Everett Berry (Head of GTM Eng @ Clay) - GTM Infrastructure

Our StackedGTM podcast deep-dive on GTM Engineering wouldn’t be complete without chatting with Clay - the company who invented the term. In fact, we decided to have that be our first episode.

We spoke with Everett Berry, Head of GTM Engineering at Clay to cut through the noise on what GTM engineering actually is, how to build the function, and where the role is headed as AI models get more capable.

Here’s what stood out from our discussion with Everett:

#1: The Definition Most Teams Are Missing

GTM engineers remove the technical constraints that stop companies from growing as fast as possible. That is Everett’s definition, and it reframes the role entirely. The bottleneck is almost never a lack of ideas. It is the process and technology blockers between idea and execution at scale, and the GTM engineer’s job is to compress that lag time as far as possible.

#2: Sequence the Build or Waste the Hire

Everett’s implementation order: data quality first, then automating broken existing processes, then net new plays. Most teams skip straight to net new and wonder why nothing lands. Starting with data, cleaner phone numbers, better firmographics, more accurate territory assignment, lets you show early results while the more ambitious build develops in parallel.

#3: Rep Ride-Alongs Are the Best Source of Net New Plays

At every high-performing org, a small number of reps are doing something unusual that lets them crush quota. Canva found reps manually screening customer social feeds for poor graphic design and pitching the CMO with screenshots. Clay helped turn that individual tactic into a signal running across hundreds of reps. The ideas are already in your org. The GTM engineer’s job is to find them and remove the human bottleneck.

#4: GTM Engineering Is Infrastructure, Not a Point Solution

Everett’s framing for Clay is GTM infrastructure: Audiences for high-volume TAM storage built on ClickHouse, Sequencer for multi-channel outreach, Agents for continuous account monitoring, and MCP connectors that let GTM engineers access Clay’s data directly from Claude or ChatGPT. The implication for buyers: purpose-built specialty platforms assume they can capture the uniqueness of your motion. Everett’s direct bet is that they cannot, and that the companies who win will build highly customized stacks on top of flexible infrastructure rather than configure their way into someone else’s playbook.

#5: The Failure Culture Has to Come Before the Hire

Everett does not expect 200-person GTM engineering orgs or VP titles. The leverage comes from AI model improvements, not headcount. But none of that leverage materializes if the culture punishes early failure. Clay’s own events invite system had empty chairs and reps who hated it for months before it consistently filled rooms with ICP. Most GTM organizations would have killed it in month one. The companies that get the most out of this function are the ones that give it enough iteration cycles to reach version four.

Listen to the full episode on Spotify or Apple for Everett’s hiring interview framework, the org design mistakes executives make when standing up these teams, and why the culture a leader builds before the first hire matters more than the hire itself.

And follow along as our podcast series on GTM engineering interviews other practitioners and vendors. At the conclusion, we will summarize all insights into an ebook.