How to check if your product appears in ChatGPT.

The slow way: do it by hand

The manual version works and costs about an hour per SKU per week. Open ChatGPT in a private browser window so your own chat history does not bias the answers. Write five to ten prompts the way a real buyer types them — "best merino base layer for cold-weather running," "waterproof jacket under $300 for Pacific Northwest hikes" — and run them one at a time. For each answer, record whether your product was named, whether only your brand was named, and which competitors appeared. Screenshot everything.

The trouble starts in week two. The same prompt produces a different answer — is that a model update, session variance, or your new blog post? You cannot tell from two data points. Most teams end up with a shared spreadsheet of screenshots that goes stale within a month, because the job is recurring and manual recurring work dies. The manual method is a fine one-time smell test. It is not a measurement system, and it cannot tell you the week a competitor takes your slot.

The eCommerce Insights way

1. Open the checker. Go to the free ChatGPT product visibility checker and paste the product URL. No signup for the preview result.
2. Add two sentences of category context. Who the product is for and what problem it solves. eCommerce Insights uses this to generate category-typical buyer prompts, and shows you the prompts before they run — nothing is a black box.
3. Run the prompts. The checker queries ChatGPT's shopping-aware surface, captures each full answer, records citation position, and flags brand-only mentions. LLM-backed checks can take up to a minute.
4. Read the per-prompt report. Each prompt gets a status — CITED, BRAND, or MISSED — plus the competitors that won the slots and the verbatim excerpt so you can read the context yourself.
5. Act on the highest-leverage gap. The gap summary names one concrete fix, typically a 150–250-word answer block on the PDP or a schema field the cited competitors have and you do not. Ship it, rerun in a week, measure the delta.

For continuous coverage, the paid product runs this check across every SKU on every refresh — weekly on Starter, daily on Growth — and folds the result into each SKU's citation score. See how eCommerce Insights tracks ChatGPT and the guide to ranking products in ChatGPT for the mechanism behind the fixes.

What "good" looks like

Directional benchmarks for a mid-market D2C brand, from eCommerce Insights audit data as of early 2026 (illustrative while the sample grows):

Zero citations across all prompts for a flagship SKU usually means the problem is not one PDP — it is catalog-wide schema gaps or an entity-clarity issue. Run the AEO Grader for the store-wide picture, then work through fixing product schema.

Related jobs