What we can tell you about the claims on your website
VERIS records what a page claims and what evidence was published with it, then watches whether that evidence is still there. Free scan, no account.
Most marketing pages make claims. Some of those claims cite evidence. Almost nobody keeps a record of which is which — and by the time someone asks, the person who ran the numbers has left and the study is a link in a deck.
What it actually does
VERIS reads a public page and records each claim as published, verbatim, alongside whatever evidence appears beside it. It dates the reading. Then it answers a narrower question than most people expect: not whether a claim is true, but whether evidence was published with it, and whether the source cited can carry what the sentence asserts.
Those are different questions and collapsing them is how a page ends up looking substantiated while being indefensible. A 99% accuracy figure linked to a blog post has evidence present. Whether a blog post can carry an accuracy figure is a separate judgement — and one this software deliberately does not make.
The distinctions it is built on
Evidence present is not evidence sufficient. A claim with a citation attached is treated as substantiated, and a claim without one is treated as unsubstantiated. Both readings skip the question that matters.
Breadth claimed is not breadth tested. A capability demonstrated in one context is described in language that covers every context, and the gap between the two is invisible in the sentence itself.
Superseding beats age. Old evidence is treated as weak evidence, and recent evidence as strong. Age is the wrong axis.
Two disclosures, not one. Commercial relationship and machine involvement are treated as a single disclosure obligation, so satisfying one is taken to satisfy the other.
The full method is published free at kema.knightbyrd.com/veris/standard — version 1.0 of The KnightByrd Claim-Evidence Method. Anyone may apply it, including people who never buy anything from us, and including competitors. A method only our customers can use is not a method.
The part that took the longest to build
Everything in this category checks a page at a moment. The failure that actually bites is slower: your claim does not change, and the evidence underneath it does. A study moves. A benchmark page is rewritten. A cited link fails to load. Your wording is word for word what it always was, so no page monitor fires, and the claim is now pointing at nothing.
Seeing that requires two dated readings and a habit of re-reading the sources, not only the page. That is what the monitoring products below do.
What you can buy
VERIS Ledger — $1,397. The claims found on the pages you submit, and what evidence was published alongside each one.
VERIS Blazon — $2,297. What a competitor claims, and what they published to support it.
VERIS Blazon Watch — $597/mo. A named competitor's published claims, read on a schedule, including when the evidence behind them moves.
VERIS Sentry — $537/mo. Your published claims, watched — so you know when the thing they measured has moved on.
VERIS Dossier — $2,297. Your substantiation file, assembled and dated before anyone asks for it.
VERIS Scan is free: A dated count of the claims on your site and how many cite evidence that carries them.
What it will not do
It does not decide whether a claim is adequately supported. It does not give legal advice or reach a legal conclusion. It reaches no verdict about anyone, you or a competitor, and it does not characterise why a page says what it says. It records what was published and what changed, with dates, and the judgement stays with the person who has to make it.
It reads public web pages. Content behind a login, inside a PDF, in a deck or in a video is outside it, and for some businesses that is where most of the exposure lives. We would rather say so here than have you find out in the first hour.
A page we cannot read is reported as unread and excluded from every count, never quietly folded into a clean result.
We ran it on ourselves first
knightbyrd.com returned 2 claims of our own that would face the same question, and we published them rather than fixing the page quietly. On 2026-08-19, 16 companies were read: 368 claims inventoried, 133 that would face a substantiation question, and 2 that came back with nothing high. A tool that flags everything is worthless and a tool that flags nothing is broken.
We also publish the sentences our own software refused to let us write about ourselves. They are on the method page, with what we wanted to say and what we published instead.
How to try it
Scan one page free at nexus.knightbyrd.com/veris. No account, no install, no code on your site — a URL is the whole setup. You get counts. If you want the record behind the counts, that is what the products above are.
This records what was claimed and what evidence was published alongside it. It is not legal advice and it does not determine whether any claim is adequate.