Diagnosis
Why AI doesn't recommend your products — and how to fix it
You have great products and real customers, but when a shopper asks ChatGPT or Perplexity for a recommendation, your brand never comes up. It's rarely about quality. It's almost always about what the AI can — and can't — read about you. Here are the nine most common reasons, with the fix for each.
The mental model: AI recommends what it can read and confirm
An AI shopping answer is assembled from two ingredients — what the model learned in training and what it retrieves from the live web right now. In both cases it gravitates to products it can describe specifically and confidently. If your data is unclear, unreadable, or unconfirmed, the safest move for the model is to recommend someone else. Every fix below closes that gap.
The 9 reasons — and the fix for each
1. Your product content is JavaScript-only
Gap: crawlers see an empty shell; the facts load after JS. Fix: put titles, descriptions, specs and price in the server-rendered HTML.
2. Descriptions are vague or "vibes-first"
Gap: "effortlessly cool" matches no query. Fix: state material, dimensions, use-case, and who it's for — in plain words the model can quote.
3. Missing Product structured data
Gap: no machine-readable attributes. Fix: emit Product/Offer/AggregateRating schema with real values.
4. No independent corroboration
Gap: only your own site mentions you, so the model can't confirm the claim. Fix: earn reviews, roundups, comparisons and forum mentions that name the brand.
5. You block AI crawlers in robots.txt
Gap: GPTBot/ClaudeBot/PerplexityBot are disallowed. Fix: allow them explicitly (see our visibility guide).
6. No content around the buying questions
Gap: nothing on your site answers "best X for Y." Fix: publish crawlable guides that answer those exact questions.
7. Ambiguous or generic product titles
Gap: "The Weekender" tells the model nothing. Fix: add category and key attribute — "The Weekender — 35L waxed-canvas carry-on duffel."
8. Thin or missing reviews/ratings
Gap: no social proof to weigh. Fix: collect reviews and expose them (with Review schema) so the model can factor them in.
9. You've never measured it
Gap: you're guessing, so you can't prioritize. Fix: sample the engines and get a baseline — the rest of this list becomes a ranked to-do.
Fixing five of these badly beats fixing one perfectly. AI visibility is cumulative: each readable, corroborated signal raises the odds the model names you instead of the competitor next to you.
Because AI answers are non-deterministic, a single manual check tells you almost nothing. You need repeated sampling to know which of these nine reasons is actually costing you — and against whom.
FAQ
Why doesn't ChatGPT mention my brand?
Usually because it can't read or confirm your product data — JavaScript-only content, vague descriptions, missing structured data, no independent mentions, or blocked crawlers. It recommends brands it can read clearly and corroborate.
Does structured data help?
Yes. Product/Offer/Review schema expresses your attributes machine-first — exactly what retrieval layers lift and cite. Missing schema is one of the most common blockers.
Is my JavaScript hurting AI visibility?
It can. If key product facts only render after client-side JS, many AI crawlers see an empty page. Put the facts in the server HTML.