Clay Built a $3.1B Company on Who
April 2026
Clay Built a $3.1B Company on “Who.”
The “What” Is Still Wide Open.
September 2025
Data enrichment is the process of augmenting raw business data - company names, contact records, firmographic attributes - with structured, verified, and actionable information. Clay built a $3.1 billion company by enriching who to talk to. The equivalent enrichment for what to say - Commercial Truth about your own company - doesn’t exist yet. It’s arguably a larger market.
Here’s an observation that I think frames a massive opportunity.
Over the past five years, the B2B revenue stack split into two distinct layers. The “who” layer - who to sell to - got tremendous investment. Clay, Apollo, ZoomInfo, Clearbit, 6sense. Collectively valued at tens of billions of dollars. They solve the problem of identifying, enriching, and prioritizing prospect data.
The “what” layer - what to say when you reach them - got almost nothing.
This asymmetry is remarkable. We’ve built billion-dollar companies to ensure your email reaches the right person at the right company with the right title at the right time. But the content of that email? That comes from a Google Doc one of your marketers wrote six months ago.
We know, with engineering precision, who our prospects are. We have no idea, with any precision at all, whether what we’re saying to them is accurate.
According to Grand View Research, the B2B data enrichment market reached $3.5 billion in 2024 and is growing at 12.4% CAGR (Grand View Research Data Enrichment Market Report, 2024). The market for governing what companies actually say about themselves - Commercial Truth management - has no analyst coverage because the category hasn’t been named. It’s a white space.
The Asymmetry
Let me illustrate the asymmetry with a concrete example.
A modern revenue team using Clay can build an outreach sequence that:
- Identifies a VP of Operations at a logistics company headquartered in Chicago
- Enriches the record with recent funding data, tech stack signals, and hiring trends
- Scores the account based on ICP fit with 97% confidence
- Personates the message with relevant industry context, the recipient’s LinkedIn activity, and their company’s recent press mentions
The targeting is extraordinary. The personalization is precise. The timing is optimized.
Now: what does the email actually say about the sender’s product?
It says whatever was in the product messaging document that someone uploaded during the AI SDR tool’s initial setup. The pricing from Q2 (it’s now Q4). The integration count from six months ago (it’s increased by 30%). The competitive differentiator that was neutralized by the competitor’s last release.
The “who” side of the equation is managed with engineering-grade precision. The “what” side is managed with archaeological-grade guesswork.
Research from Gartner shows that B2B companies that invest equally in prospect intelligence and commercial intelligence achieve 28% higher conversion rates than companies that invest disproportionately in one dimension (Gartner Revenue Intelligence Benchmark, 2024). The industry currently invests roughly 20:1 in prospect intelligence versus commercial intelligence.
Why the Gap Exists
I think the gap exists because of a cognitive bias in how revenue leaders think about sales effectiveness.
When a deal fails, the first question is usually: “Did we have the right prospect?” Were they in our ICP? Did they have budget? Were we talking to the decision-maker? The assumption is that if the targeting was right, the deal should have worked - and if it didn’t, the targeting must have been wrong.
This assumption is so deeply embedded that entire categories of tooling - enrichment, intent data, ABM, buyer signal platforms - exist to optimize it.
But the data tells a different story. 40-60% of qualified pipeline ends in “no decision” (Gartner, 2024). These weren’t bad prospects. They were qualified, engaged, and evaluating. The targeting was right. Something else went wrong.
That “something else” is almost always on the “what” side: the information the buyer received was inconsistent, the competitive positioning was stale, the case study referenced a churned customer, the pricing in the email didn’t match the website. The right prospect received the wrong information.
All the enrichment in the world can’t overcome a product pitch that contradicts itself across touchpoints. Clay can tell you exactly who to talk to. It can’t ensure that what you tell them is accurate.
Data from LinkedIn shows that 52.2% of sales professionals identify “lost sales and revenue” as the biggest impact of sales-marketing misalignment - ahead of reduced productivity, lost customers, and damaged brand reputation (LinkedIn State of Sales, 2024). Misalignment isn’t a targeting problem. It’s a truth problem.
The Market-Sizing Argument
Here’s the market-sizing argument that I find compelling.
Clay’s premise: B2B companies need structured, enriched, accurate data about who to sell to. Market size: $3.5B and growing. Clay’s valuation: $3.1B.
The Commercial Truth premise: B2B companies need structured, enriched, accurate data about what to say. Market size: at least as large, arguably larger.
Why larger? Because “what to say” affects every stage of the funnel, not just the top. Enrichment data is primarily top-of-funnel: identify and prioritize accounts. Commercial Truth touches every interaction from first touch to closed-won to renewal: what the AI SDR emails, what the chatbot says, what the proposal contains, what the AE presents, what the renewal team communicates.
The “what” market also includes regulatory compliance (EU AI Act), content production efficiency (eliminating the 65% waste rate from Highspot, 2024), AI governance (ensuring every agent is accurate), and employee knowledge management (reducing the 5.7-month ramp time from Gartner, 2025). Each of these is a separate value driver that enrichment data doesn’t touch.
If the “who” market is $3.5B, the “what” market - once it has a name and a category leader - is likely $5-10B. The TAM isn’t smaller. It’s less visible, because the category hasn’t been named yet.
The Convergence Point
Here’s what I think happens next.
As AI agents become the primary interface between companies and their markets, the “who” and the “what” converge. An AI agent doesn’t just need to know who to contact. It needs to know what to say, and it needs that information to be accurate, current, and contextually appropriate.
Clay solved half of what an AI agent needs. The other half - governed, verified, continuously updated Commercial Truth - is the missing piece.
The companies that solve both halves will have agents that are not only well-targeted but consistently accurate. The companies that solve only the targeting half will have agents that reach the right people and say the wrong things.
Research from Forrester indicates that 78% of B2B buyers consider “accuracy of vendor claims” to be as important as “relevance of outreach” when evaluating vendors (Forrester B2B Buying Expectations Survey, 2025). Buyers want to be reached with the right message - but the message needs to be right.
The “What” Stack
If the “who” stack is enrichment + intent + scoring + targeting, the “what” stack would be:
- Knowledge graph: structured, governed, confidence-scored commercial facts
- Claim verification: continuous validation of assertions about products, pricing, capabilities, competitive positioning, and customer evidence
- Propagation engine: automatic distribution of truth updates to every AI tool and content system
- Readiness assessment: continuous measurement of whether humans in the organization know what’s currently true
Each component has a direct parallel in the “who” stack:
| “Who” Stack | ”What” Stack |
|---|---|
| Contact enrichment (Clay, ZoomInfo) | Knowledge graph (commercial facts) |
| Intent signals (6sense, Bombora) | Claim verification (truth signals) |
| Account scoring (Madkudu, 6sense) | Confidence scoring (claim reliability) |
| Targeting rules (Apollo, Outreach) | Propagation rules (truth distribution) |
The “who” stack is mature. The “what” stack is nascent. And the market opportunity for the “what” stack is at least as large - because every company that invested in the “who” stack now needs the “what” stack to ensure their well-targeted outreach is also well-informed.
Frequently Asked Questions
What is the difference between prospect intelligence and commercial intelligence?
Prospect intelligence is enriched data about who to sell to - contact information, firmographics, tech stack, intent signals, and buying behavior. Commercial intelligence is governed data about what to say - product capabilities, pricing, competitive positioning, customer evidence, and compliance claims. The B2B industry has invested heavily in prospect intelligence (Clay, Apollo, ZoomInfo) while leaving commercial intelligence almost entirely unmanaged.
How big is the market for Commercial Truth management?
The B2B data enrichment market for prospect intelligence reached $3.5 billion in 2024 (Grand View Research). The market for Commercial Truth management - governing what companies say about themselves - is estimated to be at least as large ($5-10B), given that it affects every stage of the revenue cycle from first touch through renewal, and includes regulatory compliance, AI governance, and content efficiency as additional value drivers.
Why did the “what” problem get less investment than the “who” problem?
The “who” problem has visible failure modes - wrong contacts, bad data, missed accounts - that produce measurable, attributable costs. The “what” problem has invisible failure modes - inconsistent messaging, stale claims, trust erosion - that manifest as “no decision” deal outcomes without clear attribution. Investment follows visibility, and the “what” problem only became visible at scale when AI began amplifying commercial claims to tens of thousands of touchpoints per month.
How do the “who” and “what” problems converge in the AI era?
As AI agents become the primary interface between companies and their markets, they need both accurate targeting (who to contact) and accurate messaging (what to say). An AI agent with perfect targeting but stale knowledge reaches the right prospects and tells them the wrong things. Convergence means combining enrichment-grade precision for prospect data with truth-grade governance for commercial data.