Search Became One Answer. Most Brands Do Not Know If They Are In It.

Search used to hand you ten blue links and let you judge the sources yourself. Increasingly, it hands you one answer instead, written by a model that read a handful of pages and decided what to repeat. If your brand is not in that answer, most of the people who asked the question never see you, and you have no way of knowing it happened.

Plenty of tools will sell you a number that claims to measure this. Very few will tell you which prompts they ran, how many times, against which engines, or how much that number moves if you ask again tomorrow. A score you cannot audit is not a measurement. It is a claim, and I did not want to be in the business of selling claims.

So the methodology page on this site publishes the whole method, not just the headline: the exact prompt set, the engines queried, the number of runs per prompt, and a 95% bootstrap confidence interval on every score. We report Selection, whether your brand is chosen as a candidate at all, separately from Absorption, how much of the actual answer it shaped, because collapsing the two into one number hides which problem you actually have.

Nobody, including us, controls what a model decides to cite. So we do not promise a ranking, a citation, or a placement, and we say so plainly instead of rounding up. What a report commits to instead is the outcome: specific, actionable gaps against the competitors you name, delivered the way it was described before you paid.

I did not think a subscription you cannot see the value of yet was a fair way to start, so GeoHero launched selling one-time reports, Compete and Deep Research, instead of a recurring charge. And we run the exact same methodology against our own category and publish the result in the open on our benchmark page, because a method that cannot survive being pointed at its own maker is not one worth selling.