In the Room, Not on the Quota: Who Owns the FDE Promise

Ron Schmelzer's Forbes piece on September 19 has a line I've been chewing on all week. Antanas Bakšys, founder of Ace Waves, a startup in Kaunas, Lithuania, told him: "Our FDEs are part of a sales team." About a third of the company does forward deployed engineering. The engineers participate before a contract is signed, to show what the technology does inside the prospect's own environment rather than in a demo. The goal is a working system in the first month. And the contracts pay on successful customer service resolutions. Also in Kaunas, Ramūnas Zavistanavičius at Gritmind runs it the other way, small teams and workshops before anyone commits an engineer: "We don't want to give our key engineers to someone."

Now hold that next to the job data. Henley Wing Chiu at Bloomberry read 1,000 FDE postings and found that exactly zero percent are quota-carrying roles and only 8 percent mention on-target earnings. The top responsibility was working directly with customers, and in his words, "not pre-sales activities, or doing demos."

And Vinoo Ganesh, who ran a Palantir program that rotated about 250 engineers through FDE work, wrote at Latent Space on September 12 that at one industry dinner an FDE was "a sales engineer who joined 'the second call,'" and somewhere else "a quota-carrying rep who could write Python." His rule: "Point the function at sales and the incentive becomes closing the account in front of you."

So the seat is in the room before the signature and off the quota. That gap is where promises get made that nobody owns. Sales makes them, because that's the job. The FDE inherits them, because that's the job too. And nobody wrote down what was promised in terms the person who has to deliver it can read.

This is part 5 of The FDE Transition. Part 4 was the contract with engineering. This part is the contract with sales: scope, quota, credit, and the pre-contract handshake.

The short version

The promise is made before the FDE's accountability starts, so the FDE has to be in the room before the signature, without carrying the number. The instrument is a one-page handshake attached to the contract: every outcome in the proposal, sorted by the FDE into three columns, ships, builds, or no, with sales owning the number, the FDE owning the columns, and both names on the page. Chiu's zero percent is the right quota figure, so the FDE's variable pays on the customer's outcome and the renewal it produces, never on bookings, and never on an accelerator for deployments shipped, which is a quota with a lab coat on. Credit splits at the signature: the booking is the account executive's, the expansion is the FDE's. If you can afford a pre-contract week, the FDE who scopes is the FDE who lands. If you can't, run workshops first and don't call the workshop team FDEs.

The promise is made before anyone owns it

The handbook's test for the role is one line: a sales engineer's accountability ends at the signature, an FDE's begins there. I still think that's right. But the promise doesn't wait for the signature. It's made in the deck, on the second call, in the sentence the account executive says when the CIO asks "can it do X." By the time the FDE's accountability starts, the thing they're accountable for has already been defined by someone whose number is bookings.

I've run product next to sales for most of my career, and I've been in the room when it got promised. It isn't villainy. The AE is asked whether the product can do a thing, the true answer is "mostly, with some work," and "mostly" doesn't close. So the answer is yes. The gap between yes and mostly lands on whoever shows up after the signature.

Ganesh's line about pointing the function at sales is about reporting lines, and the org design chapter agrees with him: the FDE's manager should not carry a bookings number. But the incentive leaks in even when the FDE reports to product. If the FDE has no say before the signature, sales is writing the FDE's scope with no cost attached. Bakšys's answer is to put the FDE in the room. Zavistanavičius's answer is to keep the key engineers out and send a workshop team instead. Both are honest about where the engineer sits. Most companies I talk to have neither. They have an SE in the demo, an FDE at go-live, and a promise in between that nobody wrote down.

The handshake: one page, three columns, before the signature

Here's what I'd require before any contract with an FDE attached gets signed.

Take every outcome in the proposal. Not features, outcomes: the resolution rate, the workflow, the integration, the number the customer will judge the deployment on. The FDE sorts each one into three columns.

Ships. The product does it today, or it lives in the customer layer from part 4, prompts, rules, thresholds, connectors, and the FDE can build it on site in the first two weeks. The FDE puts a date on it.

Builds. It needs a harness change, a layer-two pull request with a price. The FDE writes the estimate and names the engineer who has to say yes. The date is a range.

No. It needs a core change, or it's something the product doesn't do and won't do this year. The column exists so the word can be written before the signature rather than discovered after it. Every "no" that gets sold anyway is a fork waiting to happen, and part 4 said what a fork is.

Sales owns the number. The FDE owns the three columns. Both names go on the page and it's attached to the contract. That's the whole handshake. It takes an afternoon, it kills the "mostly" problem, and it gives the landing note from part 3 something to be measured against two weeks in: what did we promise, and which column was it in.

Ace Waves does this without a form, because the engineer is in the environment before the contract. That's the strongest version, and it's expensive. The page is what you do when you can't send an engineer to every prospect, which is most companies.

Quota, credit, and the accelerator nobody should sign

Chiu's zero percent is the right number. Joe Schmidt at a16z said why in June 2025: "It's rare to find implementation leads that want to carry a quota, and that can also lead to counterproductive behavior. Instead, design incentive structures that allow services to be sold at cost." He also wrote that forward deployed teams "need to be closely aligned with account executives," and both sentences are true at once. Aligned, not incentivized the same way.

Perspective AI's comp report, the firm's own survey of 1,200 FDEs from May, says the market is drifting the other way, with FDEs "paid as a hybrid sales/engineering function, with net-new-revenue accelerators tied to deployments shipped." Hold the figure loosely. Watch the drift.

An accelerator on deployments shipped is a quota with a lab coat on. It pays the FDE to land more customers faster, which means sorting more outcomes into the ships column than belong there. The promise problem, moved one seat over. The org design chapter has the rule in one sentence: variable on the customer's outcome and the renewal or expansion it produces, never on bookings. If the company runs outcome pricing, the way Ace Waves does, the FDE's variable and the customer's invoice move on the same number, and the pay stub says whose promise it was.

Credit is the other half, and it's where sales leaders push back. If the FDE is in the room before the signature, doesn't the FDE deserve deal credit? My answer is no. The booking is the AE's, all of it. The FDE's credit is the expansion, the second workflow, the renewal, and it's paid on the outcome that earned it. Split the booking and you've built a second SE. Leave the FDE off the expansion and you've told them the part after the signature doesn't count, which is the part they signed up for.

Who gets the second call

Shyam Sankar's line on Palantir's August earnings call, via Matt Ashare at Channel Dive, was that only Palantir has FDEs and everyone else has "sparkling sales engineers." It's a good line and it's mostly marketing. It also points at the real test. The sparkling sales engineer joins the second call, sparkles, and leaves. The FDE joins the second call to write the three columns, and then lands the deployment.

That's the rule for the pre-contract week, if you run one. The FDE who scopes is the FDE who lands. Cap it at a week, not a pilot, and let the week produce two artifacts: the handshake page and the first twenty eval rows from the customer's own cases. Those rows are the ones part 4 wants attached to every harness PR and part 3 wants in the handback packet. The pre-contract week is where they come from.

Gritmind's version, workshops first and key engineers later, is the honest answer for a company that can't spare the engineer. Just don't call the workshop team FDEs. Ganesh's point stands: the engineers who mattered at Palantir, in his telling, weren't the ones who shipped the most for customers but the ones who came back and changed what got built. That's the loop, and a workshop team doesn't run it.

What this means in the seat, and at the top

If you're the FDE: ask for the proposal before the signature. Sort it. Send the three columns to the AE and to your manager with your name on them. If the deal closes with a "no" still in it, you've got the page, and the landing note two weeks in writes itself.

If you run sales: the handshake costs you an afternoon per late-stage deal and gives you a number that holds up at renewal. The FDE isn't there to slow the deal. They're there to make the yes true.

If you run the company: no quota, no accelerator on deployments shipped, and the FDE's name on the scope line of every contract they'll land. If sales owns the number and nobody owns the scope, the FDE owns it by default, after the fact, at the customer's site, with no cost attached. The economics chapter shows what that does to the margin.

One thing to try this week: take the last contract that closed with an FDE attached. List every outcome the proposal promised. Sort each into ships, builds, or no, with the FDE who landed it, no rounding up. Count the "no"s that were sold. That number is the promise nobody owned, and the handshake is how you make it zero next time.

Sources: Ron Schmelzer, "Why Forward-Deployed Engineers Are Suddenly In High Demand," Forbes (September 19, 2026, updated September 22), with Antanas Bakšys and Ramūnas Zavistanavičius quoted · Vinoo Ganesh, "The Rise of the Forward Deployed Engineer, and How To Do the Job Right," Latent Space (September 12, 2026) · Henley Wing Chiu, "What I learned analyzing 1K forward deployed engineer jobs," Bloomberry (November 18, 2025, updated January 25, 2026) · Joe Schmidt, "Trading Margin for Moat: Why the Forward Deployed Engineer Is the Hottest Job in Startups," Andreessen Horowitz (June 4, 2025) · Perspective AI, "The 2026 Forward Deployed Engineering Compensation Report: What 1,200 FDEs Earn" (May 21, 2026) · Matt Ashare, "Palantir touts forward deployed engineers, reports record revenue growth," Channel Dive (August 5, 2026), with Shyam Sankar quoted

Part of the running argument on AI Product Management: once building is cheap, the scarce work is deciding what gets built for everyone, and the handshake is where the promise gets a cost before it gets a signature.

Related answer: Who owns the promise between a forward deployed engineer and sales?