local.intel runs a loop on a weekly cron. Each pass reads every review your block collected since the last one, compares the themes and velocity to the prior run, and writes you the page that matters.
Tell us your category and city, or paste your Google Maps URL. We lock it as the subject.
We pick the top ~20 competitors near you by review count × rating. You can pin or exclude after the first brief.
Yours and theirs. Recent first. A capped sample so the brief reflects now, not 2019.
Themes you're praised and complained about, the gaps where rivals beat you, who's moving on velocity.
Ranked by likely impact and grounded in the gaps. Email + dashboard. Forever, until you cancel.
A single brief is interesting. The second one — what changed from last week — is where the value compounds. We store every run so the delta is real.
A complaint cluster from 2019 isn't a useful action. We cap and bias toward what customers said this month.
You're not running a data audit. You're running a business. We extract the page; the underlying review data stays available on request.
Not 'improve customer service.' Specific: 'three of your five nearest rivals are praised for evening hours; you aren't mentioned for them once — post 6pm slots Tue/Thu for thirty days.'
Two orthodontists in Austin share most of their competitor set. We scrape that market once and bill the synthesis twice. That's how the unit economics hold at $39.
You buy a brief, not a credit. The scrape cost is the studio's problem, not yours.