I started Prelude Robotics in 2022 with two co-founders. By the time we closed our Series A in 2025 there were three of us, but not the same three. This is what I have learned from the two co-founder changes that happened in between, and from watching twenty other pre-seed-stage hardware teams navigate the same terrain.
The three archetypes that work
In hardware specifically, the founding teams that make it to Series A almost always include three distinct profiles — not three people with three distinct profiles, but three profiles filled by two or three humans.
- The operator-CEO — shipping focused, sales-capable, comfortable in a customer conversation at minute one
- The deep technical lead — usually the CTO, has shipped at least one product at scale before, credible with senior engineering hires
- The systems thinker — the person who owns the intersection of hardware, software and manufacturing, often under-recognized externally but essential internally
The two that quietly do not work
Two research co-founders without a commercial counterweight. We have watched this team three times in the last two years. The product deepens faster than the company can sell it. The cap table is usually even, the board is usually academic, and the company arrives at month 18 with a beautiful technology and no buyers.
Two operator co-founders without deep technical ownership. Less common but equally fatal. The company appears commercial for the first six months and then hits a technical wall it cannot climb without rebuilding the team.
The conversation I recommend every team have
Before your first institutional round, sit down with your co-founders and answer the following three questions aloud. Not on a slide — aloud, with notes, ideally with an outside advisor present.
- Who is the CEO in two years, and does everyone at the table genuinely agree?
- If one of us had to step back to a non-operating role within 24 months, who would it be and why?
- What is the single decision each of us would make differently from what the team has done so far?
The third question is the hardest because it surfaces the disagreements that have been building silently. It is also the one that has saved every founder partnership I have watched go from pre-seed to Series A intact.
On cap table and vesting
Even splits are a tell to experienced investors. Use a 55/30/15 or 60/25/15 split based on who is actually driving the company, and put all co-founders on four-year vesting with a one-year cliff. If you cannot have that conversation in month two you will not be able to have it in month twenty, and month twenty is usually when it matters most.
On the two changes we made at Prelude
Our first co-founder change was amicable — we realized at month ten that one of us was better off in a senior IC role, and we restructured the cap table accordingly with the support of our lead investor. Our second change was harder — a co-founder departure after eighteen months driven by irreconcilable disagreements on go-to-market strategy. Both changes required painful conversations. Both left the company stronger.
The common thread is that both conversations were made possible because we had been practicing smaller versions of them from the beginning. Founders who have never disagreed in public by month six will not know how to navigate a real disagreement in month twenty-four. The only solution to that is practice — and the first twelve months are the time to practice.