How Claude decides which brands to recommend, and how to track it.
Claude recommends brands from what its model learned in training, and from the web when it searches. Brands that are well described on pages it can read are the ones it names. Proofsource tracks the buyer questions where Claude names you, and who it names instead.
How does Claude build an answer?
- By default Claude answers from its training. A brand with little written about it on pages Claude could learn from is easy for it to miss or mistake.
- When Claude searches the web, Anthropic's search crawler and user fetcher decide whether your pages can be read.
- Anthropic runs separate bots for training, for search and for fetching a page a user asked about, and each can be allowed or blocked on its own.
Which crawlers decide whether Claude can read your site?
As documented by the vendor. Check your CDN as well as robots.txt: an edge bot-blocking setting can turn a crawler away even when robots.txt allows it.
| User agent | What it does |
|---|---|
| ClaudeBot | Collects web content that may be used to train Anthropic's models. |
| Claude-SearchBot | Reads pages to improve the relevance of Claude's search results. |
| Claude-User | Fetches a page when a Claude user's question calls for it. |
What did Claude say about Tesla?
- Across 25 buyer questions in Tesla's markets, Claude named Tesla in 11 answers. The competitor it named most was SunPower, in 4.
- Its most-cited site was en.wikipedia.org, 16 times. It cited tesla.com once in 351 citations.
- Asked “What is Tesla?” and “Is Tesla legit?”, it answered from what the model already knew, with no sources cited.
Tesla, public AI answers, Oct 4, 2026: Proofsource's first scan of Tesla, 25 buyer questions and 5 brand questions, one answer each per engine. A small sample from one brand, shown as an example, not a benchmark.
How do you get named more often by Claude?
- Allow Claude-SearchBot and Claude-User, at your CDN as well as in robots.txt. Allow ClaudeBot if you want future models to learn about you.
- Publish a plain, factual page about what you are, who it is for and what it costs.
- Earn mentions on the third-party pages your category's answers draw on.
How does Proofsource track Claude?
Proofsource asks Claude the buyer questions you approve, alongside ChatGPT, Perplexity, Google AI Overviews, with the same wording on each. It records whether you are named, in what position, next to whom, and which pages are cited.
Paid plans ask every question daily or weekly; the free trial runs 2 scans, today and tomorrow. Each percentage carries a 95% Wilson interval, so a change is flagged only once it is bigger than the noise. See the methodology.
Where a Claude number can mislead: Most of what Claude knows changes only when Anthropic releases a new model, so improvements can take a model release to show. Track the trend, not a single week.
What people ask about Claude and brands
How do I get Claude to recommend my brand?
Give it something to learn from and something to read: clear pages about what you are, mentions on the third-party pages your category relies on, and open access for Claude-SearchBot and Claude-User.
What is the difference between ClaudeBot and Claude-SearchBot?
Anthropic documents ClaudeBot as collecting content that may be used for training, and Claude-SearchBot as reading pages to improve search results. Blocking one does not block the other.
Why does Claude confuse my brand with another company?
When a name is close to a better-known one, the model falls back on what it knows best. A single plain sentence saying what you are, repeated on your site and profiles, gives it something to tell you apart by.
- Anthropic: does Anthropic crawl data from the web support.claude.com/en/articles/8896518-does-anthropic-crawl-data-from-the-web-and-how-can-site-owners-block-the-crawler
Sources read on .
Find out if Claude says your name.
See which buyer questions name your brand, who gets named instead, and what to fix. 200 answers. 25 Prompts/Question. Top AI engines. 2 days. Free. No card.