You're in ChatGPT Shopping but never suggested. Here's the difference
Last reviewed
Short answer
Being listed means your products are in the pipe: Catalog-eligible, syndicated, findable. Being recommended means ChatGPT picks you when a shopper asks. Presence is a checklist you can finish in a week. Selection runs on relevance to the shopper's constraints, operational facts like price and availability, and evidence gathered from the wider web, which is a practice rather than a checklist.
Two problems wearing one name
Merchants say not showing up in ChatGPT about both problems, and the fixes are different, so the diagnosis matters. Listed is binary: your products either flow through Shopify Catalog into ChatGPT's product data or they do not, and OpenAI states Shopify data integrates automatically once eligible.
Recommended is competitive. When a shopper asks for the best linen shirt under a hundred dollars, several listed merchants can answer, and ChatGPT picks. OpenAI says merchants are ranked on factors like availability, price, quality, and whether they are the maker or primary seller. Passing the eligibility gates puts you in the room. It does not make you the answer.
How to tell which problem you have
Test presence first, because it is cheap and binary. Ask ChatGPT directly about your named products, your brand plus product type, in a fresh session. If your own products do not surface when named, you have a listing problem, and the troubleshooter walks the gates in order.
If named products surface but category questions never include you, you are listed and not recommended. That is the harder, better problem, and everything below is about it.
What selection actually runs on
Three layers, in rough order of how much you control them. Data fit: whether your product's stated attributes match the shopper's constraints, which is fully yours and covered by the feed spec. Operational facts: price competitiveness, availability, being the maker or primary seller, which are business decisions wearing a data costume.
Then evidence. Third-party research, labeled as such: Profound's 2026 analyses found feed-sourced offers dominating the top result position, Reddit supplying roughly a third of shopping citations, and reviews and Q&A on product pages correlating with rank. OpenAI does not publish this layer. The research keeps finding it anyway, and it matches the broader mention studies.
The uncomfortable half of the answer
Data fit is a week of work. Evidence is a year of work. A store with complete attributes and no reviews, no community presence and no third-party mentions has done everything this site's product-data pages say and can still lose the recommendation to a competitor the internet talks about.
The honest sequence is still data first, because incomplete data disqualifies you from matches evidence could have won, and because it is fast. But the ceiling on recommendations is reputation, and no product feed carries reputation. Getting recommended is the full playbook.
Frequently asked questions
My products show when I search my brand but never in category questions. Broken?
Not broken, outranked. Brand queries test presence, category queries test selection. Your listing works, and the competition is now on attribute fit for the shopper's constraints, price and availability, and the evidence layer of reviews and mentions.
How long does it take to go from listed to recommended?
Data fixes can show in days to weeks, since product data flows continuously through Catalog. Evidence accumulates on review-and-mention timescales, months. Anyone quoting a guaranteed timeline is selling something OpenAI says cannot be bought.
Can I be recommended without being listed?
Partially. ChatGPT also reads the open web through its crawlers, so a store can be cited in answers from its pages while its products are missing from shopping results. That half-presence usually means Catalog eligibility fails while crawling works, which is worth fixing for the full surface.
Does price really matter that much?
OpenAI lists price among its stated merchant ranking factors, and third-party research ties competitive prices to selection tags. You do not have to be cheapest. You do have to know that pricing is now visible to a comparison engine shoppers trust.
Sources
Keep reading
Score your own catalog
Entitled grades every product against the rules on this page, explains each failure, and fixes them with AI you approve. The score is free forever.