In one sentence
Review signals are the ratings, review counts and review text about your products across the web that AI assistants and search engines can use to judge whether your brand is worth recommending.
What counts as a review signal
- Star ratings and how many reviews you have
- What reviewers actually say: fit, quality, durability, service
- How recent the reviews are
- Where they appear: your site, marketplaces, review sites, forums
- Whether your review data is marked up with schema
How AI assistants use review signals
When a shopper asks what to buy, assistants often look for evidence that a product delivers. Reviews are some of the most direct evidence there is. Review text matters as much as the rating: "runs small" and "great for wide feet" are exactly the details shoppers ask about.
Reviews on third-party sites matter too, because assistants often cite those pages. See earned media.
How to strengthen your review signals
- Ask every customer for a review, at the right time after delivery.
- Make it easy: a short form, with photos optional.
- Respond to negative reviews and fix what they point to.
- Mark up ratings with product schema so machines can read them.
Never buy or fake reviews. It breaks platform rules and consumer protection law, and it destroys the trust the signal is supposed to show.
Review signals FAQs
Do AI assistants read reviews on my own site?
They can, if the page is crawlable. Reviews loaded only by a script after the page appears may be harder for crawlers to read, so check how your review app displays them.
Do negative reviews hurt my AI visibility?
A few honest negatives are normal and can add credibility. A pattern of the same complaint is what to fix, because assistants may repeat it.
How many reviews do I need?
There's no fixed number. More recent, detailed reviews give shoppers and AI more to go on than a handful of old ones.
Related terms