How brand visibility works

In shortBrand visibility checks whether AI engines mention and cite your brand. You save prompts (questions a buyer might ask), then run them on up to six engines: ChatGPT, Claude, Gemini, Perplexity, Google AI Overview and Google AI Mode. Each run stores whether you were mentioned, your position, the tone, competitors named and the URLs cited. Nothing runs until you run it, and reading stored results is free.

What brand visibility measures

When someone asks ChatGPT "what's the best accounting software for a small agency?", the answer names a few brands and often lists the pages it used. Brand visibility asks the same kind of question for you and records what came back. Over time that tells you how often AI recommends you, where you appear in the list, and who gets recommended instead.

The four building blocks

All results stay in your account. The pages in AI search → Brand visibility are different views of the same stored runs: Visibility overview for the headline numbers, Prompt research to manage and run prompts, Citations for the pages engines cite, Competitors for who else gets named, and Action plan for what to write next.

How a run works

1. You choose prompts and enginesOn Prompt research you click Run next to a prompt and tick the engines. Before anything starts, AI Rank Lab checks that your balance covers every prompt on every engine you picked.
2. Each engine is asked in turnThe question is sent to each engine through its API with web search switched on. The engine is asked to list its sources and report whether your domain was mentioned, where, in what tone and which competitors appeared.
3. Results are saved and chargedEach successful engine answer is saved as one result and charged at the per-engine price. If an engine fails or times out, that result is saved as failed and not charged. The rest of the run carries on.
4. The dashboards updateVisibility overview, Citations and Competitors read the saved results. Opening them, filtering and exporting are free.

What the numbers can and can't tell you

• Answers vary. Ask the same engine the same question twice and the brands it names can change. One run is a sample, not a verdict. Look at several prompts and several runs before drawing conclusions.
• API answers, not screenshots. Results come from each engine's API with web search, so they can differ from what you see typing the same question into the consumer app, which may know your location or history.
• No location for the four chat engines. ChatGPT, Claude, Gemini and Perplexity runs don't pretend to be in a particular country. Google AI Overview and AI Mode runs here use the default market, United States in English. To check a specific country, use the separate AI Overview and AI Mode scans, which take a country and language.
• Latest run counts. The overview uses the most recent result for each prompt on each engine in the date range, so running a prompt five times in a day doesn't inflate your score.
• Keep the wording stable. Old runs keep the wording they were run with, and some views match a prompt's history by its exact text. Editing a prompt's text can split its history, so add a new prompt instead of rewording one you've been tracking.

What it costs