Six Years of Obscurity, Then $100M ARR: The Origin Story of Clay

GTM Data Enrichment & Outbound Orchestration Platform

THE HOOK — An Afternoon That Should Have Taken a Sprint

It was a Tuesday afternoon. A growth operator at a mid-size SaaS company had a list of 3,000 prospects — companies recently hiring for RevOps roles, which usually meant they were about to invest in their sales stack. She needed emails. She needed company sizes. She needed the name of the VP of Sales at each one. She needed to know which ones were already using Salesforce.

Before Clay, this was a multi-week project. You'd buy a ZoomInfo license, pull what you could, then hand the gaps to an offshore VA to manually look up on LinkedIn. You'd use Hunter.io for emails and accept the 30% hit rate as the cost of doing business. You'd get an engineer to write a script that called an API, probably break it twice, and by the time the list was clean enough to actually use, the signal had gone cold. The companies had already bought something else. The moment had passed.

She opened Clay on that Tuesday afternoon. She built a waterfall. Apollo first — try to find the email. If Apollo misses, cascade to Hunter. If Hunter misses, try Clearbit. If Clearbit misses, try FullEnrich. The system tries each source in sequence, charges her only when it finds something, and moves on when it doesn't. In two hours, she had 94% email coverage on 3,000 records. She ran an AI enrichment step on top — a Claude prompt asking, for each company, what their likely pain points were based on their job postings and recent news. Each row in her spreadsheet now had a personalized first line for her outbound sequence. She launched the campaign before dinner.

This is not a story about automation. It is a story about leverage. And it is the story of why Clay, a company that spent six years building in obscurity before the world noticed, became worth $3.1 billion.

THE BACKSTORY — Who Is Kareem Amin, and Why Did He Spend Six Years Before Anyone Cared

Kareem Amin grew up between cultures — he is fluent in Arabic, English, and French, and studied at McGill University in Montreal before making his way into the technology world. He worked at Facebook during a period when the company was still figuring out how to translate a social graph into a commercial entity. The experience taught him something that would take years to fully articulate: data, when structured right, is not a resource. It is infrastructure. It is the difference between an organization that can move and one that cannot.

He co-founded Clay in 2017 alongside Nicolae Rusan and Varun Anand. The company was accepted to Y Combinator and subsequently backed by Sequoia Capital in 2019, with partners Alfred Lin, David Cahn, and James Flynn seeing something in the original thesis that most people would have called too abstract.

The original product idea was ambitious to the point of being difficult to explain. Kareem wanted to democratize the power of programming — to give non-technical people the ability to build structured workflows over data the way engineers could with code. The early Clay looked like a spreadsheet. It acted like an API orchestration layer. It was genuinely hard to categorize. Not a CRM. Not a data warehouse. Not quite automation software in the Zapier sense. Something in between.

This ambiguity was almost fatal and ultimately defining. Clay spent years as a "power user tool" — beloved by a tiny group of operators who understood what it could do, invisible to everyone else. The first six years were not a failure. They were an excavation. Kareem and the team were digging toward the exact shape of the problem they were trying to solve, getting closer with every iteration without quite knowing when they would hit bedrock.

The bedrock turned out to be data enrichment. Specifically: the broken, expensive, frustratingly fragmented market for contact and company data that every go-to-market team on earth was forced to navigate alone.

THE GRIND — Building the Waterfall, Finding the Community, Catching the AI Wave

The insight, when it finally crystallized, was deceptively simple. There is no single data provider that has everything. Apollo is strong on emails for mid-market tech. ZoomInfo owns enterprise. Clearbit is clean but expensive. Hunter is cheap but thin. Every provider has coverage holes. Every provider charges you for the attempt, not the result. You either pay one provider too much for partial coverage, or you pay three providers and stitch the data together manually in a mess of spreadsheets and deduplication scripts.

Clay's answer was the waterfall. You define a sequence — try Apollo, then Hunter, then Clearbit — and the system runs down the chain until it finds a result. You pay only for successful lookups. Coverage goes from a provider-specific ceiling of 30-40% to a composite ceiling of 80%+ on the same list. The economics flip. Instead of paying a $30,000 annual contract for partial coverage, you pay per enriched record across multiple sources and come out ahead — more data, lower cost, no annual commitment.

This was not just a feature. It was a new way of thinking about data procurement that the industry had not named yet. Clay named it. The waterfall became the product's conceptual spine.

Building it required assembling something closer to a marketplace than a SaaS tool. Today, Clay integrates with 150+ data providers — Apollo, ZoomInfo, LinkedIn, Hunter, Clearbit, Lusha, FullEnrich, RocketReach, Cognism, and dozens more, including niche providers for specific geographies and verticals. Each integration required negotiating commercial terms, building the connector, and figuring out how to expose it in a UI that looked more like Excel than like a developer console. That last constraint was philosophical. Kareem believed the operators who most needed this power — growth leads, RevOps managers, demand gen specialists — were people who understood data structurally but could not write code. The tool had to speak their language.

The community strategy was not initially a strategy. It was a consequence of who the early users were. The people who found Clay in 2020 and 2021 were a particular type: operators who had pushed every other tool to its limit and were frustrated by the gap between what they could imagine and what they could build. They were former engineers who had moved into GTM roles. They were growth hackers who had taught themselves SQL and wanted more. When they discovered Clay, they did not just use it — they built elaborate workflows and shared them. The community became a library of recipes. Clay Clubs began self-organizing in cities from Manila to Warsaw. Agencies formed specifically around helping companies implement Clay. The platform created a professional ecosystem before it had a formal community program.

Then the AI wave arrived, and Clay was sitting in exactly the right place.

In 2023, as large language models went from research curiosity to production tool, Clay did something that would define its next chapter: it made AI models callable as enrichment steps inside a workflow. You could drop a Claude prompt into a row, feed it company context from your enrichment columns, and have it write a personalized opening line for your cold email. You could ask GPT-4 to classify a company's primary use case based on their website copy. You could chain AI reasoning across your entire prospect list without writing a single line of code.

Suddenly, Clay was not just a data enrichment tool. It was an orchestration layer — a place where structured data and AI inference met, where the output of one enrichment step became the input of the next, where a single operator could construct a workflow that previously would have required a data engineer, a copywriter, and a sales operations manager working in concert.

THE BREAKTHROUGH — The Unicorn Raise and the Category That Kareem Named

In January 2025, Clay raised $40 million in a Series B expansion round led by Meritech Capital, at a $1.25 billion valuation. The round was pre-emptive — Meritech moved without being asked because the numbers demanded it. Clay had grown 6x in 2024, following 10x growth in both 2022 and 2023. They had 5,000+ customers including OpenAI, Anthropic, Canva, Ramp, Rippling, and Intercom. Enterprise net revenue retention exceeded 200% — meaning every dollar of enterprise revenue was more than doubling within twelve months. The company had never churned an enterprise customer. Not one.

The press framed it as a unicorn moment. What it actually was, was a validation. The six years of quiet excavation had found the vein. The world had finally caught up to the problem Clay had been solving since 2017.

By August 2025, CapitalG led a $100 million Series C at a $3.1 billion valuation. Tender offers at $1.5 billion (May 2025) and $5 billion (January 2026) gave employees liquidity at a cadence almost unheard of in startup culture — Kareem's philosophy being that equity should serve people on their own timeline, not just at an IPO that may be years away.

But the more consequential breakthrough was cultural, not financial. It was a phrase.

In 2023, Kareem and co-founder Varun Anand began using the term "GTM Engineer" to describe what the best Clay users had become. Not salespeople. Not growth hackers. Not RevOps analysts. Something new: operators who built revenue engines the way software engineers build product features — with systems thinking, versioned workflows, and iterative experimentation. People who used tools like Clay the way a developer uses an IDE.

The naming act was itself a product strategy. By giving this person a job title, Clay gave thousands of operators a professional identity. The GTM Engineer was not just a Clay user — they were a category of professional that now had a career path, a community, and a growing market of employers posting job listings specifically for this role. Within months of the term circulating, hundreds of GTM Engineer positions appeared monthly on LinkedIn. Companies like Cursor, Lovable, and Webflow hired for the function explicitly. Agencies built service offerings around it. First-time entrepreneurs built seven-figure businesses as Clay-specialized consultants.

Clay had not just built a product. It had manufactured a labor market around the product's power users and then sold into that labor market. This is one of the most elegant go-to-market loops in recent startup history.

By December 2025, Clay crossed $100 million in ARR — scaling from $1M to $100M in two years after six years of foundational work. The company described itself, accurately, as an "eight-year overnight success."

THE AFTERMATH — AI, CRM, and Where Clay Fits the Plumbing

The GTM stack of 2025 looks nothing like the GTM stack of 2019. Outbound is not a numbers game anymore — it is an engineering discipline. The teams winning at pipeline generation are not the ones sending the most emails; they are the ones with the most signal fidelity, the most precise enrichment, and the tightest iteration loops between what they know about a prospect and what they say to them.

Clay sits at the center of this new stack as an orchestration layer. The inputs are: a list (from LinkedIn, a database, a CRM export, a web scrape), a set of enrichment providers, a set of AI reasoning steps, and a set of signal triggers (hiring, funding, tech stack changes, job postings). The outputs are: enriched records, personalized messaging, CRM updates, and segmented audiences for paid campaigns.

This last output — CRM updates — is where companies like Stacksync enter the picture. Clay enriches a contact record beautifully: you know the person's verified email, their job title, their company's revenue range, their tech stack, which cloud their company runs on, and what their likely pain points are based on their recent job postings. But that enriched record needs to live somewhere it can drive action across the entire revenue team. It needs to flow into Salesforce. Into HubSpot. Into whatever CRM the company uses as its system of record.

The problem is that moving enriched data from a tool like Clay into a CRM is not trivial. Data schemas differ. Object relationships are complex. Bidirectional sync — where changes in the CRM update the enrichment workflow and vice versa — requires infrastructure that neither Clay nor most CRMs provide natively. This is the exact seam that Stacksync is built to stitch.

Kareem's vision has always been that data enrichment should not be a one-time operation. You do not enrich a list once and call it done. Companies change. People change jobs. The contact who was a VP of Sales last quarter might be a Chief Revenue Officer today. The startup that was a 50-person Series A when you first enriched them might have raised a Series C and grown to 300 people. Enrichment, in Kareem's framing, is a living workflow — a continuous process of keeping your understanding of the market current. The CRM is where that living understanding must eventually land to be useful.

The full picture, then, is this: Clay generates the intelligence. Stacksync moves it into the system of record. The revenue team acts on it. The loop closes.

5 THINGS NOBODY KNOWS ABOUT CLAY

1. The waterfall is not just a feature — it is a philosophical inversion of the data industry's business model.

Every major data provider wants you to sign an annual contract for bulk access to their database. The business model assumes you will overpay — that you will buy coverage you do not need because the alternative is a patchwork of point solutions. Clay inverted this entirely. By buying data in bulk from multiple providers and passing the savings to customers on a per-successful-lookup basis, Clay became the largest buyer of B2B data on the market and passed that purchasing power downstream. The waterfall means you never pay for a miss. The industry hates this, which is why no incumbent has replicated it. The incentive structures are fundamentally at odds.

2. The "GTM Engineer" category was a product strategy disguised as a job description.

When Kareem and Varun Anand coined the term in 2023, they were not documenting a role that already existed at scale. They were naming an aspiration — a vision of what Clay's best users could become — and then building content, community, and internal function around that vision to make it real. By the time companies started posting GTM Engineer job listings, Clay had already built the playbook, the case studies, the community clubs, and the internal team that validated the category. The naming was the act of creation. The market followed the map.

3. Clay spent more than six years not growing fast before the two years it grew fast.

The company was founded in 2017. It did not hit $1M ARR until around 2023. Six years of building, pivoting, narrowing, and rebuilding — funded by patient capital from Sequoia and sustained by a team that had internalized the belief that the right idea just needed the right moment. The moment was AI democratization meeting the collapse of traditional outbound. But the infrastructure to capture that moment had been under construction for half a decade. The overnight success had a six-year foundation most people never saw.

4. Clay's AI integration made it a de-facto orchestration layer for the entire AI GTM stack.

When Clay added the ability to call Claude, GPT-4, and other models as enrichment steps inside a workflow, it became something its competitors were not positioned to become: a node in the AI reasoning chain, not just a data pipe. You can now build a Clay workflow where step one is a data lookup, step two is an AI classification, step three is an AI-generated personalization, and step four is a push to your CRM or email sequence — all without code, all traceable, all re-runnable as the underlying data changes. No other tool in the GTM stack sits at this intersection of structured data and unstructured AI reasoning at scale.

5. The community is not a marketing channel — it is the product's quality assurance layer.

The 70+ self-organized Clay Clubs globally, the hundreds of agencies, the thousands of workflow templates shared by power users — these are not just acquisition channels. They are a distributed R&D function. When a new use case emerges in the wild (GTM engineers using Clay to track job postings as a buying signal, for example), Clay's internal team sees it in the community before they would have discovered it through traditional product analytics. The community acts as an antenna for the next problem worth solving. This is why Clay's product roadmap consistently tracks reality rather than speculation — it is built by people who use the product to do their actual jobs, sharing what they build in public.

BY THE NUMBERS

THE STACKSYNC CONNECTION

Clay is where enrichment happens. Stacksync is where enriched data becomes operational.

A GTM engineer builds a Clay workflow that identifies high-fit accounts, enriches them across 12 providers, scores them with AI, and writes a personalized first line for each one. That is the intelligence layer. But the revenue team lives in Salesforce or HubSpot. The Account Executive needs to see the enriched firmographic data when they open a record. The CS team needs to see when a contact changed jobs. The pipeline review needs accurate headcount and revenue range data.

Getting Clay's output into the CRM — cleanly, bidirectionally, without data loss or field mapping chaos — is the integration problem. It is not a CRM problem and it is not a Clay problem. It is the seam between them. That seam is exactly where Stacksync operates: real-time, bidirectional sync between the tools where intelligence is generated and the systems where it drives action.

Kareem's insight that enrichment should be continuous — not a one-time data pull but a living workflow — implies a continuous sync layer beneath it. The intelligence is only as valuable as the infrastructure that moves it into the places where decisions get made.

Research compiled March 2026. Sources: Clay.com blog, Sequoia Capital portfolio pages, TechCrunch, company announcements.