Five pages, written in the same words the app uses. If something on a screen is unclear and it is not answered here, write to us and we will fix the page.
How to read your Index — what the 0–100 number is, why it always appears next to your category count, what the rank and percentile are counting, and what does and does not move it.
Why a category says Not named — the five standings, and why Not named is evidence rather than missing data.
What each requirement means — every row on the badge screen, including the one that is not a criterion and the one we deliberately do not show yet.
The formal version is the badge & claims policy, which also covers what may and may not be said with a badge.
What each product group means — Ready, Waiting on us, We couldn’t confirm these, and Outside what we measure. Three of the four are ours.
What this product does not do — no outside validation of our method, no advice on how to rank, no per-SKU signal, and no trend until the second round. Timestamping is partial: the 2026·09·05 round is independently timestamped as existing on or before 2026·09·19, which does not prove the round is from the 5th. And there are two emails, both transactional — one when a category you have products in starts being measured, one seven days before a first charge.
Methodology — engines, questions, chance levels, intervals and the full criteria table with thresholds.
The AIProof Index — the published formula, why it is logarithmic and why the anchor is fixed.
Pricing — what the free tier includes and what $39 a month buys.