FDE services: product moat or consultancy in disguise? The test is whether field work flows back into the product
An October 18 FDE Times analysis contrasts traditional services firms' 30-40% gross margins with Palantir's 80%+ and argues the difference comes from a feedback loop that turns one customer's deployment into the next customer's product, which is what FDE candidates should probe in interviews.
FDE Times looks at a question that has grown louder since OpenAI set up the OpenAI Deployment Company, a separate subsidiary with more than $4bn in committed capital for placing engineers inside enterprises: is FDE work building a software product, or is it consulting sold under a new name? Citing investment fund N47, it notes that traditional services firms run at 30-40% gross margins, while Palantir, the company most associated with the FDE model, keeps margins above 80% despite heavy field deployment.
The piece argues the gap is not about how many engineers a company sends to customers. Drawing on a16z, TSIA, KDnuggets and Valletta Software, it says customization pays off only when one-off deployments are turned into reusable templates, as happened at ServiceNow and Workday, whose margins rose from 63.2% and 54.1% at IPO to 79% and 75% by 2024. If nobody owns feeding customer-specific work back into the product, the article says, "FDE" is just a label. It also notes Gergely Orosz's observation that many new FDE roles sit in semi-independent units away from where the product is built, though a separate entity does not automatically break the loop.
Why it matters for FDEs: the same title can lead to a product-engineering career or a consultant's one. The article suggests asking two questions in interviews: who on the product team reads what FDEs learn in the field, and how often; and who, by name, is responsible for turning customer work into product features. It also recommends writing CV bullets about components built for one customer that others later reused.
Source: fdetimes.net