What job do reviews do in an AI answer?
Reviews have always done conversion work: a person choosing between two businesses reads them. That job has not changed. The newer job is that review platforms are among the sources assistants consult when building an answer about who to call.
Volume, recency, distribution and what the reviews actually say all become inputs to a recommendation you never see being made.
Which review signals matter most?
Four things carry the weight in a review profile: recency, consistency of cadence, distribution across platforms, and specificity about the job that was done. None of them is the star rating. They are the properties we weigh when we judge whether a review profile is useful as machine input, in rough order.
| Signal | What it means |
|---|---|
| Recency | A steady trickle reads better than a burst two years ago. Recency signals an operating business. |
| Consistency of cadence | Forty reviews in one week followed by silence is a worse pattern than two a week for a year, to platforms and to readers. |
| Distribution | Reviews across the platforms that matter in your category, not all concentrated in one place. |
| Specificity | Reviews that mention the actual service, the actual place, the actual problem are more useful as machine input than “great service, highly recommend.” |
What are the compliance rules for asking for reviews?
Everything here runs inside Google's review policy and the FTC's rule on fake and deceptive reviews. That means no incentives conditioned on positive sentiment, no gating (asking how someone feels before deciding whether to request a review), no writing reviews on a customer's behalf, and no suppressing negative ones.
This is not caution for its own sake. Review manipulation is enforceable, and a business that gets caught loses the asset it spent years building. The compliant version is slower and it compounds.
Do replies to reviews matter?
Replies are content too, and they are content you control on a surface you do not own. A thoughtful reply to a negative review does more for a reader, and gives more usable signal to a machine, than the review itself does damage.
Read in order, or jump
Twelve chapters, in the order the work has to happen. The first four are the spine; the platform chapters at the end include two surfaces we deliberately do not score, and say why.
01 How AI search picks local businesses 02 Eligibility: can AI even see your site? 03 Robots.txt: block AI crawlers or allow them? 04 Measuring AI visibility honestly 05 Entity foundations for local businesses 06 Reviews as AI input 07 Citation surfaces: where AI engines look 08 Getting ChatGPT to recommend your business 09 Showing up in Perplexity 10 Showing up in Google AI Overviews 11 Showing up in Gemini 12 Copilot, and why we don't sample itSee where you actually stand
The free Visibility Check applies this guide's measurement to your own business.
Get my free Visibility Check