Does KI Score show real user queries to AI systems?
Where the answers in KI Score come from and why you see reproducible measurement queries instead of real users’ conversations.
Last updated 23 days ago
This article explains where the questions and answers in KI Score come from. Afterward, you’ll know what the data tells you, what it doesn’t, and which analyses are available to you instead.
Short answer
No. KI Score doesn’t show conversations that real people have with ChatGPT, Perplexity, or other AI systems. KI Score doesn’t import user logs from these services. There’s also no metric for how often a question is asked, and no search volumes.
Everything KI Score analyzes is based on measurement queries that KI Score itself sends to the AI providers.
Where the answers come from
Every answer comes from KI Score’s own queries. The AI providers are queried through their APIs with web search turned on, not through their chat apps. That’s why the ChatGPT, Google AI Overview, and Copilot profiles are approximations of the respective product interface. The dashboard glossary explains this under Näherung (approximation).
There are two kinds of measurement queries:
- Monitoring: your own prompts, which you create under Prompts. In every run, KI Score sends your active prompts to the selected AI providers.
- KI Score measurement: a market’s question set, which measures every brand in that market the same way. Only the KI Score team sees the question set internally. Customers can’t view it, including workspace admins.
Why measurement queries instead of user logs
KI Score deliberately works with its own measurement queries because they’re reproducible: the same question is asked again and again under the same, traceable conditions. That lets you compare runs over time, and changes reflect the AI’s answers rather than shifting questions.
There’s one limitation: an API measurement is an approximation of the real apps. What a person sees in a provider’s app may differ.
Note: How representative your monitoring is depends on your prompts. Phrase them the way your customers actually ask.
What you can analyze instead
- Responses shows the measured prompt–answer pairs from your monitoring, including the cited sources.
- Prompts shows, among other things, the share of answers per prompt that mention your brand or your competitors, and how often your domain is cited as a source.
- Sources shows which domains and URLs the AI answers cite.
- Fanouts shows the searches the AI models send to their web search while answering your prompts, with how often each occurs. These are searches from KI Score’s measurement, not search volumes and not queries from real users. Search texts are only delivered by ChatGPT, Copilot, Grok, Claude, and DeepSeek.
- The KI Score places your visibility in your market in comparison with other brands.
- Traffic Overview shows real visits to your website from AI services such as ChatGPT or Perplexity under KI-Verweise (AI referrals), provided you’ve set up the integration. It doesn’t show which question preceded a visit.
What teams say about KI Score
“We finally see week by week where our brand shows up in ChatGPT and Perplexity.”
“The score finally gave us one number the whole team actually understands.”
“Instead of guessing which prompts matter, we now have real data.”
“Within the first week we already saw which competitor was beating us in AI answers.”
“Setup took one afternoon, not weeks of onboarding.”
“Finally a tool that takes AI Overviews as seriously as classic SEO.”