Lesson 6.5 · The team shape
Where this gets you: you’ll be able to walk into any FDE engagement, recognize the company shape you’re working with, and adjust your cadence, conventions, and politics to fit.
The idea
Solo, small, big. Each one is a different engagement. The same Forward Deployed Engineer who’s effective in one can stall in another. The team shape is the size and process of the customer you’re embedded in, and reading it right on day one is most of the job.
Solo founder or very small team (1–3 people). You move fast. Nobody’s asking permission. Your decision-maker is a founder, and you talk to them daily. No SOC2, no procurement, no legal review. You’re a temporary cofounder. The risks: founder burnout, everything depending on you, scope creep nobody pushes back on. Agents change the wedge — one person becomes five, with discipline. Without discipline, one becomes half.
Small company or scaleup (10–50 people). Some process, mostly still fast. The engineering leader is reachable, maybe weekly. Real users, real revenue, real consequences if you break something. The stack has been hardened by reality, which is harder than greenfield and more durable. The risks: priorities shift mid-engagement, the team you’re embedded with isn’t always who decides, the customer’s roadmap collides with yours. Agents change the wedge — leverage compounds. What you ship in week 4 multiplies what the team ships in month 4.
Big company or enterprise (500+ people). Lots of process. Security review, procurement, legal, change management. The decision-maker sits several layers above the engineers you work with. Slow loops. Real distribution and real capital are the prize, which is why anyone puts up with the rest. The risks: politics, mismatched incentives, a project that lives or dies on org dynamics you don’t control. Agents change the wedge — individual productivity gains are huge, but org-wide impact means changing incentives, which is a different game.
What getting this wrong looks like. An FDE spends a year at a three-person startup, shipping to production twice a day. She lands at a bank and does what worked: pushes a small fix on day three. It’s correct, it’s tested, and it bypasses change management. Security freezes her account. Six weeks of trust to rebuild, and nothing wrong with the code. Same engineer, same discipline, wrong shape.
How to adjust:
- Solo shop: ship every day, retro every Friday, write everything down because nobody else will. Build the Lesson 6.3 conventions from day one even though it feels like overkill. It isn’t — you’re the redundancy.
- Small company: the engineering leader is your champion, so make them look good. Document decisions obsessively; priorities will shift and you’ll defend choices three months later. Evals are your political shield. When the debate starts, point at the eval suite.
- Big company: find the executive sponsor in week one. Spend more time than feels natural on documentation, security review, and procurement. Use the agentic discipline to deliver inside the process, not around it, or you’ll get bounced.
Your exercise
Pick a real FDE engagement you’ve seen or want to do. Identify which shape it is. Write down three things you’d do differently from the other two shapes.
You’re done when you can defend the three differences in a one-paragraph email to a customer’s lead engineer.
Practice proof: save your shape analysis in NOTES.md under “team shape.”
Build on it: build a one-page intake questionnaire that asks eight questions about a customer and prints the shape plus its three adjustments.
Why this matters
Most engineers walk into every engagement the same way. The ones who win adjust cadence, conventions, and politics to the shape they’re actually in. The customer never sees the adjustment. What they see is that everything goes smoothly. That’s you, reading the shape right on day one.