Clay raised a $100M Series C in August 2025 at a $3.1 billion valuation, two years after the company had effectively no revenue and a few months after it crossed $100M in ARR. The growth got the coverage. What caught my attention was that Clay did it while inventing a job category, one that OpenAI, Anthropic, Cursor, Rippling, Canva and more than 280 other companies are now hiring for.
The GTM Engineer.
Job postings for the role grew 205% year over year in 2025. Median pay sits around $160,000, roughly 20% above traditional sales and marketing operations roles, and seven bootcamps now teach it. The role exists because the old way of building revenue stopped working, and a handful of companies found a better one.
Fewer People, Faster Growth
Benchmarkit’s 2025 SaaS Performance Metrics report puts the median cost of acquiring $1.00 of new ARR at $2.00, up 14% from the prior year. The bottom quartile spends $2.82. Customer acquisition costs have climbed 222% over eight years, and CAC payback for private SaaS companies now averages 23 months, which means a new customer runs at a loss for nearly two years before contributing anything. Then there’s retention: 75% of software companies reported declines in 2024, which feeds the loop, because every point of lost retention has to be bought back with more acquisition spend.
The traditional response was to hire: more SDRs, more AEs, more coverage. That playbook ran into a new constraint when the Federal Reserve took rates from near zero to above 5% and capital efficiency became existential. Boards stopped asking how fast a company could grow and started asking how efficiently, and the answer they wanted, ARR growing faster than headcount, is one the traditional GTM structure cannot produce.
The 2024 SaaS Benchmarks Report from High Alpha and OpenView shows what the companies that could produce it looked like. Among companies under $5M ARR, median headcount fell 25 to 41% year over year while top-quartile growth rose from 150% to 250%. The median headcount at those top performers was seven people, down from twelve the year before. A seven-person company growing 250% a year is running on systems.
Revenue Is A Systems Problem
Clay coined the term in 2023 and their definition remains the clearest: a GTM Engineer builds revenue engines using AI and automation. The traditional GTM org treated revenue as a coverage problem, solved by adding reps and territories. GTM engineering treats it as an engineering problem, solved by building machinery that compounds, improves with each iteration and scales without proportional headcount. Varun Anand, Clay’s co-founder, describes the company’s own ambition the same way: “With our product, we’re taking the engineering discipline and bringing that to go-to-market.”
People compare the role to a sales engineer, and the accountability is where the comparison ends. A sales engineer supports someone else’s deal. A GTM Engineer owns pipeline, builds the systems that produce it and carries a number, and that ownership is what makes it a separate job.
One caveat before the examples. Clay coined the term, and the public case studies run almost entirely through Clay customers. The results below are documented, and they were also published by the company selling the tool that produced them, so read them with that in mind.
What The Role Looks Like
When Adam Wall joined Anthropic as its first Head of Sales Operations, the inbound interest form was a few weeks old and already producing a flood of leads, on top of signups arriving through Amazon Bedrock, Google Vertex AI and Anthropic’s own API. The default answer would have been a team of SDRs researching, scoring and routing by hand. Adam built an automated enrichment and scoring pipeline in Clay instead: 3x the enrichment coverage of the previous single-vendor setup, four hours a week saved on Salesforce automation alone, and access to more than 100 data providers without a separate contract for each. The lean sales team works inside Salesforce with clean, enriched data and spends its time selling. “Even if we enriched every lead, we wouldn’t have bandwidth to handle everyone,” Adam said. “So scoring became super important.” The pipeline decides what deserves a human before a human sees it.
OpenAI hit the same wall when ChatGPT Enterprise launched in 2023. Keith Jones, its GTM Systems Lead, described the team as “under-staffed for the demand we were facing,” with inbound enrichment running through a single provider and leaving large gaps. Clay’s multi-source model took coverage from 40% to over 80%. Individual sellers now run as many as 150 enrichments on a busy day, especially at the start of a quarter when they’re building pipeline, work that would have needed a dedicated research team a few years ago. The architecture matters more than the coverage number: RevOps improves the enrichment logic and waterfall prioritization behind the scenes without disrupting what sellers see, and the team is now looking at applying the same approach to data science and recruiting.
At Rippling, GTM engineering sits inside the growth teams, both the US team and Noah Adelstein’s international team, which run constant experiments including automated outbound and direct mail campaigns. Because the growth team could build and adjust workflows without pulling engineers off the product, it experimented more, and cold email performance doubled year over year in 2023. Putting a builder next to demand gen opens up experiments traditional marketers never get to run, programmatic landing pages and live intent harvesting among them, because they no longer wait on a ticket and a sprint. The GTM Engineer executes faster and, more importantly, widens the set of things it’s feasible to try.
What Happens To The SDR?
Emergence Capital’s survey of more than 560 B2B software companies found that 36% cut SDR/BDR headcount in 2025, the largest reduction of any sales role, ahead of account executives, customer success and sales management. Another 19% grew their SDR teams, and my guess is that group is seeing outsized returns from the people it kept. The version of the function being cut is the large team making hundreds of cold calls a day. The version being built is smaller, better equipped and pointed at the conversations that move deals.
The productivity numbers explain the reallocation. Reps spend roughly 70% of their time on non-selling work: data entry, prospecting research, coordination. Automating that work returns about 23 selling days per rep per year. GTM Engineers build the systems that absorb the 70%, and a smaller number of SDRs, with more skill and better tooling, take the conversations.
What’s The Comp?
Bloomberry’s analysis of 1,000 GTM Engineering postings puts the median lower, at $127,500, and both numbers hide the premium the best-funded companies pay. Vercel tops the list at $252,000, followed by OpenAI at $250,000, LILT AI at $221,500, Air at $208,500 and Ramp at $184,000. Squint, Tandem, Clay and Mux round out the top ten, all above $170,000, and AI companies and developer tool startups dominate the top of the range.
They pay it because one hire replaces several. One GTM Engineer can amplify 100 sellers by automating the research, signal tracking, data entry and messaging that otherwise eats their week. A $200K GTM Engineer can produce the pipeline impact of three to five traditional SDRs at similar total cost, and the systems improve over time while an SDR team churns and gets retrained. Betts Recruiting puts it more aggressively: a single GTM Engineer books more demos than a team of five SDRs, through automated workflows competitors can’t easily copy.
Who Does The Job?
Clay describes the ideal GTM Engineer as “half commercial thinker, half builder,” or less politely as “marketers who are bad engineers” and “engineers who care more about moving revenue than writing perfect code.” Technical fluency matters in a specific way. Familiarity with Clay, n8n, Zapier or Lovable helps. Python, SQL or TypeScript helps more. The signal Clay hires on is a willingness to learn by tinkering, what they call vibe coding. Spencer Chemtob, one of Clay’s own GTM Engineers, started as a product designer and learned SQL on the job by asking Claude. He got to the role by building.
Commercial instinct carries equal weight. GTM Engineers think about ROI constantly and judge every workflow by whether it helps someone close a deal. They carry quota or pipeline targets and get measured on business outcomes, which is the line between this job and a pure ops role measured on tickets closed.
Clay’s interview reflects that. Candidates get a fuzzy problem, something like “our trial-to-paid rate won’t budge,” and the strong ones start with the data: current performance, where the drop-off happens, what’s been tried. Good answers work both market levers, such as new intent data sources, and internal levers, such as slow SDR-to-AE routing. The red flags are tunnel vision on a single channel, ignoring obvious risks like deliverability or targeting fatigue, and needing full clarity before acting. Ambiguity is the working condition.
The breadth is also the job’s main hazard. A day can run from building enrichment flows in the morning to discovery calls in the afternoon to conversion analysis at night, and that range burns people out unless scope, tooling and targets are set deliberately.
Where The Model Works
It works best in high-velocity sales to technical buyers who expect fast, informed responses. That covers transactional B2B SaaS, product-led companies with rich usage data and expansion motions, technical products like developer tools and data infrastructure, and high-volume, low-ACV SMB businesses. It struggles where the sale is slow, regulated or built on trust. Pure enterprise deals with 18-month cycles and heavy executive involvement resist it, regulated industries constrain what can be automated, and strategic consulting or high-touch professional services run on rapport that has no automation equivalent. Complex services with heavy scoping and delivery need human judgment throughout.
There is also a ceiling. Past 50 or so reps, specialization tends to return. The model trades depth in one discipline for breadth across several, which is the wrong trade for an enterprise-only company with complex, high-touch deals and the right one for a startup trying to reach scale efficiently. Its best work is getting a company to 50 reps, which happens to be the hard part.
How To Structure The Team
Clay runs two distinct GTM engineering functions. Forward-deployed GTM Engineers work directly with customers, helping prospects evaluate Clay, scope implementations and build specific use cases; many came from RevOps or growth roles. Internal GTM Engineers build and maintain the company’s own operations: deal automation, QBR generation, partnership workflows. Their internal customers are the customer experience, marketing, growth and partnerships teams. The two groups could swap roles if they had to. What sorts people is temperament: engineering-minded people gravitate to internal infrastructure, and people who like customer conversations end up forward-deployed.
Rippling embeds the function in growth teams. Notion, Intercom and Canva are building hybrids, with embedded GTM Engineers sitting alongside RevOps and the internal AI team and working across sales, marketing and product. The common pattern is to incubate under RevOps and push people outward once the foundations hold. Practitioners give the same advice: start in ops and federate later, balance rapid prototyping with stable implementation, and treat data hygiene as non-negotiable, because every result in this piece traces back to clean, enriched, well-modeled data.
What To Do About It
For founders and revenue leaders, the capacity model (quota divided by average deal size equals reps needed) has stopped describing how revenue gets built. The number to manage instead is output per headcount dollar. GTM infrastructure is becoming a competitive moat, and budgets should say so. If I were making the first hire, it would report into RevOps, carry a pipeline number rather than a ticket queue, and spend its first quarter on data before automating anything. An enterprise-only or heavily regulated motion should expect less from the model and plan accordingly.
For people in RevOps, sales ops or growth, the skills that defined the job in 2020 are the ones being automated: process documentation, tool administration, manual optimization. The skills that matter now are systems design, AI orchestration and rapid experimentation, and all three are learnable. Revenue teams are picking up the habits of product teams: build once, iterate constantly, measure everything. The practical way in is the one Spencer Chemtob took, a fuzzy revenue problem, a tool you can build in, and the SQL learned along the way.
For SDRs, the volume model is the thing being cut, and the 30% of the week spent in actual conversations is the part companies are keeping. The move is toward one side of the new split or the other: the conversations, or the systems that feed them.
The infrastructure for learning this is arriving faster than the postings.
How Big is This?
There is one GTM Engineering posting for every five RevTech and MarTech postings, one for every 14 in RevOps and one for every 92 SDR postings. The growth rate matters more than the level: ZoomInfo data shows hiring doubling year over year for two years running, with more than 3,000 listings in January 2026. Clay’s certification program has more than 2,500 alumni. Several customers have started GTM engineering agencies, Claygencies in Clay’s vocabulary, and scaled them past $1M in ARR in under a year. Clay projects $50M in 2025 revenue for its data and integration partners alone, and its network of 108 agencies generates hundreds of millions collectively.
There are now more than 60 Clay clubs worldwide, from Warsaw to Manila. At Clay Club Lahore, doctors and software engineers are picking up the skills as a path to economic opportunity. In Ukraine, organizers held their first meetup during the war and secured bomb shelter access for the 50-plus people who came to learn a job that had no name three years ago.
✌🏽SR
Related Reading:
