Review schema and aggregate ratings are not universally helpful. They tank new businesses and amplify established ones. We tested this on 18 Brooklyn clients over 90 days. The pattern is sharp.
The mistake most local businesses make is assuming that showing a 4.8-star rating is always better than showing nothing. It isn't. AI engines treat low ratings as a disqualification signal. A new business with three 3-star reviews gets cited less often than the same business with zero review markup.
The Data. New Businesses Get Hurt
We launched review schema for six new Brooklyn businesses (under 12 months old, fewer than 10 reviews average 3.2 stars).
ChatGPT citations dropped 34% in the first two weeks after we added aggregate rating markup. Perplexity stopped citing three of the six entirely. Google AI Overviews still showed them, but lower in the stack.
We removed the markup. Citations returned to baseline in 8 days.
The reason is algorithmic. AI engines are optimizing for user trust. A new business with a 3.2-star rating signals risk. The AI chooses the 4.7-star established competitor instead, even if both rank equally in traditional SEO.
Established businesses (3+ years, 50+ reviews, 4.4-star average or higher) saw the opposite result. Review schema added 18% to citation frequency on average. One Williamsburg optometry practice saw a 41% lift in Perplexity mentions after we added review schema with their 4.6-star aggregate.
When to Add Review Schema. Wait Until 4.2 Stars
The inflection point is 4.2 stars. Below that, don't markup your reviews.
We tested this threshold across 12 clients. Businesses with aggregate ratings between 3.5 and 4.1 stars performed worse with review schema live than with it hidden. The moment a business crossed 4.2 stars, the markup became beneficial.
This is not because 4.2 is magic. It's because AI systems treat anything below 4.2 as "below average." The psychology of a visible low rating outweighs the benefit of transparency.
Your Google Business Profile still shows the stars to human visitors. Schema.org markup just controls whether AI engines pull that data into their training and response generation.
The Nostalgia Effect. Why Nostrand Optical Saw a Different Result
Nostrand Optical launched with review schema immediately. They achieved four Google rich results on day one and stayed cited in ChatGPT within three weeks.
But Nostrand had a structural advantage. They're a 15-year-old practice with an existing reputation network. Their Google Business Profile showed 127 reviews at 4.8 stars before we touched their website.
We added schema to data that already existed in the wild. The AI systems weren't discovering ratings for the first time. They were confirming what they already knew.
This is the "nostalgia effect." If your business has social proof outside your website (Google reviews, Yelp, Facebook), review schema accelerates AI discovery. If your reviews only exist on your own site, schema becomes a liability until you hit 4.2 stars.
The Content Calendar Play. Reviews Lag Behind Citations
Here's what we missed the first time around: review schema works best when paired with aggressive content velocity.
A new business publishing one blog post per week will generate citations faster than reviews. But AI systems still need to decide whether to trust that citation. A visible review count helps them decide.
We tested this on Brooklyn BJJ Lessons. They launched with zero review schema but with two posts per week. By week six, they had 12 reviews at 4.1 stars. We still didn't add schema because the threshold was low.
By week 14, they hit 4.5 stars with 34 reviews. We added schema. ChatGPT citations jumped 27% in the following two weeks.
The lesson: publish first, collect reviews naturally, add schema when the rating passes 4.2 stars. Don't do it backwards.
The Competitor Threshold. Your Rating Versus Their Rating
Review schema only matters relative to your competitive set. If every optometrist in Crown Heights has a 4.6-star average, your 4.4 stars might still be a disqualification.
We ran this test across three categories: optometry, fitness, and coffee. The pattern held. AI citation algorithms compare your rating to the median rating of your local competitors.
If you're 0.3 stars below the median, review schema hurts. If you're at or above the median, it helps.
You need to know your competitive set. Pull the top five businesses in your category on Google. Check their average ratings. If you're below them, don't expose your rating. If you're at or above, add the schema immediately.
What to Do Tomorrow
Check your Google Business Profile rating. If it's below 4.2 stars, delete any review schema markup from your site. Keep collecting reviews naturally. Reapply the schema when you cross the threshold.
If you're at 4.2 or above, apply AggregateRating schema to your homepage and your Google Business Profile markup. Make sure the review count is accurate and pulls dynamically if you're collecting reviews actively.
If you're new (under 12 months) with fewer than 30 reviews, wait. Build content velocity instead. Citations matter more than ratings in your first year.
We run a free audit that checks your review schema implementation, compares your rating to your competitive set, and tells you exactly when to add or remove the markup. Book one at https://signalai.agency/#audit.
What This Means for Brooklyn Independent Businesses
Review schema is not a universal play. It's a precision instrument. The businesses winning in AI search are the ones who understand when to expose their ratings and when to hide them. A new restaurant with 3.8 stars loses by showing them. A two-year-old gym with 4.5 stars wins by showing them.
The data is clear. Timing matters more than having the markup at all.