Monitor How Claude Describes Your Brand
Claude is often asked to research and compare: "summarise what [company] does", "pros and cons of [product]", "[company] vs [competitor] for a small team". Its answers tend to be detailed — which leaves more room for a wrong fact, an outdated limitation or an unflattering comparison. Brand monitoring keeps those answers in view.
Research-style questions about your brand
Example branded questions people ask Claude
- what does [company] do and who is it for?
- pros and cons of [product]
- [company] vs [competitor] for a team of 10
- is [product] GDPR compliant?
- what are the limitations of [product]?
Questions about limitations and compliance are worth special attention: the answer often decides whether a buyer keeps you on the list.
What Claude bases its answer on
Without a web search, Claude answers from what its model learned — which can predate your latest features, prices or policies. With a search, it reads current pages and cites them. The same question can get either kind of answer, so it matters to know which one you're looking at.
What to check
- Strengths and weaknesses — are the limitations real and current, or long fixed?
- Facts — pricing, features, integrations, locations, compliance claims.
- Comparisons — whether competitors are described on the same terms as you.
- Sources — when Claude searched, which pages it cited and how old they are.
When Claude's knowledge is out of date
You can't update a model's training data, but you can make current facts easy to find when Claude searches: a clear, dated page for pricing, features and policies, accurate third-party profiles, and crawler access for Claude-SearchBot. How Claude chooses what to mention is covered in how to get mentioned by Claude.
Brand citations in Claude: what gets recorded
When Claude searches the web for a question about your brand, it attaches citations to the parts of its answer that came from a page. Those brand citations in Claude show which pages it trusted to describe you — your own, or someone else's.
- Answer text — the full answer, so you can read what was said about you.
- Web search used — whether Claude searched at all; answers from model knowledge have no citations.
- Searches run — the queries Claude sent to its search tool.
- Cited pages — the pages attached to the answer, flagged as your site, a competitor's or a third party.
- Searched, not cited — results Claude's search returned but it didn't use, kept separate from citations.
An old review or an outdated comparison among the cited pages is usually the reason a detail is wrong. Making your own pages the ones worth citing is covered in how to get mentioned by Claude.
Monitoring Claude with YourAIRank
Add the branded questions you want to monitor as your own questions. Each one is asked through Anthropic's API with Claude's web search tool; the answer is stored in full with the model that produced it, whether a search was used, and the sources cited.
Pro re-checks up to 10 questions every 7 days on ChatGPT, Perplexity, Claude, Google AI Mode and Google AI Overviews. Pro isn't available to buy yet — the free check works today.
Branded questions measure what's said about you, not whether you're found. For discovery, track unbranded questions with the Claude rank tracker.
Frequently asked questions
How do I track Claude citations for my brand?
+
Ask Claude the questions people ask about your brand, under the same conditions each time, and record which pages it cites when it searches. YourAIRank does this through Anthropic's API with web search and keeps the cited pages separate from results it didn't cite. Unbranded discovery questions belong in the Claude rank tracker.
Is this what people see in claude.ai?
+
Not exactly. The Claude apps add their own instructions, memory and settings. We use Anthropic's API with web search under fixed conditions, and every report says so.
Can products built on Claude say different things about us?
+
Yes. Each product adds its own instructions and data. Your own pages and independent sources are what they all have in common.
Should I monitor ChatGPT and Google too?
+
Buyers use several assistants. See ChatGPT brand monitoring and Google AI brand monitoring.
Start with a free check, then track it over time
The free check asks ChatGPT, Perplexity, Claude, Google AI Mode and Google AI Overviews 3 real customer questions about your business. Pro monitors up to 10 questions on ChatGPT, Perplexity, Claude, Google AI Mode and Google AI Overviews every 7 days — Pro is launching soon.