rate but to find the weakest point in your funnel and make that the lever. That lines up with how we think about organic too, where the aggregate number is often the least useful thing you can look at.
🔗 Blend Commerce: 2026 ecommerce conversion rate benchmarks
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.