/ what the ai gets wrong

When the AI gets you wrong

The whole market measures whether the AI names you. We also measure whether it tells the truth about you.

How it works

You give us your hard facts: founding year, location, what you do, and explicitly what you do not do. We ask for that last part on purpose, because it is exactly where the expensive misunderstandings start.

ECHO compares every measured answer against them and shows you the sentence verbatim when ChatGPT, Claude, Gemini or Perplexity claims otherwise. Tick each finding off, or wave it away as correct after all, and we stop raising it.

EvidenceNo quote, no finding. ECHO reports only what appears word for word in a stored answer. If a quote is not in the answer character for character, we discard the entire evaluation of that answer. An invented accusation against an AI answer costs more trust than the feature is worth. The check runs over the answers of the past 14 days.

Why this is a different conversation

For a managing director, «ChatGPT says you are based in Zurich and do not do maintenance» is a different conversation from «your mention rate is 12 per cent».

One is a metric. The other is a sentence a customer reads right before she calls you, or does not call you.

EvidenceWith no facts on file, nothing happens, and it happens quietly. An empty library is not a bad result, it is no result at all.

Limits that belong here

  • The fact watch belongs to the pilot and the larger plans, not to the entry tier.
  • It covers the four chat services and the Google AI box.
  • With no facts on file it reports nothing rather than inventing something.

Which plan includes this? →

Want to know what the AI gets wrong about you?

Give us five hard facts about your company and we show you where the AI says something else.

Request a free first measurement → ← Back to site
The AI box above the Google resultsWhen the AI gets you wrongGerman, French, Italian, EnglishHow honest the numbers are