Clay's search strategy has two layers that pull in opposite directions. The first is a community flywheel: Slack, Clay Clubs, Clay University, cohorts, the Sculpt conference and a marketplace. That flywheel made "Clay" a searched-for word and created the GTM engineer job title. The second is a Programmatic SEO push: 27,593 sitemap URLs carry a June 2026 lastmod, and the bulk of the sitemap is dossier pages on named executives and company funding, the template Apollo has just retired with HTTP 410.
Footprint
There are two sitemaps and 33,939 URLs, crawled 5 October 2026. 16,509 sit under /fr/.
| Section | URLs | What it is |
|---|---|---|
| /dossier | 20,677 | "Who is the CEO of 0x in 2026?", "How much did X raise?" |
| /gift | 6,835 | Per-account personalised pages, noindex |
| /integrations | 3,339 | Data points, signal points, provider pages |
| /glossary | 536 | GTM terms |
| /blog | 523 | Company news and playbooks |
| /tools | 472 | "How to append data to records with AI — Sculptor by Clay" |
| /faq | 252 | Product and referral FAQs |
| /providers-signal-catalogue | 210 | Signal types by provider |
| /livestreams | 202 | Webinar archive |
| /templates | 198 | Workflow templates |
| /changelog | 168 | Release notes |
| /guides | 108 | Long-form guides (own sitemap) |
| /data-tests | 24 | "Best catch-all email verifiers 2025" benchmarks |
Lastmod by month: 27,593 URLs in June 2026, then between 900 and 2,700 a month since. The stack is Webflow; the crawler's HubSpot hint is a false positive.
What ranks
From Semrush's public overview for August 2026, a third-party estimate: 179.46K organic search visits, down 12% month on month. Total visits were 1.57M (+12%), paid search was only 5.58K, Authority Score 51, 11.44K referring domains (+11%) and 193.11K backlinks. The United States accounts for 19.3% of visits; Bangladesh's 30.7% looks like panel noise.
Two things stand out. Organic traffic fell 12% in August, two months after the June wave of roughly 27,600 re-dated URLs, so the dossier push has not visibly paid off yet. And at about 11% of visits, organic search is a minor channel next to direct and branded traffic, which the community drives.
Whether Google indexes the dossier pages at scale, and for which queries ("[company] CEO", "[company] funding"), is not verified. No site: counts or SERP data were collected, and Semrush's keyword table is paywalled.
Programmatic and template plays
Dossier (20,677). Who is the CEO of 0x in 2026? is about 3,800 words with ProfilePage and FAQPage schema. It shows the executive's LinkedIn URL, work history and a work email address marked "Status: Verified", followed by a "Use Clay's email finder" call to action. Eight dossier types exist, including company-email and company-funding. This is person-level data published for search traffic. For EU-resident executives it raises the GDPR Article 14 and legitimate-interest questions that public directories attract. See GDPR and ePrivacy.
The French layer is English. /fr/dossier/10-minute-school-funding carries lang="fr" and hreflang fr, but its title and body are English ("How Much Did 10 Minute School Raise?"). About 16,500 /fr/ URLs are likely duplicates of this kind. That doubles crawl load without adding a market.
Gift pages (6,835) contradict themselves. /gift/10x-genomics is titled just "Clay" and carries noindex, nofollow, yet every gift URL is listed in the sitemap. They are ABM pages personalised for target accounts and were never meant to rank. Listing them wastes crawl budget and sends Google conflicting signals.
What works. The ABM orchestration glossary entry runs about 4,400 words. Tools pages carry HowTo schema. The data-tests series benchmarks providers such as catch-all verifiers and European mobile providers by country, original research no competitor publishes. With 3,339 integration and provider pages, Clay presents itself as the hub of the data layer.
Placeholder copy everywhere. The site-wide navigation shows "Link long form description will go in this slot here." under several menu items. It appeared dozens of times in the text of a single dossier page. robots.txt disallows lorem-ipsum CMS slugs such as /data-inputs/aliquam-debitis one by one, and the sitemap still lists /admin/* and /blog-posts-test-locale/*. The output is large and lightly reviewed.
AEO posture
- No
llms.txt,llms-full.txtorpricing.md. Clay is the only large player in this group with none of them. - No AI-bot rules, so all crawlers are allowed by default.
- JSON-LD: the homepage has SoftwareApplication, AggregateRating and Review; the pricing page has only WebSite.
- Agent surfaces live on subdomains:
agent-docs.clay.com,developers.clay.com(Mintlify),chatgpt-app-widget.clay.comandmcp-app-widget.clay.com. Clay builds for agents inside its product and neglects agent-readable marketing.
For a company whose users ask ChatGPT "how do I do waterfall enrichment", the missing llms.txt is an odd gap. See llms.txt, pricing.md and AI crawlers.
Subdomains
There are 71 hosts in certificate logs and 41 live. This is the richest set of community properties in the market:
| Host | What it is |
|---|---|
| university.clay.com | Clay University (docs. and www.docs. redirect here) |
| community.clay.com | Forum; robots allows /members/ and /events/ |
| cohorts.clay.com | Paid and live cohorts |
| marketplace.clay.com | "GTM skills built by the community" (Astro, indexable) |
| sculpt.clay.com | Sculpt, Clay's GTM conference |
| events.clay.com | "cro-summit" |
| sequencer.clay.com | Page title "Smartlead". The Smartlead-powered sequencer is visible in cert logs |
| partner. / rewards. / terra. | Partner, referral and Terra apps |
| web-staging.clay.com | Live staging site; canonical points to www |
Putting university, community and marketplace on subdomains splits their authority from www. See Subdomain strategy.
Off-site
This is the engine. Per Clay's own post (February 2025, vendor-claimed): "~20k person Slack community", "40+ Clay Clubs now meet in 20+ countries", "more than 100 agencies run on Clay", and one partner bootcamp with "150+ students". The YouTube channel @GrowWithClay has 17.2K subscribers (5 October 2026). Clay is hiring a "GTME University Lead", so it treats education as a product line.
Agencies and freelancers who sell Clay builds publish tutorials and LinkedIn posts about Clay, probably more than clay.com itself (inference). See Content engines.
Weak spots
- Legal exposure from publishing named executives' emails at scale. This is the risk Apollo walked away from.
- No visible payoff yet: organic traffic fell 12% month on month in August, after the June publishing wave (Semrush estimate).
- Fake French: about 16,500
/fr/URLs with English bodies. - Hygiene: noindex pages in the sitemap, placeholder navigation copy, admin and test URLs listed, lorem-ipsum slugs.
- No llms.txt or pricing.md, even though Clay's dual credit meter is exactly the kind of pricing assistants get wrong without a machine-readable source.
What this means for an entrant
- Clay's moat is community, not pages. An entrant cannot out-publish 34,000 URLs, but it can own one community Clay does not serve: EU agency operators who care about compliance. Clay Clubs in Warsaw and Berlin show the demand exists. See European outbound agencies.
- Do not copy the dossier play. For an EU-domiciled company with the trust positioning, publishing named people's emails would undermine the pitch. "Is Clay GDPR-compliant?" is itself a query worth answering. See Trust as positioning.
- Copy data-tests. Original benchmarks, such as inbox placement by provider and by EU country, earn links and AI citations. Clay's 24 tests are its best pages. See Inbox placement testing.
- Integrate early. Clay is where GTM engineers build. A Clay integration page and a marketplace listing ("route Clay-built campaigns' replies into the reply desk") put the The reply desk in front of its most active users.
- Answer the questions Clay leaves open. "Clay pricing explained", "Clay credits calculator" and "Clay vs [x] for agencies" go unanswered because Clay publishes no pricing.md or comparison pages.