Bob McGrew's Operating Manual for Running the FDE Model
Over 100 YC startups are now hiring Forward Deployed Engineers, up from basically zero three years ago, and nobody can tell you exactly how the model actually works.
So we went through Bob McGrew’s 51-minute breakdown of the FDE playbook he ran as Palantir’s head of product and pulled out the full mechanics:
when to use it
who to hire
how field hacks become durable product
how to price for outcomes without sliding into consulting
McGrew was an early engineer at PayPal, an early executive at Palantir and most recently Chief Research Officer at OpenAI, where he led ChatGPT, GPT-4, and the o1 reasoning model. He was in the room when the FDE model was invented. His most surprising advice to founders copying it: don’t.
The whole thing is a trap that’s dangerously easy to turn into a low-margin consulting firm.
We previously broke down the first decade of this model in The Palantir Origin Playbook (2003–2013): how the CIA became an alpha customer, how Gotham was deployed through embedded teams and how software and consulting were fused from day one. This guide picks up where that history ended. Instead of retelling the origin story, it turns the model into an operating playbook for AI founders.
Inside you will find:
When the FDE Model Is a Moat and When It’s a Trap: The No-Incumbent Test
- How Palantir Lands Its First Customers: Build a Demo, Then Turn Services into Product Discovery
How Field Hacks Become Product: The Gravel Road, the Ontology, and the Loop That Keeps You Out of Consulting
Who to Hire: Domain Rebels and Prototypers, Not Loyalists and Craftsmen
Price the Outcome, Clear IT: How AI Startups Go Zero to Eight Figures in a Year
The Metrics That Prove It’s Working, and Why the Adoption Gap Is the Whole Opportunity
This issue is available exclusively to paid subscribers!
If you’re not subscribed yet, you can unlock this and our full library by upgrading below: Let’s begin!