Method

Why we publish a band, not a number

Revenue cannot be known from the outside. Two independent routes and their assumptions are more useful than one confident figure.

Aug 2026 · Editorial

What we will and will not state
Measures in the catalogue260counted
Written as a band or a word6not a single figure
No winner by design182own scale, price, positioning

Of 260 measures we hold, 6 are written as a band or a word rather than a number.

Details Part 1

Two routes that share no source

One route takes their estimated visits and applies an ordinary conversion rate. The other takes the number of new reviews they collect in a week and applies an ordinary review rate. Neither route uses anything the other one uses.

Where the two land is the band. If the bottom were the truth they would convert at one rate; if the top were, a different share of their buyers would leave a review. Both are ordinary, which is exactly why it is a band.

See a size band on a real brand

Benchmark Part 2

A confident number is the easy lie

It would be trivial to publish a single figure. It would look better in a headline, it would be easier to compare, and it would be wrong in a way nobody could check.

The band is harder to read and harder to sell. It is also the only version that survives someone asking how we got there.

See how a measure is marked

Changes Part 3

The same rule on the smaller numbers

Visit estimates are modelled, not measured, and they say so on the page. 182 of our measures have no winner at all, because a price is positioning and a search index is scaled against a brand's own past.

Marking those as a win for somebody would make the scoreboard look decisive. It would also make it useless.

See what has no winner

Pick one competitor. See what we find.

Your first benchmark is free. One domain, measured against you, no card and no call.

All notes