Clay's $3.1bn Valuation Is a Verdict on Apollo & ZoomInfo
Clay raised $100mn Series C at $3.1bn in August 2025, becoming the orchestration layer on top of Apollo, ZoomInfo, Lusha, and Cognism. Buyers now treat the data vendors as interchangeable utilities — and pay Clay the margin to route between them.
In August 2025, Clay closed a $100mn Series C led by CapitalG at a $3.1bn post-money valuation. Total funding now sits at $204mn, and 2025 revenue is reportedly tracking to $100mn — a roughly 3x increase year over year. Customers include OpenAI, Anthropic, Canva, Intercom, and Rippling: the entire AI-native B2B SaaS cohort, plus a long tail of growth-stage outbound teams.
What Clay actually does is route lookups across Apollo, ZoomInfo, Lusha, Cognism, and a long list of smaller enrichment vendors. It does not own a contact database. It does not collect emails. It is a workflow shell that calls other people’s APIs and stitches the results into a single record.
The $3.1bn valuation is the unusual part — because it is being paid by buyers who, by definition, are also paying the underlying data vendors. The market has decided that the orchestration layer is worth more than the layer it sits on.
That has consequences for everyone selling lead data, including the long tail of LinkedIn scrapers documented in the Q1 2026 lead and contact extractors report on this site.
What the data vendors actually charge
The pricing on the vendors Clay sits on top of has moved sharply in the last year. The direction is up.
Apollo raised its basic plan from $19/seat/month to $59/seat/month over roughly twelve months. Mid-tier plans run $49–$119/seat/month annual. Apollo’s pitch — “self-serve, transparent pricing, large database” — has not changed, but the entry-level price has tripled.
ZoomInfo has not published a public price card in years; reported enterprise floors sit around $15K/year per seat, scaling to $26K+ for the larger packages. That is roughly 3–5x Apollo on a per-seat basis, justified historically by data depth and intent signals.
Lusha runs $29–$51 per seat per month with a credit-based model: ten credits per phone number, fewer per email. Buyers who hit the cap often layer in Apollo or another vendor for overflow.
Cognism does not publish pricing publicly, but is consistently described as a 20–50% premium over Apollo for GDPR-compliant EU mobile data — the segment most exposed to the EU AI Act Article 5 and to noyb-style enforcement.
Sales Navigator, the original source of much of the data the others resell, runs $119.99/month for Core, $159.99 for Advanced, and $1,600+/seat/year for Advanced Plus annual contracts.
These are not interchangeable products by feature. They are increasingly interchangeable by outcome: a verified email address and phone number for a target persona at a target company. And once buyers treat them as interchangeable by outcome, the question becomes which one to call for which lookup — which is the question Clay answers.
Why the orchestration layer captured the margin
The Clay valuation works because buyers were already running this orchestration logic in spreadsheets, Zapier flows, and sales-engineer Python scripts. Clay productized the workflow. The product is not the data; the product is the routing.
A typical Clay enrichment for a single contact might call Apollo first (cheapest), fall back to ZoomInfo if the email is missing, fall back to Lusha if the phone is missing, validate the email through a third vendor, and stitch the result into a CRM record. Each call has a cost. Each cost is metered. Clay’s margin is the difference between the per-record price the customer pays Clay and the sum of the per-call prices Clay pays the underlying vendors.
That margin is sustainable as long as the routing logic is non-trivial. Right now it is. The vendors do not interoperate, the success rates are inconsistent, the credit models are different, and the data quality varies by region and seniority. A team building this routing in-house can do it, but it takes engineering time that doesn’t scale. Clay charges for not having to build it.
The risk to that margin is that the underlying vendors converge on a common interface — exactly the thing MCP is doing for scraping APIs. If Apollo, ZoomInfo, and Lusha all expose MCP servers with comparable tool surfaces, an LLM client can do the routing without Clay in the middle. Whether that happens in 2026 is the most consequential unresolved question in this market.
What the Apify Store data shows about the bottom of the stack
The Q1 2026 lead and contact extractors report measured 2,031 actors holding 43,027 30-day active users on the Apify Store. The category is dominated by LinkedIn-targeting scrapers — 65% of demand. The single most predictive marketing phrase in the entire category is “No Cookies”: 48 actors using it in the title hold 56.6% of segment demand.
That is the bottom of the stack Clay sits on. Buyers in that segment are explicitly trying to avoid Apollo and ZoomInfo’s published rates and running their own LinkedIn extraction at marginal cost — fractions of a cent per record against fifty cents through Apollo’s API.
The open question is whether Clay (or a Clay-like product) will eventually route through these primitives directly. The economics suggest it must. A LinkedIn profile lookup that costs $0.001 through an Apify actor cannot, indefinitely, be re-priced at $0.50 through Apollo’s API and routed through Clay at a higher rate. Either the bottom of the stack gets absorbed into the orchestration layer, or the orchestration layer’s margin gets compressed.
What this means for publishers in the segment
For Apify Store publishers running LinkedIn scrapers, the Clay valuation is good news in the same way the rising tide is good news. It validates the market: there is enough enterprise spend on B2B contact data to support a $3bn orchestration company, which means there is enough underlying volume to support a healthy long tail of low-cost data extractors.
The risk is positioning. The actors that win the orchestration channel will be the ones whose API surface is clean, predictable, and benchmarkable — exactly the same shape the MCP integration layer is rewarding. An actor with a confusing input schema and a 60% success rate will not be selected by an orchestrator routing thousands of calls per minute. An actor with a clean schema, 95% success, and a published per-record price will be.
That is a different selling motion than ranking in the Apify Store search bar. It is closer to becoming a wholesale data primitive. The publishers who build for that motion — and who are willing to be a tool inside someone else’s product — will capture the next leg of growth in this segment. The publishers who insist on owning the buyer relationship directly will keep playing on the SEO and Store-discovery surface, which is real but capped.
The Clay valuation says one of these two strategies has roughly thirty times the revenue ceiling of the other.
Sources