If this was forwarded to you, welcome to Built GTM. I’m Heath, a VP Revenue who has built and rebuilt go-to-market at Uber, Airwallex, and Notch. Every week I read the field, test what holds, and share the systems I build, what broke, and what I learned. Join the operators reading it.

[ ] TL;DR
  1. GTM Engineering is more than pipeline or outbound. Your Tuesday is six jobs. Your title is one.
  2. The job is the 20 percent: system design, AI context, workflow. AI does the 80. Most operators have it inverted.
  3. Solve · Stack · Split. The loop I run on the one workflow that costs the most.
  4. If you think GTM Engineering falls on one person, you are in for a rude awakening. Level yourself up.

My Tuesday last week, before lunch: I scored a list of accounts, rewrote a positioning line, debugged a routing rule, wrote a cold email, sat in a renewal call, and shipped a small tool to enrich a list without paying for another seat.

Six jobs. One title on my LinkedIn. The GTM engineer is seven jobs wearing one title. So is the VP of Sales, the Head of RevOps, and the CS lead. The lanes between those titles dissolved. The org chart has not caught up. The work already has.

The 20 percent isn’t the polish anymore

For years the rule was don’t let good be the enemy of great. So we chased great. We poured our hours into the last 20 percent, the polish, the perfect deck, the line nobody asked us to rewrite. The other 80 percent, the actual production, we ground out by hand and called it the job.

AI flipped that. It made good enough free. It does the 80 percent now. Which means the polish stopped being where the leverage is. The new 20 percent is upstream: the context you give the model and the structure you build so it can execute. That is the job now. Not the output. The system that produces it.

I learned this the slow way. I spent months blaming the model and the prompt for generic output. The real problem was that I had never structured the context the model needed. I was doing the 80 percent by hand and skipping the only 20 percent that mattered.

AI does the 80 percent. Your 20 percent is the context, not the polish.

How other operators are solving it

I read the field before I land on my own take. Three moves show up over and over, and the receipts are already in.

Bolt AI onto outbound · Mutiny, How are B2B sales teams using AI in 2026

TL;DR  Mutiny paired 100,000 outbound emails and found AI-written ones replied at 4.1 percent against 5.2 percent for human-written. More volume, lower reply.

My take  This is the receipt behind faster spam. AI is fine for the draft. It is death as the strategy. The lever was never more emails. It is targeting and timing, the 20 percent AI cannot do for you.

Open a GTM engineer req · Kyle Poyar, GTM Engineers: Hype or Reality?

TL;DR  Kyle Poyar called the GTM Engineer the most talked-about role of 2025, with OpenAI, Ramp, and Webflow hiring.

My take  The role is real. The reflex to fix GTM by hiring one is not. A req names a hire, not the workflow that is bleeding. Level the whole team up instead of outsourcing the muscle to a title.

Buy another tool · MarTech, Vibe coding is hollowing out the martech stack

TL;DR  Marketers are swapping SaaS point solutions for AI-built, vibe-coded apps, driving churn and consolidating the stack.

My take  Consolidation is right. Buying another tool for this week’s fire is the trap. Build the thin edge that is yours, buy the plumbing, and stop adding logos nobody opens.

None of them is crazy. They are the reflexes of a busy team. They just start in the wrong place: the tool or the headcount, instead of the work.

How I run it: Solve, Stack, Split

Here is the loop I run on the one workflow that costs the most, instead of hiring or buying my way around it.

01
Solve the problem.Get to the crux, not the symptom. “Reps are busy” is not it. Name the workflow, framed as work, not a headcount.
02
Stack the tech.Not a shopping trip. Give the workflow the right context: the tech that carries the data, the signals that say who matters, the brief in your voice. This is the new 20 percent.
03
Split the ownership.Split the drag from the judgment. The scoring, the enrichment, the first-draft prep goes to the system. The relationship and the decision stay with you. AI owns the 80, you own the 20.

Want the receipt that this is not one job? I read 50 GTM Engineer job descriptions and no two described the same job. The proof is in I Read 50 GTM Engineer Job Descriptions.

The move

Stop describing yourself by your title. Look at your last Tuesday and count the jobs. Then pick the one workflow that costs you the most and run it: Solve, Stack, Split. Do that one thing this week.

And do not wait for a hire to do it for you. GTM Engineering does not fall on one person. It is a muscle every operator on the team has to build now. If you think one req fixes it, you are in for a rude awakening. Level yourself up. The work is already yours.