How to Get Your Brand Mentioned in Claude Answers

This check asks Claude questions about your brand and your product category, then looks at whether your brand appears in the answers. If Claude doesn't name you or describes you wrongly, buyers who use it for research won't find you. This guide covers letting Anthropic's crawlers in, publishing pages Claude can quote, building the outside mentions it relies on, and re-testing the result.

By Updated ⏱ 3-4 hours of setup, then 2-6 months for third-party mentions to have an effect🎯 Advanced● Medium impactGEO / AI
Illustration: Brand in Claude answers fix guide
GEO / AI · FIX GUIDE Brand in Claude answers IMPACTEFFORT TIME3-4 hours of setup, then 2-6 months for third-party mentions to have an effect LEVELAdvanced

Why it matters

Claude answers from two sources: what it learned in training, and live web search, which runs when a question needs current information. If your site blocks ClaudeBot or Claude-SearchBot, or if few independent sites mention your brand, Claude either leaves you out of category recommendations or guesses facts about you. Every missed mention sends a recommendation to a competitor, and it happens without a click you could measure.

Wrong descriptions, such as old pricing or a discontinued product, also hurt trust before a prospect ever reaches your site.

✕BeforeAsked "What are good invoicing toolsfor freelancers?", Claude lists fivecompetitors and never names Example✓AfterClaude names Example Co among itsrecommendations, describes itaccurately (pricing, audience, key

What you need

Access to robots.txt and any CDN/WAF bot settings (e.g. Cloudflare)
Ability to edit page templates or add JSON-LD
A free or paid Claude account at claude.ai with web search available
A list of 10-20 real buyer questions in your category
Time for PR/outreach to review sites, directories and industry publications

Step-by-step fix

1

Record a baseline of Claude's current answers

Open claude.ai and run each prompt in a new chat, once with web search off and once with it on. Use three kinds of prompt:

  • Brand prompts: "What is Example Co?"
  • Category prompts: "Best invoicing software for freelancers"
  • Comparison prompts: "Example Co vs [competitor]"

For each answer, note whether you're mentioned, whether the facts are correct, and which sources are cited. This tells you whether the gap is in training data (missing with search off) or in retrieval (missing with search on).

text
Prompt | Search | Mentioned? | Accurate? | Sources cited
What is Example Co? | off | no | - | -
What is Example Co? | on | yes | pricing wrong | example.com/pricing (2023)
Best invoicing tools for freelancers | on | no | - | g2.com, competitor.com
Pro tip: Answers vary from run to run. Run each prompt 3 times and count mention rate (e.g. 1/3) instead of a single yes/no.
2

Allow Anthropic's crawlers in robots.txt

Anthropic uses three user agents:

  • ClaudeBot collects training data.
  • Claude-SearchBot indexes pages for Claude's search results.
  • Claude-User fetches a page when a user asks Claude about it.

Blocking any of them reduces your visibility. Check https://example.com/robots.txt for Disallow rules under these names or under a broad User-agent: * block, and add explicit allow groups.

text
User-agent: ClaudeBot
Allow: /

User-agent: Claude-SearchBot
Allow: /

User-agent: Claude-User
Allow: /

User-agent: *
Disallow: /cart/
Disallow: /account/

Sitemap: https://example.com/sitemap.xml
Pro tip: robots.txt is only half of it. Firewall rules and CDN 'AI bot' toggles can return 403 to these agents even when robots.txt allows them. Check your server logs for ClaudeBot hits that got 403 or 429 responses.
3

Serve your key facts in the initial HTML

Anthropic's crawlers are not guaranteed to execute JavaScript. Pricing, product descriptions and FAQs that load only after client-side rendering may be invisible to them. Fetch your important pages as plain HTML and confirm the facts appear in the response.

If they don't, switch those pages to server-side rendering or static generation.

text
curl -s -A "Claude-SearchBot" https://example.com/pricing | grep -i "per month"
Pro tip: Also check the HTTP status code with `curl -I -A "ClaudeBot" https://example.com/`. Anything other than 200 or 301 points to a bot-blocking rule.
4

Publish an unambiguous entity page with Organization schema

Claude needs to know exactly what Example Co is and must not confuse it with similarly named companies. Make sure /about states clearly:

  • what the company does
  • who it serves
  • where it is based
  • when it was founded
  • its flagship products

Add Organization JSON-LD with sameAs links to your official profiles. The first paragraph should work as a quotable one-sentence definition.

html
<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Example Co",
  "url": "https://example.com",
  "logo": "https://example.com/logo.png",
  "description": "Example Co makes invoicing and expense software for freelancers and agencies with under 50 staff.",
  "foundingDate": "2019",
  "address": {"@type": "PostalAddress", "addressLocality": "Austin", "addressCountry": "US"},
  "sameAs": [
    "https://www.linkedin.com/company/example-co",
    "https://www.crunchbase.com/organization/example-co",
    "https://www.wikidata.org/wiki/Q000000",
    "https://github.com/example-co"
  ]
}
</script>
Pro tip: Use exactly the same company name and one-line description on your site, LinkedIn, Crunchbase and directory listings. Consistent wording across sources helps the model treat them as the same entity.
5

Create pages that answer your category prompts directly

For each category or comparison prompt where you were missing, make sure one page on your site answers it. Examples: /invoicing-software-for-freelancers or /example-co-vs-competitor.

  • Open with a direct one or two sentence answer.
  • Follow with specifics: a comparison table, real prices, limits, integrations, and a 'last updated' date.
  • Use H2s phrased like the questions people ask.

Keep it factual and fair to competitors. Pages that read like ads are less likely to be quoted.

html
<h1>Invoicing software for freelancers</h1>
<p>Example Co is invoicing software for freelancers that sends recurring invoices, tracks expenses and accepts card payments, starting at $9/month.</p>
<p><time datetime="2026-01-15">Updated 15 January 2026</time></p>
<h2>How much does Example Co cost?</h2>
<table>
  <tr><th>Plan</th><th>Price</th><th>Clients</th></tr>
  <tr><td>Solo</td><td>$9/mo</td><td>Unlimited</td></tr>
  <tr><td>Team</td><td>$29/mo</td><td>Unlimited</td></tr>
</table>
Pro tip: Wherever you found outdated facts in your baseline, update or redirect the old page. Claude's search often surfaces stale pages that still rank.
6

Earn independent third-party mentions

Category recommendations depend heavily on what other sites say about you, so this is the slowest step and the one with the most effect. Prioritise sources Claude cited for your category prompts in step 1. Typical targets:

  • Review platforms such as G2, Capterra or Trustpilot
  • Industry 'best of' lists and roundups
  • Trade publications and podcasts
  • Relevant Reddit and forum threads where you participate transparently
  • A Wikidata item, plus Wikipedia only if you genuinely meet its notability rules

Each mention should state your category in plain words, e.g. 'Example Co, an invoicing tool for freelancers'.

Pro tip: Ask a reviewer or journalist to update an existing article that already ranks for your category. That usually pays off faster than pitching a new one.
7

Re-test in Claude and track the mention rate

Every 4-6 weeks, rerun the same prompt set from step 1 in fresh chats with search on and off, three runs each.

  • Search-on results should improve within weeks once crawlers have access and your pages are indexed.
  • Search-off results only change when Anthropic releases a model trained on newer data, which can take months.

Log mention rate, accuracy and cited sources, then send outreach effort toward the sources Claude keeps citing.

text
Prompt | Run 1 | Run 2 | Run 3 | Rate | Cited sources
Best invoicing tools for freelancers (search on) | yes | yes | no | 2/3 | example.com, g2.com
What is Example Co? (search off) | yes | yes | yes | 3/3 | -

Platform-specific notes

Cloudflare

AI crawler blocking can be switched on in the dashboard under Security > Bots and in AI Crawl Control. Newer zones may block AI bots by default. Allow ClaudeBot, Claude-SearchBot and Claude-User there, and turn off any managed robots.txt that adds Disallow rules for them.

WordPress

Edit robots.txt via Yoast SEO > Tools > File editor, or Rank Math > General Settings > Edit robots.txt. Security plugins such as Wordfence or a hosting firewall can still rate-limit bots, so check their blocked-traffic logs for Claude user agents.

Shopify

You can't edit robots.txt directly. Create the robots.txt.liquid template under Online Store > Themes > Edit code > Add a new template, and add Allow groups for the Anthropic agents above the default rules.

Wix

Go to Marketing & SEO > SEO > Robots.txt Editor (under Tools and settings) to add the user-agent groups. Wix pages are server-rendered, so the content-in-HTML step usually needs no changes.

Next.js

With the App Router, define rules in app/robots.ts and return separate rules entries for ClaudeBot, Claude-SearchBot and Claude-User. Render key pages with Server Components or static generation, not client-only fetching.

Check that it worked

✅ How to verify

  • Run `curl -I -A "ClaudeBot" https://example.com/` and `curl -I -A "Claude-SearchBot" https://example.com/`. Both should return 200 (or a 301 to a page that returns 200), not 403 or 429.
  • Check server or CDN logs for requests from ClaudeBot and Claude-SearchBot that received 200 responses after your changes.
  • In claude.ai with web search on, ask 'What is Example Co?' and confirm the answer is accurate and cites example.com.
  • Compare your prompt-set mention rate with the baseline after 4-6 weeks. Category prompts should show your brand more often than before.

⚠️ Common mistakes

  • Allowing ClaudeBot in robots.txt while a CDN bot-management rule or WAF still returns 403 to it.
  • Blocking Claude-SearchBot and Claude-User while allowing only the training crawler. This keeps you out of live search answers, which update fastest.
  • Publishing promotional pages full of superlatives and without concrete facts like prices, limits or audience, which gives Claude nothing specific to quote.
  • Testing once with a single prompt and judging success from it, even though answers vary between runs and between search-on and search-off modes.

Is this still an issue on your site? Run a free audit — ZORO checks this and 294 other SEO, AI-visibility, accessibility and security signals, with fixes for your platform.

Check my site free

FAQ

How does Claude decide which brands to recommend?

Claude combines patterns from its training data with live web search results when search is used. Brands that are described consistently across their own site and many independent sources, such as reviews, lists and press, are named more often and more accurately.

Should I block ClaudeBot to protect my content?

Blocking ClaudeBot keeps your pages out of future training data, which lowers the chance of being mentioned when Claude answers without search. If you want visibility, allow ClaudeBot. If you have licensing concerns, allow at least Claude-SearchBot and Claude-User so live answers can still cite you.

Does an llms.txt file help Claude mention my brand?

There is no confirmed evidence that Claude uses llms.txt for ranking or recommendations. It costs little to add, but crawler access, clear on-page facts and third-party mentions matter much more.

How long until Claude starts mentioning my brand?

With web search on, improvements can appear within a few weeks once crawlers can access your pages and they are indexed. Answers without search only change when a new model trained on newer data is released, which typically takes several months.