How should we split the equity?
The split is the second decision. This asks about the first one — vesting — before it shows you any percentages.
First: is there vesting?
Asked before the split because it matters more. A co-founder who leaves in eighteen months with no vesting keeps everything, contributes nothing, and blocks the next round.
Then: score each other, 0 to 10
Do this separately and compare. The value is in finding where you disagree, not in the number at the end.
| Factor | Weight | ||
|---|---|---|---|
Full-time commitment, going forward Who is actually doing this, and for how long. The most predictive single factor and the one most often scored lowest, because it is about the future rather than the past. | 5 | 5 | |
Responsibility for outcomes Who carries it when something goes wrong at two in the morning. Not seniority - accountability. | 5 | 5 | |
Skills the company cannot buy Genuinely scarce capability. Be strict: if you could hire it at market rate, it belongs in salary rather than equity. | 5 | 5 | |
Money put in Cash actually at risk. Consider a loan or a convertible instead - equity converts a one-off cheque into a permanent share of everything after it. | 5 | 5 | |
What each person gave up Salary forgone, job left, opportunity dropped. Real, and already in the past. | 5 | 5 | |
Whose idea it was Weighted lowest on purpose. The idea is the cheapest input in any company and weighting it heavily is the most common cause of a split people resent in year two. | 5 | 5 |
A structured way to find out whether you disagree, not a legal document. Get the agreement drafted by a lawyer before anyone relies on it.
The split is the second decision
Founding teams spend weeks on the percentages and an afternoon on the vesting, and it is precisely the wrong way round. The split allocates ownership of a company that does not exist yet, on the basis of things people have mostly already done. Vesting determines what happens when one of you leaves in eighteen months — which is the event that actually kills companies at this stage, and it is common enough that every investor you meet will ask about it before they ask about anything else.
The failure looks the same every time. Two people split it fifty-fifty, or sixty-forty after a careful conversation, and sign nothing about what happens if someone stops. One of them leaves after a year and a half, for entirely ordinary reasons. They keep the whole allocation. They now own a large share of a company they are not working on, and the next investor looks at that cap table and passes — not because the business is wrong but because the problem is unfixable without the departed founder’s agreement, and they have no reason to agree.
The other thing worth being suspicious of is the precision. Every calculator of this kind takes subjective scores on an invented scale and returns a number with a decimal point. That decimal point is doing real work: it makes the output feel like a measurement rather than a summary of opinions, and it makes it much harder for the person who comes off worse to push back. If a one-point change to one slider moves the answer by two points, the answer was never in the sliders — and the useful output of the exercise is not the percentage but the factor where your two scores are furthest apart.
Last, the principle that resolves most of the hard cases: equity pays for the future. Money already spent and work already done are debts, and a company can repay debts in cash once it has any. Paying for them in shares converts a finished contribution into a permanent claim on a decade of work that has not happened yet. Recognise it, price it, and pick an instrument that ends.
What this tool does not do
Every input is a subjective score on a scale this tool invented, weighted by numbers you chose. It cannot know anything about your founders and it is not measuring - it is organising opinions so the disagreement becomes visible. That is why the result is rounded to 5% and why the sensitivity is shown: where a one-point change moves the answer materially, the answer was never in the scores. The strongest use of it is for each founder to fill it in separately and compare.
- Drafting anything - the vesting schedule, the shareholders' agreement and the share issue are all legal documents and all belong with a lawyer
- Telling you whether the split is fair, which is not a computable property
- Acceleration on a change of control, good leaver and bad leaver terms, or what happens on a founder's death or incapacity
- Tax on founder shares, elections that must be made within days of issue, or anything that depends on where you or the company are
- Share classes, voting rights, board composition or any of the terms that actually determine control
- Advisors, early employees, or anyone who is not a founder
- Revisiting a split that already exists - it cannot tell you how to unwind one that is wrong
Frequently asked questions
Why does this ask about vesting before it shows me a split?
Because it is the more important decision by a wide margin, and putting it second is how it gets skipped. A co-founder who leaves after eighteen months with no vesting keeps every share they were allocated on day one — contributing nothing further, sitting on the cap table, and presenting the next investor with a problem they will decline to fund around. An equal split with four-year vesting survives that. A meticulously calculated split without vesting does not.
Why round to 5%? My other calculator gave me a decimal.
Because the inputs are subjective scores on an invented scale, and a decimal point implies a precision that does not exist. 53.7% is the same guess as 55% wearing a lab coat — and the lab coat makes it far harder for your co-founder to question, which is exactly why it is worth stripping off. The sensitivity figure shown beside each name tells you how far the result moves on a one-point change; where that is large, the split was never in the numbers.
Should we just split it equally?
Frequently, yes, and the tool says so when the spread falls inside the noise. Among founding teams where everyone is genuinely all-in, an equal split is the most common arrangement that survives, and it removes a conversation that otherwise resurfaces every time the company has a bad quarter. Where one founder is part-time or joining much later, that is a real difference and the numbers will show it.
One of us put in the starting capital. Shouldn't that count for a lot?
It should count, and there is usually a better instrument for it. Cash is the one contribution a company can actually repay. A loan or a convertible note recognises it and keeps it finite; equity converts a one-off cheque into a permanent share of everything the business ever does, including a decade of work somebody else has yet to do. Founders who funded the start rarely intend the second thing.
Whose idea it was barely moves the number. Is that right?
It is deliberate. The idea is the cheapest input in any company, it is rarely unique, and it changes beyond recognition within two years of anyone building it. Weighting it heavily is the single most reliable way to produce a split that whoever is doing the work resents by month eighteen — which is a more expensive outcome than whatever the founder with the idea gained from the extra points.