Learn / Assistants
Which assistants send shoppers to your own checkout
Shopify documents ChatGPT shoppers completing on the merchant’s own checkout in an in-app browser, and Google AI Mode, Gemini and Microsoft Copilot completing in the AI channel’s Shopify-powered direct checkout. The split determines where your risk sits, and it has moved once already, so it is worth rechecking rather than memorising.
There is one documented split worth knowing, and it decides where your risk sits.
The documented positions
Shopify’s help centre says ChatGPT users "complete their purchase on your online store checkout in a ChatGPT in-app browser, or in a new tab when customers use ChatGPT web".
The same documentation says shoppers using Google AI Mode, Gemini or Microsoft Copilot complete their purchase "in the AI channel’s Shopify-powered direct checkout without having to leave their chat conversation".
What each side demands
- Your own checkout. Your data has to get you chosen, and then your checkout has to survive a cold session in a narrow in-app browser. See how to test checkout in an in-app browser.
- In-channel checkout. Your checkout is not in the path, so your published data carries the whole sale with no page and no second chance. See will AI agents recommend my products.
The overlap is large: both require correct price, variant level availability, identifiers and stated terms. The difference is whether a checkout defect can cost you an order.
Do not memorise this
The assignment has already changed once. Retail Dive reported on 25 March 2026 that OpenAI was retiring the first version of Instant Checkout, moving purchases that had been completing inside ChatGPT back to merchant sites.
Anything written confidently about this in early 2026 was wrong by spring. Treat the split as a thing to recheck on a cadence, not a fact to learn once. See what changed when Instant Checkout was retired.
The practical position
Build for both. That sounds like twice the work and is not: the data work is shared, and the additional cost of the checkout half is one test run cold on a narrow viewport, tied to checkout and theme changes.
Questions
Can I just optimise for one pattern?
Not safely. Both are live simultaneously, and the assignment has already changed once during 2026. A store optimised only for in-channel selling is exposed the moment an assistant starts sending traffic to its checkout.
Which pattern is better for merchants?
Checkout on your own store keeps the customer and the order data, at the cost of owning the failure. In-channel checkout removes your checkout as a risk and leaves your data carrying the sale alone. Neither is strictly better.
How do I keep this current?
Read your platform’s documentation on a cadence rather than relying on any article. This is the single most changeable fact in the whole subject.