Claude Mentioned 20 Brands Across 20 Queries. It Linked to Zero Brand Websites.

Claude mentioned 20 brands across 20 B2B SaaS queries in Wave 4 of our citation study. It cited only 4 of those brands with source links, the lowest citation rate of any AI search engine at 20.0%. More striking: across 160 total citation URLs produced by Claude, not one pointed to a brand-owned website. That is a 0.0% brand-own-site rate. ChatGPT, running the same 20 queries, linked to brand websites 19.1% of the time.

The practical difference matters for anyone measuring AI-driven referral traffic. A brand that appears in Claude responses is getting name recognition. It is not getting clicks.

Why This Matters If You Track AI Visibility

Most "share of voice" metrics count mentions. You run a query, see if your brand came up, mark it as a win. That approach treats Claude and ChatGPT as equivalent channels. They are not.

When ChatGPT mentions your brand in a recommendation, there is roughly a 1 in 5 chance it also links directly to your website. Users can click through. You get referral traffic. The mention is actionable.

When Claude mentions your brand, the recommendation arrives without a link. Users who want to investigate further have to search for you separately. That friction is real. In our Wave 4 data, Claude mentioned brands 20 times across 20 queries and provided source links for brands in only 4 of those responses.

For context on how AI citation rates are changing overall, our Wave 4 analysis found AI engines now cite brands with source links 27% more often than three weeks prior. Claude is moving in the opposite direction.

The Engine Citation Rate Table

As of April 2026, here is the per-engine breakdown from Wave 4, running 20 queries across 5 B2B SaaS categories (CRM, Project Management, Email Marketing, Analytics, Dev Tools):

Engine	Brands Mentioned	Brands Cited (with source link)	Citation Rate
ChatGPT	20	14	70.0%
Gemini	22	10	45.5%
Perplexity	23	10	43.5%
Grok	22	7	31.8%
Claude	20	4	20.0%

Claude's 20.0% citation rate is not even close to the next-lowest engine. Grok, second-to-last, cites at 31.8%. ChatGPT cites at 70.0%, three and a half times Claude's rate.

The 4 brands Claude did cite with source links: ActiveCampaign (in the email marketing comparison query), Amplitude (analytics tools for startups), Vercel (Vercel vs Netlify), and Netlify (also Vercel vs Netlify). In 16 of 20 queries, Claude named brands and gave no links to any of them.

The Brand-Owned URL Finding

The source type breakdown is where Claude's behavior becomes most distinct.

Engine	Brand Own Site	Third-Party Review	Reddit/Forum	Aggregator	Wikipedia	Other	Total URLs
ChatGPT	22 (19.1%)	53 (46.1%)	3 (2.6%)	0 (0.0%)	4 (3.5%)	33 (28.7%)	115
Gemini	12 (5.2%)	154 (67.0%)	5 (2.2%)	0 (0.0%)	0 (0.0%)	59 (25.7%)	230
Perplexity	7 (4.3%)	110 (67.1%)	0 (0.0%)	1 (0.6%)	0 (0.0%)	46 (28.0%)	164
Grok	4 (0.9%)	354 (77.8%)	13 (2.9%)	22 (4.8%)	0 (0.0%)	62 (13.6%)	455
Claude	0 (0.0%)	111 (69.4%)	0 (0.0%)	9 (5.6%)	0 (0.0%)	40 (25.0%)	160

Every other engine links to brand-owned websites at least some of the time. Gemini, which is not known as a particularly brand-friendly engine, still links to brand sites 5.2% of the time. Claude lands at 0.0% across 160 URLs.

When Claude provides citations, 69.4% go to third-party review sites and 25.0% go to an "other" category (press coverage, industry publications, reference content). Aggregators like G2 and Capterra account for 5.6%. Brand websites get nothing.

What This Looks Like in Practice

When we asked Claude for the "best CRM for startups" in Wave 4, it recommended Salesforce first. The response mentioned Salesforce, HubSpot, and Pipedrive. No source links accompanied any of those recommendations. Users reading that response have no direct path to any brand's website from within the Claude answer.

Four of the other five engines cited at least one CRM brand with a source link for that same query. Claude was the only engine that gave a pure-text recommendation with no supporting citations for CRM brands.

The same pattern appeared across categories. For "analytics tools for startups," Claude cited Amplitude with a source link but mentioned Mixpanel, Google Analytics, and PostHog without links. For the "best PM tool for engineering teams" query (the one where all 5 engines unanimously recommended Linear), Claude named Linear but did not link to it. ChatGPT, Perplexity, Gemini, and Grok all provided source citations for Linear in that same response.

What This Means for AEO Strategy

The "mention without credit" problem has two practical consequences.

First, measuring Claude mentions as equivalent to mentions from other engines overestimates Claude's referral value. If your reporting shows 10 Claude mentions and 10 ChatGPT mentions, those are not interchangeable. The ChatGPT mentions may include 2-3 direct links to your site. The Claude mentions deliver zero.

Second, Claude appears to use a different threshold for when citations are warranted. For direct comparison queries ("Vercel vs Netlify"), Claude does provide citations. For broader recommendation queries ("best CRM for startups"), it does not. The pattern across 16 of 20 queries without brand citations suggests Claude treats most recommendation responses as conversational rather than sourced.

This distinction matters when deciding where to focus content investment. Building structured, citable content still matters for Claude, because it affects whether Claude mentions your brand at all. But optimizing for Claude citation links, using the same approach as optimizing for Claude AI mentions, requires accepting that Claude will rarely send users directly to your site.

The contrast with ChatGPT is worth sitting with. ChatGPT is already the most independent AI search engine, disagreeing with other engines on brand recommendations more than any other pair (56-67% overlap vs 73-83% for the rest of the cluster). It is also the engine most likely to link directly to a brand's website when it recommends them. For traffic-focused AEO, ChatGPT and Claude are operating in different modes.

For comparison, Grok's citation behavior shows yet another distinct pattern: massive URL volume (455 URLs vs Claude's 160) but concentrated in third-party review sites and Reddit, with only 0.9% going to brand-owned sites in Wave 4.

What You Can Do About It

Count citations, not just mentions, when measuring Claude share of voice. Claude mentions your brand in responses, but only 4 of 20 brands received source links. If you are tracking Claude visibility, separate "mentioned" from "cited" in your reporting.
Do not treat Claude mentions as traffic drivers. Claude responses link to brand websites 0% of the time. Budget accordingly. Attribution models that assign equal value to all AI mentions will overcount Claude's contribution to referral traffic.
Prioritize direct comparison queries for Claude. Claude's 4 citations all came from explicit comparison queries (email marketing comparison, analytics comparison, Vercel vs Netlify). If you can appear in Claude's responses to head-to-head queries, you are more likely to receive a source link than in broad recommendation queries.
Use the invisible brands data as a floor check. Before worrying about citation quality on Claude, confirm you are being mentioned at all. Brands like Height had zero Claude mentions across all 4 waves. Getting mentioned is step one; getting cited with a link is step two.
Cross-check across engines. Claude's 0.0% brand-site citation rate is a structural pattern, not a one-week anomaly. Claude's citation counts: 6, 6, 6, 4 across four waves, consistently the lowest. Multi-engine tracking captures the full picture that single-engine measurement misses.

Methodology

We ran 20 queries across 5 AI search engines (ChatGPT, Perplexity, Gemini, Grok, Claude) on April 7, 2026, covering 5 B2B SaaS categories: CRM, Project Management, Email Marketing, Analytics, and Dev Tools. Each query was sent as a real-time API call, simulating how actual users interact with these platforms. We tracked 25 B2B SaaS brands across those categories, categorizing each engine response by which brands were mentioned (named without a source link) and which were cited (named with a direct source URL). We also categorized citation URLs by type: brand-owned sites, third-party review sites, Reddit/forums, aggregators (G2, Capterra, etc.), Wikipedia, and other. This is the fourth wave of a longitudinal study that began March 6, 2026.

FAQ

Why does Claude link to so few brand websites? Claude produces citation URLs that skew heavily toward third-party review sites (69.4% of all Claude citations) and aggregator content. Claude appears to treat brand-owned websites as a less authoritative source type than independent reviews, defaulting to third-party validation over direct brand links. This is a structural pattern: 0.0% brand-owned site citations across 160 Claude URLs in Wave 4, consistent across the prior three waves.

Does Claude mentioning your brand still matter for AEO if it doesn't link to you? Yes, but the value is different. Claude mentions contribute to brand familiarity and may influence users who then search independently. However, mentions without links generate zero direct referral traffic. For marketers measuring AI-sourced traffic, Claude mentions should be tracked separately from ChatGPT or Perplexity mentions, which do link to brand sites at meaningful rates (19.1% and 4.3% respectively).

Which AI engine is most likely to link directly to a brand's website? ChatGPT, at 19.1% of citation URLs pointing to brand-owned sites in Wave 4. Gemini is second at 5.2%, followed by Perplexity at 4.3%. Grok dropped to 0.9% in Wave 4 (down from 8.5% in Wave 3). Claude is at 0.0%.

Is the 0.0% Claude brand-citation rate a new development? Claude has been the lowest-citing engine by brand count across all four waves of our study (March and April 2026). The 0.0% brand-own-site rate is specific to Wave 4, but Claude's pattern of producing fewer brand citations than any other engine is consistent throughout the study. What changed in Wave 4 is that Claude's brand-cited count dropped from 6 to 4 for the first time.

Should I optimize differently for Claude versus other engines? For Claude, the primary goal should be getting mentioned at all, because Claude's overall brand citation rate is the lowest of any engine and brand-site links appear to be absent by design. For engines like ChatGPT and Gemini, optimizing for citation quality (getting a source link, not just a mention) is a more achievable second-order goal once visibility is established.