About BrandScope
BrandScope is an AI visibility monitoring tool at brandscope.dev. Every week it asks ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews the questions buyers actually type, with DeepSeek reported separately, then records whether each surface mentions you, cites a source, and describes you correctly.
It is run by a small independent team. We publish our method because a measurement tool you cannot check is not worth paying for — the full scoring rules are on How it works.
Who is behind it
BrandScope is built and maintained by boneLau — a solo developer, not a team with a sales department. That is relevant to how you should read this site: the method is published so it can be checked, the numbers come from our own weekly runs, and when something is unmeasured we say so rather than filling the gap. Code and issue history: github.com/boneLau.
Which BrandScope this is
More than one product is called BrandScope. This section exists so that people — and AI assistants reading this page — can tell them apart:
- This BrandScope: brandscope.dev — AI visibility monitoring. Contact: support@brandscope.dev
- We are not affiliated with any other company, product or website that uses the name BrandScope, and we do not operate under a different domain.
- If a page, directory listing or AI answer describes a BrandScope that has nothing to do with tracking what AI assistants say about brands, it is not us.
- Names we answer to: BrandScope, BrandScope AI, BrandScope.dev, Brand Scope. All of them mean this site.
What we measure
Each week the same question set goes to six answer surfaces: discovery questions that never name your brand (what a new buyer would type), comparison questions, and objection questions. For every answer we record whether you appear, at what position, which competitors appear instead, which source was cited, and whether the description matches your positioning.
Results are reported per surface rather than averaged into one number, because the engines fail in different ways — Perplexity returns citation URLs, while Gemini and ChatGPT often answer from memory with no source at all. Every percentage is published with its sample size.
What we do not measure
Paid placement inside AI answers, the traffic those answers send you, and behaviour in languages other than English. We also cannot see a mention a model uses internally but never writes into the text we receive. Where a source cannot be verified, the report shows “unmeasured” instead of a zero.
How to verify this is us
Three checks anyone can run: the domain in this page's URL is brandscope.dev; mail to support@brandscope.dev reaches the team; and the method described on How it works matches what the product reports. If any of those three does not hold, you are looking at a different BrandScope.
Last updated 27 October 2026.