How PromptSignal measures your AI visibility.

PromptSignal runs the prompts your buyers ask in the engines they use, keeps every answer and citation, and reports how often each engine names you.

Illustration with sample data: seven AI engines answer a buyer's question, “ What's the best project management software for a creative agency?”, for a fictional brand, Tallyfield. PromptSignal reads each answer: Tallyfield is mentioned by 5 of 7 engines, at an average position of 1.8.

You choose the prompts, markets and competitors.

Every number starts from your setup: your brand and its aliases, your prompts and the competitors you compare against. PromptSignal suggests prompts and competitors, and you decide what to keep.

Your prompts

Write prompts in your buyers' words and group them by product, audience or market with segments and tags.

Suggested prompts

You get 20 suggested prompts at a time, written for your company, market and language.

Countries and languages

Each prompt has one country and one language, chosen from 72 countries and 32 languages.

Competitors

Share of voice and gap to the leader compare you with the competitors you track, so your list shapes them.

Answers come from the interfaces your buyers use.

PromptSignal tracks ChatGPT, Google AI Overviews, Google AI Mode, Gemini, Perplexity, Microsoft Copilot and Claude. It captures six of them from those interfaces and measures Claude through its API with web search.

How often prompts runEach prompt runs in each engine you track every 48 hours by default.

Every answer keptEach answer is kept in full, with its citations, your position when you are named, and the prompt's run history.

CitationsSources shows the domains and pages the engines cite, and the pages that name your competitors but not you.

Each engine on its ownThe Models scorecard shows each engine separately, including where you are strongest and weakest.

Illustration: the seven AI engines PromptSignal tracks, all switched on, six captured from their live interfaces and Claude through its API with web search, beside prompts set to run in 72 countries and 32 languages.

Visibility is the share of runs that name you.

A run is one prompt asked once in one engine. Metrics cover the runs in the period you pick, in 7, 14, 30 and 90 day views.

MentionsHow many times answers name your brand in the period.

Average rank and first mentionYour average position in answers that name you, where lower is better, and how often those answers name you first.

Share of voice and gap to the leaderYour share of mentions among tracked brands, and how far you trail the leader, as a percentage of the leader's visibility.

SentimentThe average 0 to 100 score for how each answer that names you describes you, where higher is more positive.

Illustration with sample data: the PromptSignal dashboard for a fictional brand, Tallyfield. Its visibility rose to 48 percent over 30 days and passed its closest rival, Huddlewise; the panels rank five competitors and show visibility in each of seven AI engines.

The numbers are not search volume or a guarantee.

Visibility measures how often engines name you for the prompts you track. It does not measure how many people ask them, and no tool can guarantee what an engine will say.

Questions. Answered.

How many runs stand behind a visibility number?

Each prompt runs in each engine you track every 48 hours by default, and visibility is the share of those runs that name you. A longer period rests on more runs, so read a single prompt over 30 or 90 days rather than 7. Open any prompt to read every answer behind its number.

Why does PromptSignal run the same prompt more than once?

Because the same prompt can get a different answer each time. Prompts run every 48 hours by default, and visibility is the share of those runs that name you, so one unusual answer does not decide the result.

Is AI visibility the same as search volume?

No. Visibility measures how often engines name your brand when they answer the prompts you track. It does not measure how many people ask those prompts.

Can PromptSignal guarantee that AI engines will recommend us?

No. Each engine decides its own answers, and no tool can guarantee them. PromptSignal measures where you stand and shows whether your changes moved the numbers.

Can we see the answers behind a number?

Yes. Every answer is kept in full for each prompt, with its citations, your position and the run history, so you can read exactly what each engine said.

How does PromptSignal capture answers from each engine?

PromptSignal captures answers from ChatGPT, Google AI Overviews, Google AI Mode, Gemini, Perplexity and Microsoft Copilot through the same interfaces your buyers use, and measures Claude through its API with web search.

See what AI says about your brand.

Start a free trial and track your first prompts today, or book a demo on your own brand.