Claude and AI assistants: being described accurately.
Assistants that answer from what they have read reward clarity and consistency. Being left out is one risk. Being described wrongly and never knowing is the other.
Check your visibility freeSources
Key takeaways
- Does Claude browse the web? It can, depending on how it is being used. Either way, what it says about a business comes from what is publicly written about that business.
- What is the real risk? Being described inaccurately. An old address, a service you dropped or a competitor's claim can be repeated confidently to a customer who never checks.
- How do you fix that? One authoritative entity page, consistent facts everywhere, Organization and Person schema, and sameAs links to the profiles that corroborate them.
How does Claude choose who to cite?
From what has been written publicly about you, on your site and everywhere else. It rewards consistency of name, address, services and claims, is sensitive to contradictions with your own directory listings, has no reason to prefer your marketing language, and will state something confidently even when the source is out of date.
- It answers from what has been written publicly about you, across your site and everywhere else.
- It rewards consistency: the same name, address, services and claims wherever they appear.
- It is sensitive to contradictions, and a site that disagrees with its own directory listings is a weaker source.
- It has no incentive to prefer your marketing language over a third party's plainer description.
- It will state something confidently even when the underlying source is out of date, which is the part worth watching.
What does our AI assistant work involve?
One entity page stating what the business is, where it operates, what it sells and who runs it; Organization and Person schema with sameAs links; an identical name, address and phone everywhere; llms.txt and a Master FAQ; and the public sources behind any wrong answer corrected when one turns up.
- Build one entity page that states what the business is, where it operates, what it sells and who runs it.
- Install Organization and Person schema with sameAs links to the profiles that confirm it.
- Make the name, address and phone identical across the site, the Business Profile and the directories.
- Publish llms.txt and a Master FAQ, so there is an obvious, current summary to read.
- When an assistant gives a wrong answer about you, trace it to the public source and correct that source.
What does an authoritative entity page contain?
The business name stated once and exactly, what it does, where it operates, who runs it and how to reach it, with Organization and Person schema that repeats the same facts and sameAs links to the profiles that corroborate them.
- Name, address and phone exactly as they appear on the Google Business Profile, Bing Places and every directory.
- The services actually sold, in the words a customer would use, each linking to its own page.
- The people behind the business, with real credentials and Person schema, so an assistant can tell a named owner from an anonymous site.
- Founding date, hours and service area stated plainly: ours are 2015, Monday to Saturday 9 am to 5 pm, and businesses anywhere in the USA.
How do we keep the public record about you consistent?
By fixing the sources an assistant reads. There is no arguing with the assistant itself. The site, the Business Profile, Bing Places, llms.txt and the Master FAQ are brought into agreement, and when a wrong answer turns up we trace it to the page that caused it and correct that.
- One set of facts, copied exactly to the site, the schema, both business listings and the directories that matter.
- llms.txt and a Master FAQ give any model an obvious, current summary to read first.
- When you tell us an assistant said something wrong, we find the source. It is usually a stale directory entry, an old page of your own or a review site.
- Corrections show up slowly and unevenly, because models retrain and caches refresh on their own schedules. We report what changed and never promise a date.
What do the statistics say about how people use Claude?
Anthropic's own research shows personal use rose to 42% of Claude.ai conversations in early 2026, and explanations are what Claude produces most often. Claude crawls far more pages than it sends visitors back to.
Statistics last updated 10-30-2026. Each figure links to the organization that first published it.
- Personal use rose from 35% to 42% of Claude.ai conversations between the November 2025 and February 2026 samples, according to Anthropic's March 2026 Economic Index report. Source: Anthropic, 2026.
- Explanations were the most common output of Claude conversations at 17%, followed by documents and reports at 15% and guidance at 11%, according to Anthropic's June 2026 Economic Index report. Source: Anthropic, 2026.
- Anthropic had the highest crawl-to-refer ratio among leading AI and search platforms on Cloudflare in 2025, reaching as much as 500,000 crawls per referral. Source: Cloudflare, 2025.
- 42% of US adults say they use AI chatbots to search for information, according to Pew Research Center's February 2026 survey. Source: Pew Research Center, 2026.
Claude questions
Can I control what an AI assistant says about my business?
Not directly, and anyone claiming otherwise is overselling. What you can control is the quality and consistency of the public material it reads, which is what the answer is assembled from.
What if it says something wrong about us?
Tell us what it said, and we trace the claim to its source. Usually it is a stale directory entry, an old page of your own, or a review site. Fixing the source is what eventually fixes the answer.
Do I need a separate page for every assistant?
No. The work that makes you legible to one makes you legible to the others, because they are reading the same public material. The differences are in indexing and freshness, not in what counts as a clear fact.
Is llms.txt required?
No, and we will not pretend it is. It is an emerging convention the engines are not obliged to read. It costs little, cannot hurt, and puts you ahead if adoption grows.
How quickly do corrections show up?
Slowly, and unevenly. Models are retrained and caches refresh on their own schedules. Correcting the source is the only durable fix, and we report what changed. We will not promise a date.
Is the content you publish for us checked by a person?
Yes. We keep a human in the loop on everything: AI writes the first draft, then a senior marketer checks every fact against your own material and signs the page against an eight-point content standard. Engines quote pages that are specific and correct, so the edit is what makes a page worth citing in the first place. Separately, every client site gets a 300+ point audit at onboarding and a 300+ point scan in every monthly report.
Related pages
AI search, by platform
Reference: the schema.org vocabulary.