On June 17, 2026, Shopify's Summer '26 Editions moved AI Collection Sort, predictive cross-sell, and a merchandising insights panel into the core platform, for free. That logic used to live in paid apps. For founders in the sort, upsell, and search-merchandising category, this is an extinction-level event, and pretending otherwise will cost you a year.
- Native AI Collection Sort and predictive cross-sell replace the baseline that most sort and upsell apps were selling.
- Apps that sold the baseline are exposed. Apps that sell judgment, control, and cross-channel depth are not, yet.
- The honest exposure test is one question: does Shopify's free default do roughly what your app does for most of your installs?
- Survivors differentiate hard, move up-market, expand surface area beyond the absorbed feature, or sell while the asset still has value.
On June 17, 2026, Shopify shipped native AI merchandising in its Summer '26 Editions, and a category of apps got smaller overnight. AI Collection Sort orders products by predicted conversion. Predictive cross-sell handles cart-aware upsell. A merchandising insights panel reports on what is converting. All three are now in core Shopify, free, and all three were paid app categories the day before. If you build in this space, the question is not whether this affects you. It is how exposed you are and what you do this quarter.
I have watched Shopify absorb features from every seat in the ecosystem. I built the Partner Program, so I understand the platform's logic for what it builds versus what it leaves to apps. I founded and sold a software company that lived in this ecosystem, so I know what it feels like when the platform moves toward your roadmap. And I operated brands that paid for these exact apps, so I know which ones merchants kept and which ones they cut the moment a native option appeared. This post is the honest read, not the reassuring one.
The merchant-facing version of this drop, which apps to cancel and which to keep, is in the Summer '26 read for brands. Read it, because that is the cancel-or-keep decision your customers are making about you right now. This post is the other side of that table: what to do when you are the app on the cancel list, or one drop away from it. The broader pattern of where native AI compresses the opportunity is in how AI is reshaping Shopify apps, and this is that pattern arriving in a specific category.
The instinct after a drop like this is to argue that your app is different, that the native version is worse, that merchants will not actually switch. Some of that is true. Most of it is cope. The native version does not have to be better than your app. It has to be good enough for the median merchant who was paying you for the baseline, and it is free and already installed. That is the bar, and it is lower than founders want it to be.
What Shopify moved
into the core, and
what it used to cost.
Be precise about what changed, because precision is what separates a real threat from a panic. Shopify did not vaguely add AI. It absorbed three specific jobs that mapped one-to-one onto three app categories.
| Native feature | App category it hits |
|---|---|
AI Collection Sort | Collection sort and product ranking apps Rule-based and conversion-based sorters |
Predictive cross-sell | Cart upsell and cross-sell apps |
Merchandising insights panel | Merchandising analytics and reporting apps |
Native A/B testing | Basic CRO and split-test apps |
Notice the pattern: every one of these absorbed the baseline version of a category, not the advanced version. Shopify shipped good-enough conversion sorting, not a full merchandising rules engine. It shipped cart-aware cross-sell, not segment-specific bundling logic. That distinction is the whole game. If your app is the baseline that Shopify just shipped, you are in trouble. If your app is the advanced thing the native version stops short of, you have room, for now. Knowing exactly where the native line falls is the most valuable thing you can do this month.
Why the platform wins
this fight by default,
not by being better.
The reason native features beat apps is not quality. It is distribution and friction. The native feature is already installed on every store, it is free, it carries no security review, no extra app in the admin, no monthly line item, and no integration risk. For the median merchant, that combination beats a slightly better paid app that they have to find, vet, install, and pay for. The app has to be a lot better to overcome free-and-already-here, and most apps in an absorbed category are only a little better.
This is the structural disadvantage every app carries against its own platform, and it is why I keep telling founders that churn is the symptom, not the problem. When Shopify absorbs your baseline, churn does not spike because your product got worse. It spikes because the alternative became free and frictionless. You can lower churn at the margins with better onboarding, but you cannot onboard your way out of the platform giving away the job you charge for.
"The native version does not have to be better than your app. It has to be good enough for the median merchant, free, and already installed. That is the bar, and it is brutal."
There is a pricing consequence too. Once a free native baseline exists, your pricing power collapses toward the gap between native and you. If your app does 20 percent more than the native feature, you can charge for that 20 percent, not for the whole job. Founders who priced for the whole job and now compete with free have to re-anchor their pricing on the increment they still uniquely deliver, which is a much smaller number. I walk through that re-anchoring in the app pricing strategy post, and it is the conversation most exposed founders avoid for too long.
The one-question test
for whether your app
is actually exposed.
Strip away the reassurance and the exposure test is a single question. For most of your installs, does Shopify's free native default now do roughly what your app does? Not for your best customer, not for the power user with the complex setup, for the median install. If the answer is yes, you are exposed and you should act this quarter. If the answer is no, you have room, and you should spend it widening the gap before the next drop closes it.
Run the test honestly per segment. It is common for an app to be safe for its enterprise installs and dead for its long tail of small merchants. That split is itself the strategy signal: if the native feature only threatens your low end, the move is to lean into the up-market customers the native default underserves, which is exactly the direction the next two sections point.
What the survivors sell:
judgment, control,
and cross-channel depth.
Apps survive an absorption event by selling the things a one-size native default structurally cannot. There are three, and the strongest survivors sell more than one.
Judgment. The native sort optimizes for predicted conversion, full stop. It cannot run a merchant's real strategy: lead with margin, push the new launch, clear the overstock, protect the hero SKU, merchandise by segment. An app that lets a merchant encode strategy the platform ignores is selling judgment, and judgment does not commoditize.
Control. The native feature is a black box tuned for the average store. Merchants with real merchandising teams want rules, overrides, explainability, and the ability to say why a product ranks where it does. Control is the opposite of a default, and the customers who need it are the customers who pay the most and churn the least.
Cross-channel depth. Shopify's native merchandising lives on Shopify. An app that unifies merchandising across Shopify, marketplaces, retail, and the agentic surfaces opening up is solving a problem the platform is not even trying to solve. Surface area beyond the absorbed feature is the most durable moat, because it is the one Shopify is least likely to chase. The brands the strongest apps serve are the ones thinking about the full Sidekick versus app stack question, where native AI keeps expanding into the app layer year over year.
Differentiate, move up,
expand, or exit.
Pick deliberately.
If you are exposed, there are four honest roads, and the failure mode is taking none of them while telling yourself the native feature will not catch on. Pick deliberately, this quarter.
| Move | When it fits |
|---|---|
Differentiate hard | You can deliver judgment or control the native default cannot, fast. |
Move up-market | Native threatens your long tail but not your complex, paying installs. |
Expand surface area | You can credibly extend beyond Shopify into channels it does not own. |
Sell the asset | The category is closing and your revenue still has value to a consolidator. |
The road founders avoid most is the last one, and it is often the right one. A clean app business with real revenue is worth more to a consolidator the day before a category gets commoditized than the year after, and there is an active market for these assets. I covered who is buying and why in the app M&A market post, and the uncomfortable truth is that the best exits in an absorbed category happen while the asset still looks healthy, not after the decline is obvious in the numbers. If you are early enough to read this and act, you have optionality that a founder one drop later will not.
For founders not in this category but watching it happen, the lesson for what you build next is to start from a defensible position, not a feature the platform is one drop from shipping. The how to build a Shopify app in 2026 guide frames the same point: build on surface area, judgment, and cross-channel depth, the things the platform is structurally unlikely to absorb.
This is the pattern,
and it runs every
Editions. Plan for it.
None of this is new, and that is the point. Shopify has absorbed app categories at a steady clip for years. It watches which third-party features the platform cannot live without, builds a good-enough native version, and ships it to every store. Merchandising was simply next. Whatever category you build in, assume the most popular, baseline version of your feature is on a list somewhere, and that the platform will eventually ship it for free.
The founders who survive this pattern over multiple drops are not the ones with the best version of the absorbed feature. They are the ones who treated every Editions as a clock, kept moving up-market and out into surface area the platform does not own, and never let the baseline become their whole value. The founders who get caught are the ones who built a great version of a feature and then defended it emotionally when the platform shipped a free one. The same read from the partner side, drop by drop, is in what the Editions mean for app owners.
Native AI merchandising ate a category this June. It will not be the last category, and the only real protection is to build where the platform is structurally unlikely to follow. Run the exposure test honestly, pick one of the four roads, and do it this quarter rather than next year. In an absorbed category, the founders who move early keep their optionality. The ones who argue that the native version will not catch on lose it.
What app founders ask me
about native AI
merchandising.
It did not kill all of them, but it absorbed the baseline of the category. Native AI Collection Sort, predictive cross-sell, and a merchandising insights panel now do for free what most basic sort and upsell apps charged for. Apps selling only that baseline are in serious trouble. Apps selling judgment, control, and cross-channel depth the native default cannot match still have room. The broader compression pattern is in how AI is reshaping Shopify apps.
Ask one question honestly, per segment: for most of your installs, does Shopify's free native default now do roughly what your app does? Judge it for the median install, not your best power user. If yes, you are exposed and should act this quarter. It is common for an app to be safe up-market and dead in its long tail, and that split tells you which way to move. The pricing consequence is covered in the app pricing strategy post.
There are four honest roads: differentiate hard on judgment or control the native version lacks, move up-market to complex installs the default underserves, expand surface area into channels Shopify does not own, or sell the asset while it still has value. The failure mode is taking none of them and assuming the native feature will not catch on. The market for these assets is real, as covered in the app M&A market post.
Adding AI to the same job Shopify now does natively means racing the platform on its own field with a fraction of its data and distribution, which you will lose. The better move is to differentiate on a job Shopify is not doing: encoding merchant strategy, giving control and explainability, or unifying merchandising across channels the platform does not own. Build on surface area, not on the feature the platform just shipped, as the build a Shopify app guide argues.
Because the native feature is already installed on every store, free, with no security review, no extra app in admin, no monthly fee, and no integration risk. For the median merchant, that beats a slightly better paid app they have to find, vet, install, and pay for. Your app has to be dramatically better for a paying segment, not marginally better, to survive. This is why churn is the symptom, not the problem when the platform absorbs your baseline.
Is your app exposed to native AI, and what is the right move?
I built the Shopify Partner Program, sold a software company in this ecosystem, and operated brands that bought these apps. When the platform absorbs a feature, I have seen it from every seat. That is the view I bring to whether your app differentiates, moves up, expands, or exits.
Start a conversation See the case studies →A note on sources: the June 17, 2026 Summer '26 Editions stream, the native AI Collection Sort, predictive cross-sell, merchandising insights panel, and native A/B testing features are drawn from Shopify's Editions announcements and contemporaneous reporting on the drop. The exposure framework, the four roads, and the read on what survives are mine, from building the Partner Program, founding and selling a software company in this ecosystem, and operating brands that bought these apps.