Skip to content
Open dashboard

Help centerGet started

How GEO Radar works

Prompts, AI answers, scans, mentions and sources: the ideas every page is built on.

More and more people ask ChatGPT, Gemini or Google’s AI answers which product to choose. GEO Radar asks those AI assistants your buyers’ questions every day. It records which brands they name, which sources they rely on, and what you can do to be recommended more often.

The daily loop

Monitor → Sources → Actions → Impact

  1. 1MonitorAI answers your prompts every night
  2. 2SourcesSee which sites shape the answers
  3. 3ActionsFix your site, pages and placements
  4. 4ImpactCompare before and after

Each step has its own pages in the sidebar. Most teams check Answer audit and Insights weekly, work through Recommendations, and look at Impact once actions are done.

Key ideas

Project
One brand and its website. A workspace can hold several projects, for example one per brand or market.
Prompt
A question buyers ask AI, such as “What is the best AI humanizer for students?”. You choose which prompts to track. Prompts are grouped into topics and can carry tags.
AI model
An AI assistant GEO Radar asks: ChatGPT, Perplexity, Google AI Overview, Google AI Mode, Gemini or Copilot. Each project uses three.
Scan
The nightly run that asks every enabled prompt on all three models. 50 prompts give 150 answers a night.
Answer
One model’s reply to one prompt. Everything in GEO Radar is counted from answers.
Mention
The answer names a tracked brand. Each mention is read for its role (recommended, compared, mentioned or warned against), its rank in the answer, and its sentiment.
Source
A web page the AI used. Cited sources are shown to the reader; retrieved ones were only read.
Competitor gap
An answer that cites a source and names a competitor, but not you. These are the answers you can most clearly win.
Recommendation
An action to improve your visibility, either suggested by GEO Radar from the data or saved by your team.

The four numbers to know

Visibility
In one sentence: How often AI names you, out of all answers.
Recommended
In one sentence: When AI recommends a product, how often it recommends yours.
Position
In one sentence: How early in the answer you are named (#1 is first).
Sentiment
In one sentence: How positively AI talks about you, from 0 to 100.

Every formula, minimum and marker is explained in Metrics explained.

What you can rely on

  • Real answers only. Every number links back to the answers it comes from. Click through to read them.
  • A dash is not a zero. When there are too few answers for a fair number, you see “—”.
  • Fair comparisons. Changes always compare with the previous period of the same length, using the same models.

Related guides