How we calculate your score
Nævnt measures your AI Share-of-Voice — how often and how strongly the AI engines mention you when customers ask real buying questions. Here's the method, no marketing.
1. Real buyer prompts
For each industry we have a fixed set of Danish and English questions customers actually ask — e.g. “best accountant in Aarhus?” or “which accounting software for small businesses?”. We do not put your own name in the questions; the score is about whether you show up when the customer doesn't know you yet.
2. Four AI engines
Every question is put to ChatGPT (OpenAI), Gemini (Google), Perplexity and Google AI Overviews. We store the exact answer, so you can always read precisely what the engine said.
3. From answer to score
For each answer we check: are you mentioned? Are you recommended, not just referenced? Are you among the first? That scores each answer, summed into a 0–100 score per engine and a weighted average across the engines.
The score maps to a level: Invisible · Rarely mentioned · Visible · Strong · Dominant.
4. Honest about uncertainty
AI answers are partly random — the same question can give slightly different answers. So we smooth your score over time and only show changes that are real, not noise. Your score is based on your fixed prompt set, so the trend over time is comparable.
See the method applied to real companies
We ran exactly this method on Denmark's largest accounting firms — none of them are customers, none paid, and none had their name written into the questions. Under each name is the verbatim quote from the engine, so you can check the claim instead of taking our word for it.