Not based on who had the idea. Based on what each person brings from here — and it should be a real conversation, not a reflex 50/50.
12 min read · Level 1
Splitting on who had the idea values the wrong four years. Ideas are cheap and they change; the next eight thousand hours are what actually builds the company, and someone doing equal work for thirty percent will notice — not loudly, and not immediately, but permanently. A defensible split is one both people can explain out loud without flinching. If you cannot say the reason, it is not the reason.
The founder equity split is how ownership is divided at the start. It should reflect what each person contributes from this point forward — commitment, risk, and the work of the next four years.
It is not payment for the past.
Because the most common formula — 'it was my idea, so I take 70%' — values the wrong thing. Ideas are cheap and the next four years are not. A co-founder doing equal work for 30% will notice, and resentment compounds quietly until it is fatal.
And because an unequal split needs a reason both people accept out loud. An unexplained one poisons the partnership slowly.
Two founders. A had the idea and built a prototype over three months. B joins full-time from day one.
A proposes 70/30, on the basis of the idea and three months of work.
Count forward instead. Over the next four years both will work roughly 8,000 hours. A's head start is about 400 hours and an idea that will change substantially anyway. On a forward view, the contributions are close to equal.
A more defensible split: 55/45 — recognising the head start without pricing it as though the past outweighed the future.
Factors that legitimately justify a difference: who is taking more financial risk (leaving a job versus staying employed), who has committed full-time versus part-time, who brings capital, and who brings a decisive network or reputation. All of these are forward-looking. 'I thought of it' is not.
The default 50/50 with no conversation is also a mistake — it is chosen to avoid an awkward discussion rather than because it is right. Sometimes it is right; it should still be discussed.
The bigger mistake: splitting equity without vesting. A 50/50 split where one founder leaves in month four and keeps 50% for ever is the single most damaging thing you can do to a young company.
Third: not writing it down. Verbal equity agreements diverge in memory, always in the direction of the person remembering.
The next four years dwarf anything that happened before. Value the future contribution.
Financial risk taken, full-time versus part-time, capital contributed, decisive network. Not who spoke first.
Both propose a number and explain it. The explanations matter more than the numbers.
55/45 or 50/50 is fine. Splits like 52.5/47.5 signal a negotiation that should have been a conversation.
Four years, one-year cliff. Non-negotiable. See the next lesson.
A founders agreement, signed. Not an email, not a memory.
After the conversation and the trial project, before incorporating or issuing shares.
Do not decide this in the first week of knowing someone. It is one of the few genuinely hard-to-reverse decisions available to you.
The typical claim is three to six months of prior work and "the idea". Counted against four years of full-time effort from both parties, that head start is a few hundred hours out of roughly eight thousand each — and the idea will have changed substantially by month twelve anyway.
Count forward, not backward. A 55/45 that recognises a head start is defensible; a 70/30 that prices the past over the future rarely is.
A default 50/50 chosen to avoid an awkward conversation is a different thing from a 50/50 both founders argued for. The first one holds until the first serious imbalance in contribution; the second one holds because the reasoning was made explicit and can be revisited.
Both founders should propose a number and explain it. The explanations matter more than the numbers, and the gap between them is the real conversation.
Create a free account to read the rest and apply it to your own startup.