Discussion starter: a local SEO question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.
- AI Search
- Local SEO
Reviews on Yelp, Trustpilot or industry sites versus GBP reviews: which seem to shape local AI answers more in your markets?
Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.
1 reply
-
adminAdmin
It depends on which assistant, and that is the answer I would build strategy around. Google’s AI surfaces, including AI Overviews, AI Mode and Ask Maps, sit on top of Google’s own local data, so Google Business Profile reviews carry the most weight there. Other assistants rely more on the open web and third-party data, so platforms such as Yelp, Trustpilot, Tripadvisor and industry directories matter more in those answers, especially in categories where those platforms rank well in ordinary search. The practical approach is to find out which sources each assistant actually cites for your category and area, then prioritise Google Business Profile plus the one or two third-party platforms that keep showing up.
Why the sources differ by assistant
Each assistant has different access to local data. Google has its own place database, reviews and photos, and its AI features draw on them directly. OpenAI’s help documentation says ChatGPT search uses third-party search providers and rewrites prompts into targeted searches, which means its local answers are shaped by whatever pages rank for those searches, often review platforms and directories. Microsoft’s assistants draw on Bing’s index and Bing Places. Apple’s services have historically shown third-party reviews from providers such as Yelp alongside their own ratings. None of these companies publishes a full list of local sources or their weighting, so observation is the only way to know what matters in your market.
What seems to matter in the reviews themselves
AI answers tend to summarise what reviewers say rather than simply repeat star ratings: “praised for fast response times”, “some reviewers mention higher prices”. That makes review content more important than it was for classic local rankings. A business with a 4.8 average but reviews that say nothing specific gives an assistant little to work with; one whose reviews mention the services, neighbourhoods and qualities people ask about gives it material to match against a request like “an electrician who can come the same day in Brunswick”. Recency and volume also matter for how confidently a summary can be made.
Where each type of platform tends to matter
Platform type Where it tends to carry weight Notes Google Business Profile Google AI Overviews, AI Mode, Ask Maps, the local pack The foundation for every local business General review platforms (Yelp, Trustpilot, Tripadvisor) Non-Google assistants; categories where the platform ranks in search Strength varies by country and sector Industry directories and review sites (for example trade platforms, healthcare and legal directories) Specialist queries, where these sites rank strongly Often the most cited for professional services Community discussions Assistants that draw heavily on forums; recommendation queries You cannot manage these directly, only earn them How to find out what matters in your market
- Build a prompt set that reflects how customers ask: “best [service] in [area]”, “[service] near [landmark] open now”, “who should I call for [problem] in [suburb]”.
- Run it across the assistants your customers use, from a location in the service area where possible, and repeat each prompt several times because answers vary.
- Record the businesses mentioned and the sources cited or linked. Note which review platforms appear and how often.
- Compare with your review profile. If a competitor is recommended and you are not, check where their reviews live and what those reviews say.
- Repeat quarterly. Assistants change their sources, and one quarter’s snapshot can mislead.
Review policies differ, and they matter
Before building a review programme around a platform, read its rules. Google allows businesses to ask customers for reviews but prohibits incentives and selective solicitation. Trustpilot allows businesses to invite customers, with rules about inviting everyone rather than only happy customers. Yelp asks businesses not to solicit reviews at all, and its systems may filter reviews it thinks were solicited. A programme that sends every customer to Yelp after a job can do more harm than good there. Plan the programme per platform, not with one message for all.
Watch for this:
Spreading effort thinly across ten review platforms usually produces a handful of reviews on each and a strong profile nowhere. It is better to be clearly strong on Google plus the one or two platforms your prompt testing shows are cited in your sector.
Hypothetical example:
A plumbing business in Melbourne tests fifteen prompts across Google AI Mode, ChatGPT and Copilot. Google’s answers draw almost entirely on Business Profile data and reviews. ChatGPT’s answers repeatedly cite a trade directory where the business has three reviews while two competitors have dozens. The business keeps its Google review requests going and adds a follow-up asking recent customers to review it on the trade directory, within that platform’s rules. Over the next two quarters it starts appearing in ChatGPT answers for its suburb, while its Google visibility stays stable.
Negative themes travel further in AI summaries
One consequence of assistants summarising reviews is that recurring complaints become part of the description of your business. A single bad review has always been visible, but an assistant that says “several reviewers mention missed appointments” turns a pattern into a headline. Two things follow. First, read your reviews across platforms for repeated themes, not just ratings, and fix the operational causes; no amount of review generation will outweigh a pattern customers keep describing. Second, reply to negative reviews with specifics about what changed. Replies are read by prospective customers, and while nobody outside the platforms knows how much weight assistants give them, a reply that shows a problem was fixed gives any reader, human or machine, a more complete picture than the complaint alone.
What to prioritise
- Google Business Profile reviews, steadily and within Google’s rules, with replies that address specifics.
- The one or two third-party platforms your testing shows assistants cite for your category and area.
- Consistent business details across all platforms, so assistants do not merge or confuse listings.
- Encouraging customers to mention what the job was and where, which gives summaries something specific to say.
- Prompt testing each quarter to see whether the mix of cited sources is changing.
Google Business Profile first for Google’s AI surfaces, then the specific third-party platforms your own testing shows other assistants rely on in your sector. The question “which platform matters more” has no universal answer, but it has a findable one for each market.
Need help with this on your own site?
Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.
Add a reply Cancel reply
Replies are open to approved members. Log in or apply to join.