Your Most Threatening Competitor Just Activated Unified, Agentic Commerce on Shopify

On March 24, 2026, Shopify turned on Agentic Storefronts by default. Overnight, millions of stores became discoverable and shoppable inside ChatGPT and Microsoft Copilot, with Google AI Mode and the Gemini app rolling out. No project charter. No steering committee. No ownership politics. No integration roadmap.

Now ask the question every enterprise B2B commerce leader in the room is thinking but won't say out loud: how long would that have taken at your company?

If your honest answer is "eighteen months and a re-platforming," you're in good company. And I want to argue that the reason isn't the one we usually reach for.

We wanted to. We struggled to.

I spent years leading digital commerce inside a large enterprise manufacturer: well-run, well-resourced, with leadership that genuinely wanted to win in digital. We achieved some incredible results in spots and starts, but struggled to significantly transform for fully unified ecommerce. It's a story I often hear repeated by my peers — strong leadership alignment and talented teams, but "half-way there" stalls during enterprise scale. That's what makes this story worth telling.

In one of our water filtration businesses, we rebuilt the brand store on a major marketplace and activated AI-powered guided selling: an end-to-end needs-based buying experience that asked shoppers about their water concerns in their language, then recommended the right filtration system. It let us target non-brand and non-technical searches we had never been able to engage and convert. The results were not subtle:

So we did the obvious thing: we tried to bring the same capability home. We scaled guided selling to our owned site and to other sites.

And that is where it stalled. Not because anyone said no. Because the systems couldn't say yes. The owned site we connected it to didn't have the commerce capabilities to deliver the same end-to-end experience. Product data lived in one place, customer data in another, transactions and conversion data somewhere else. Each channel had a different set of capabilities, so the experience a shopper got depended on where they happened to land.

The practical result: we deferred customer experience and sales to platforms that could close them. Customers we had earned with our own content and our own guidance completed their purchase on someone else's storefront, at lower margins, with almost no data coming back to us. Repeat purchase and cross-sell — the economics that make a filtration business beautiful — became nearly impossible to run.

We sure did want to. We just struggled to.

The tax nobody budgets for

Here is what I've come to believe after years in rooms like that one: most enterprise commerce programs don't fail on strategy or alignment. They fail on activation. Leadership aligns in a quarter. Systems take years.

We spent a decade calling the ambition "omnichannel" — one connected experience wherever the customer shows up. What most large companies actually built was multichannel instances with integration glue: the same capability rebuilt for each channel, each rebuild a little more expensive and a little less faithful than the last. Call it the multichannel repetition tax. You pay it in licensing, in agency hours, and in the customer journeys that quietly break between touchpoints.

Shopify draws the distinction cleanly. Omnichannel connects separate systems; unified commerce runs everything on one: inventory, customers, orders and channels in a single source of truth. That's not semantics.

It's the difference between maintaining fifteen bridges and living on one continent.

The companies with the most complex ERPs, the most customized sites and the most siloed CRM data are exactly the organizations that need unification the most, and they are structurally the least able to accomplish it. That's the paradox.

AI makes the paradox expensive

Agentic commerce raises the stakes because AI agents are ruthless about consistency. An agent recommending a product needs one trustworthy version of the truth: specs, price, availability, fit. Fragmented systems produce fragmented answers, and an agent that gets a blank or a contradiction simply moves to a merchant that doesn't give it one.

At that enterprise, I used "The Killer Be's" — four plain-language principles as the test for every buyer-journey investment. They turn out to be a pretty good test for AI readiness, too:

• Be found. Is your product data structured so an agent can read and evaluate it?
• Be relevant. Does it surface content that addresses the problem and questions the buyer actually has, not just the part number?
• Be helpful. Can the experience engage and guide a decision the way your best salesperson would?
• Be accessible. Can the buyer act, check out, reorder or ask for help, wherever intent peaks?

If your answer to any of these depends on which channel or system the buyer happens to touch, you don't have an AI strategy. You have an AI pilot.

Meanwhile, in the mid-market

Now picture a mid-market brand that started on a unified platform. What does it get, more or less by default?

Its catalog is set up once and surfaced across ChatGPT, Copilot, Google AI Mode and Gemini from a single admin. Its operators work alongside Sidekick, Shopify's merchant assistant, built on Anthropic's Claude. And in September, Anthropic released blueprints for shopper and merchant agents, and Shopify published a public reference implementation built on them. That means a brand can stand up its own guided-selling agent, inside its own branded experience, on top of the same unified data.

Read that again with the filtration story in mind. The capability we have fought to activate — AI-powered semantics and guided selling that converts early-journey buyers — is becoming something a 20-person company can switch on and maintain. No additional ad spend. No additional fees. And it lands on a platform where the sale, the data and the next purchase all stay with the brand.

And the quiet power behind those filtration results was semantic matching: a shopper searching for "blue" finds the product your catalog calls "azure," without a team endlessly rewriting product records. On a unified platform, that's built in. Shopify's Semantic Search matches intent rather than keywords, and Shopify says it has structured more than a billion products so AI agents can find the right one. In an ERP-attached commerce suite, that's usually another vendor, another integration, another budget line.

So here's the uncomfortable question: if your largest competitor came to market with these capabilities before you did, would you be under threat?

Well, you probably are under threat. You're just watching the wrong competitor. The brand coming for your category isn't your enterprise peer, stuck in the same re-platforming quagmire that you are. It's the focused challenger that launched last year on a unified stack — and it's already showing up in the AI answer your prospective buyer just asked.

So: should large B2B enterprises move to Shopify?

The honest answer is many of them, now. Most of them, at the edge first.

The momentum is real. On its recent earnings calls, Shopify reported B2B GMV up about 80% in Q1 2026 and then 76% in Q2 — more than double the pace of the platform overall. And Shopify's growth was broad-based across regions: North America 33%, Europe 42%, Asia Pacific 30%. Wholesale and direct-to-consumer now run from the same admin, and native B2B features have moved well beyond the targets of the stalled enterprise plan.

But be clear-eyed about the fit. Commerce for complex manufacturers often means RFQ workflows, EDI, multi-level purchase approvals and contract pricing driven in real time by the ERP. Those still tend to require middleware, apps or custom development — and that's the part of the conversation ERP vendors emphasize and commerce vendors gloss over.

The smarter path for most large B2B organizations isn't a big-bang migration. It's starting at the edge:

1. A new brand, a new channel, or a new segment launched unified from day one, where the multichannel tax hasn't been paid yet.
2. The commerce layer on a unified platform, with the ERP staying the system of record, connected deliberately rather than rebuilt for every channel.
3. AI capability deployed where the data is already consistent, so pilots become scalable programs instead of demos.

The risk isn't moving too fast. With every IT budget getting a fresh look, the risk is dismissing the model outright — and widening the gap while new entrants race to the top of the answer.

A note for the platforms and their partners

There's a lesson here for the ecosystem, too. Default enrollment made millions of merchants eligible for AI channels. Eligible isn't active. The gap between being listed and being recommended is where adoption is won or lost, and it gets closed by education, not announcements: showing merchants — and the enterprises watching them — exactly how to be found, relevant, helpful and accessible on every surface.

The partnerships that win this next chapter will be the ones that keep teaching long after the launch press release. Fewer, better, bigger bets.

One truth, every surface.

We sure did want to. The next generation of commerce leaders won't have to struggle to.

Ryan T. Satre is a Revenue Growth Architect who has led digital commerce, marketplace and partner-channel growth inside large enterprise organizations. He is a member of the Master B2B Forum and the author of the Total Market Viability™ (TMV) framework. ryantsatre.com