How to Block Amazon Buy for Me on Your Shopify Store (and When Not To)

How to block Amazon Buy for Me on Shopify: detect buyforme orders across your stores, opt out via Amazon, and decide fulfill-or-block per store.

Linh Nguyen · Updated 2026-07-04

The first time I saw one, I almost refunded it as fraud. An order landed on one of the stores we operate with a shipping address that didn't match the buyer's name, a customer email in the shape of buyforme.amazon..., and — a week later — a phone call asking why there was no tracking. The buyer had ordered our product on Amazon. We had never listed it on Amazon. Nobody at our end had agreed to any of it.

That is Amazon "Buy for Me," and if you run Shopify stores you should assume it can reach yours whether you opted in or not. This post is what I actually did about it: how to recognize these orders, how to opt out (and why opt-out alone hasn't been enough for a lot of sellers), and the harder question of whether you should block Amazon Buy for Me on your Shopify store at all — because they are real orders with real money, and blocking blindly is its own mistake.

What Buy for Me actually does to a store you don't control

Buy for Me is an Amazon AI shopping agent. When a shopper searches Amazon for something Amazon doesn't stock, the agent surfaces products from outside merchants' own websites, and — with the shopper's go-ahead — Amazon's system goes to that merchant's site and completes the purchase on the buyer's behalf. The merchant is not consulted. There is no opt-in step. Amazon scraped the listing, put it in front of an Amazon shopper, and placed an order through your normal checkout as if it were any other customer.

It started as a limited pilot in 2025 and has been scaling since. Amazon's own announcement puts the broader Shop Direct catalog it feeds at more than 100 million products from over 400,000 merchants, with Buy for Me itself covering "tens of millions" of those — a subset, but a subset measured in the tens of millions. That scale is the whole point: this is not a niche edge case you can wait out.

The backlash caught up on January 6, 2026, when CNBC reported that more than 180 businesses across Shopify, Squarespace, WooCommerce and Wix had found their products listed on Amazon through Buy for Me without permission. Modern Retail's reporting collected the same complaint from brand after brand — including one, Bobo Design Studio, whose founder described the effect as being turned into a dropshipper for Amazon, against her will.

Shopify's president Harley Finkelstein has been publicly blunt about the underlying model. "You can't just scrape information and expect a great experience," he told Modern Retail on January 22, 2026, arguing that agentic commerce should run on merchants deliberately publishing their catalog details — "if you don't want to be part of the Catalog, you don't have to be" — rather than on an agent extracting them. Whatever you think of the framing, it draws the line the whole fight is about: consent. Buy for Me sits on the wrong side of it by default.

The part that actually hurts operations: the tracking gap

The intellectual-property argument is real, but it is not what showed up in my day. What showed up was support load.

When Buy for Me places the order, the buyer's real relationship is with Amazon, not with you. So your Shopify store dutifully sends its shipping confirmation and tracking email — to an Amazon-routed address the buyer never checks. From the customer's side, they paid on Amazon and then heard nothing. They don't have your order status page. They can't see the tracking your store already generated. So they do the only thing left: they contact Amazon, or they find your store's contact form, and they ask where their order is. Mesa's write-up of the same problem describes exactly this loop — the shipping notification your store fires never reaches the human who needs it.

This is a specific, nasty failure mode because it defeats the systems you built to prevent it. We run automated shipping and tracking precisely so customers don't have to ask "where's my order." A Buy for Me order routes around all of it. The tracking is correct; it just never lands with a person who's looking at it. And an order the buyer thinks has gone silent is a chargeback and a bad review waiting to happen — the exact outcome our whole chargeback-prevention setup exists to avoid, arriving through a door we didn't know was open.

The week I checked the whole fleet, not one store

The genuinely dangerous version of this problem is not "we got a weird order." It's not knowing whether you're getting them at all — and across a portfolio, "check each store by hand" is not an answer.

The week the CNBC story broke in January 2026, I stopped treating it as one store's oddity and searched every store we operate for the pattern. The signal is boringly reliable: orders tied to a buyforme.amazon sender pattern, often with a name/address mismatch because the shipping identity is the end buyer while the ordering identity is Amazon's agent. An order from an unrecognized source, on an email you can't actually reply to, is precisely the shape of thing our self-built order-ops and tracking lifecycle was designed to catch. So I didn't want a per-store manual grep. I wanted one rule that runs across the fleet and flags these the moment they arrive.

That is the whole StoreFleet reflex, and it applies here cleanly: one detection rule — match the sender pattern, tag the order buyforme, raise it in the same store-health alerting feed we already watch — applied to N stores at once, instead of hoping I remember to look at each admin. Detection is the part you cannot skip, because every decision below depends on it. You can't choose to block, fulfill, or ignore a category of order you can't see.

Opting out — and being honest that it may not stick

Amazon's stated path to remove your products is to email [email protected] and ask to be excluded. Do it. But I'm not going to tell you it's a clean switch, because the sellers documenting this haven't found it to be one: multiple reports, including Heart Coding's walkthrough and the Mesa post above, describe merchants who emailed the opt-out and kept receiving Buy for Me orders afterward. Treat the email as a request that may or may not be honored on a timeline you don't control — not as a solved problem.

Which is exactly why detection has to come first and stand on its own. If opt-out worked instantly and permanently, tagging would be a nice-to-have. Because it doesn't, tagging is the load-bearing part: it's what lets you keep serving affected customers well while the opt-out does or doesn't take effect, and it's what tells you whether your opt-out actually worked — you'll know it landed when the tagged orders stop, not when Amazon says so.

If you want to go further than opt-out, you can technically try to block Amazon's fetchers at the edge (robots directives, blocking known agent user-agents/IP ranges). I'd caution that this is a moving target and a blunt one — you can break legitimate crawlers or miss the traffic entirely — so I treat network-level blocking as a supplement to detection and opt-out, never the first move.

So should you actually block Buy for Me? My take.

Here's where I'll plant a flag, because "it depends" is a cop-out and this genuinely has a right answer for most people.

Don't block blindly, and don't ignore it either. A Buy for Me order is real money from a real buyer — refusing it on principle is throwing away revenue to make a point the buyer will never hear. But it's revenue where you don't control the experience: the buyer is Amazon's customer, the margin may be thinner once you account for the support overhead, and the tracking gap means every one of these orders carries above-average odds of a "where is my order" ticket or a dispute. The right call is a function of two numbers only you know — your margin on the product and how much support load each of these orders actually generates.

• Thin margin, high support cost, brand-control matters? Lean toward blocking — opt out, and if orders keep coming, cancel-and-refund the tagged ones with a polite note pointing the buyer to your own store.
• Healthy margin, and you'd rather have the sale? Fulfill them, but fulfill them knowing — tag every one, and consider a proactive touch (a manual email to the shipping contact with tracking) to close the notification gap before it becomes a ticket.
• Not sure yet? Tag and watch for a few weeks before deciding. That's the honest default.

The one option that is never acceptable is the one most stores are living in right now: not knowing they're receiving these orders at all. Everything else is a legitimate business judgment. Blindness is not.

The open question Shopify hasn't fully answered

There's a policy backdrop worth watching, because it may make this decision for you eventually. In mid-July 2025, Digiday spotted that Shopify had quietly added agent rules to its merchants' robots.txt, declaring that "automated scraping, 'buy-for-me' agents, or any end-to-end flow that completes payment without a final review step is not permitted." The principle underneath is blunt: no agent should finish a payment without a human taking a final look. The policy pushes back on agents that drive an end-to-end payment flow with no final human review step on the buyer's side.

I'm not a lawyer and I won't pretend to adjudicate where an Amazon agent completing a checkout on a Shopify storefront lands against that policy — that's a genuinely open question, and it's Shopify's and Amazon's to resolve, not mine. But it's the reason I'd file Buy for Me under the same heading as every other agentic-commerce shift: something to monitor deliberately rather than react to in a panic. If you're thinking about how your catalog should show up in AI shopping surfaces — on your terms, with consent — that's the constructive flip side, and it's the subject of our guide to selling into ChatGPT and agentic storefronts and the broader map of AI agents in Shopify for 2026.

For the defensive posture, though, the playbook is the same one that protects a multi-store operation from any external shock: know what's hitting your stores, apply one rule across all of them, and decide on purpose. Buy for Me is just the newest reason that preventing account and operational surprises across a portfolio is a system, not a one-time fix. If detecting and tagging these orders across every store from one place sounds like the missing piece, that unified order-and-alerting layer is exactly what StoreFleet is built to be — book a demo on your own stores and we'll set the buyforme rule up live.