What star ratings and review counts don’t show
A 4.3 rating from 180 reviews is a number — it says nothing about why customers were happy or unhappy. A competitor with fewer reviews can still be winning if their handful of reviews consistently praise one specific thing, while a business’s own 180 quietly bury a recurring complaint the average score never surfaces. What people actually write in a review stays invisible in a plain star rating.
What the click sets in motion
AI review analysis pulls a linked business’s recent Google reviews (up to 50) and has an AI extract three things: a sentiment breakdown across positive, neutral and negative, a list of recurring themes with their tone — “wait times” running negative, “staff friendliness” running consistently positive, say — and a few representative quotes per theme instead of a bare number. The result appears directly on the project page, timestamped with when it was last run.
Deliberately opt-in, deliberately throttled
The analysis doesn’t run alongside every scan automatically — it only fires from its own button, so anyone who doesn’t need it never sees it in the way. Server-side, it’s also capped at one refresh per project per week, regardless of how many times the button gets clicked. That keeps the cost envelope tight: neither the Google review data nor the AI classification draws on scan credits, both run on a separate daily budget of their own.
What it’s actually useful for
For an agency that has to summarize a client’s reputation regularly, the theme-and-sentiment picture replaces manually reading through dozens of reviews. For a single-location business, it surfaces which specific issue has been coming up over recent weeks — a lead that usually points to an action more directly than a single star score ever could.