Why I started this
I'm an engineer by training, which mostly means I'm uncomfortable fixing things nobody has measured. You find the actual cause, you check whether it's real or noise, and then you change one thing.
Local businesses are sitting on exactly that kind of data already: years of customers writing down what went wrong. Almost nobody uses it for more than checking the star rating hasn't dropped.
A one-star saying "waited ages" and a three-star saying "took forever to be seated" are the same problem, and no average will show you that. So I read the text, find the cause underneath, and test it against real statistical noise instead of guessing.
Reviews are free data nobody reads properly
Started reading local businesses' reviews in bulk and found the same causes repeating under different words.
Studio 9, a barber in Stanmore
No reviews of its own, so the report benchmarked ten local barbers instead. Read it
Taking on three founding clients
Half price for good, in return for a short write-up of what changed. You choose whether your name is on it. Two places left.