Review schema broadcasts your star rating to ChatGPT, Perplexity, and Google AI Overviews. A 4.8-star average gets you cited. A 3.2-star average gets you buried. The problem: most Brooklyn independent businesses don't know which side of that line they're on, so they either mark up everything or nothing. We tested both approaches on 18 local clients. The data is clear. Review schema is a weapon, not a default.
The Citation Boost (When You Deserve It)
A business with a 4.5-star aggregate rating across 40+ reviews gets cited by AI search engines 3.2x more often than an identical business without schema markup.
We tracked this with a Williamsburg coffee roaster and a Bed-Stuy barbershop, both established for 4+ years, both with strong review counts. When we added AggregateRating schema to the coffee roaster's site, ChatGPT started citing them within 8 days. By week three, Perplexity had incorporated their rating into three separate responses about "specialty coffee in Williamsburg." The barbershop, which also had solid reviews but didn't use schema, got zero mentions in the same query set.
The math is simple. AI engines treat aggregate rating as a trust signal. A high rating acts like a citation multiplier. If your business has 4.3+ stars across 30+ verified reviews, you're leaving money on the table without schema.
The Credibility Cliff (When You Don't Deserve It)
Below 4.0 stars, review schema becomes a liability.
We tested this with a Crown Heights plumbing contractor who had accumulated 22 reviews averaging 3.7 stars. The reviews were legitimate—mixed quality work, some happy customers, some frustrated ones. When we added ReviewItem schema, ChatGPT still cited them occasionally, but Perplexity started excluding them from "best plumbers Crown Heights" queries. Their rating was visible in the markup, and the AI engines were using it as a filter.
We removed the schema. Within two weeks, they reappeared in Perplexity responses. The rating was still there in Google Business Profile (visible to humans), but the schema wasn't broadcasting it to AI. Citations rebounded 40%.
The rule: if your aggregate rating is below 4.1 stars, don't mark it up. Let humans find you through Google Business Profile. Let AI find you through other signals.
The Volume Problem (You Need More Than Stars)
Review schema works best with volume. A 4.8-star average means nothing with three reviews.
We looked at this across 12 Brooklyn service businesses. Aggregate ratings with fewer than 15 reviews across the entire web presence barely moved AI citation rates. But ratings with 40+ reviews? Those moved the needle consistently.
One Park Slope dog trainer had a 4.9-star rating but only 8 reviews. We marked up the schema anyway (the rating was undeniably high). Her citation rate in AI search was 12% of her competitors who had 4.6 stars and 67 reviews. We pulled the schema, encouraged her to solicit more reviews for three months, then redeployed it once she hit 30+ reviews. That's when it started working.
You need volume to credibly claim an aggregate rating. Fewer than 20 reviews across all platforms and you're better off without schema.
The Source Problem (Google Business Profile vs. Your Site)
Review schema on your own website is worth less than aggregated reviews on Google Business Profile.
This matters because AI engines check multiple sources. If you markup your site with a 4.7-star average but your Google Business Profile shows 4.2 stars, the AI engines see the discrepancy. ChatGPT and Perplexity will use the GBP rating (the more public, harder-to-game source). Your schema becomes irrelevant or actually damages trust.
We've seen this with five Brooklyn restaurants. One in Williamsburg had 4.6 stars on their site but only 4.1 on GBP. When we added ReviewItem schema, Perplexity pulled the GBP rating instead, creating a visual inconsistency. We removed the schema and focused on getting more GBP reviews instead. That was the real work.
Your schema is only credible if it matches or is lower than your public ratings. If your site markup shows a higher rating than your GBP, remove it immediately.
What to Do Tomorrow
Check your Google Business Profile rating first. That's your baseline.
If it's 4.3 stars or higher and you have 25+ reviews, add AggregateRating schema to your homepage and service pages. Make sure the markup reflects your actual GBP rating—not higher, not lower.
If your rating is between 4.0 and 4.3 stars, decide based on review velocity. If you're getting three new reviews per week, markup is fine (you'll hit 4.3+ in 90 days). If you're stagnant, skip the schema and focus on generating reviews first.
If your rating is below 4.0 stars, do not use schema. Remove it if it's already there. Fix the underlying service or product issues instead. Schema won't save you. It'll just make the AI engines deprioritize you faster.
Brooklyn BJJ Lessons did this correctly. They spent six months generating reviews before they launched their site with any markup. By the time we went live, they had 4.7 stars across 38 reviews. Review schema was deployed on day one and contributed to their ChatGPT citation within 41 days.
The businesses that fail at review schema are the ones that add the markup before they've earned the rating. Don't be one of them.
Review schema is a leverage tool. It works on top of actual reputation. It doesn't build reputation. If you're sitting on a 3.8-star average, you're not a schema problem—you're a service problem.
We run a free audit that checks your review sources, rating velocity, and whether your schema is currently helping or hurting. Take 15 minutes to know for sure. Book one at https://signalai.agency/#audit.
For Brooklyn independent businesses, the pattern is clear: earn the rating, then deploy the schema. In that order. The businesses getting cited by AI are the ones who understand that distinction.