Co-Founder Equity Splits: Frameworks That Survive Year Three

The conversation usually takes about an hour. Two people who like each other, in the early excitement of a new company, neither wanting to be the one who makes it awkward. Somebody says “fifty-fifty, obviously” and everyone is relieved.

Three years later one of them has been working eighty-hour weeks on no salary while the other kept a full-time job for the first year and now contributes on weekends. They own the same amount. Nobody has said anything, and that silence is now the largest problem in the company.

This guide covers how to actually decide a split, what genuinely deserves weight, the fixed-versus-dynamic debate, how to run the conversation itself, and what to do when it is already wrong.

1. Start Here: Vesting Matters More Than the Split

Before any framework, the single most important point.

An imperfect split with proper vesting largely self-corrects. A co-founder who leaves at month eighteen on a four-year schedule keeps a fraction of their stake and the rest returns to the company. The mistake is survivable.

A perfect split with no vesting is permanent. Somebody leaves after seven months owning half the company forever, and every investor who ever looks at your cap table sees a dead stake.

If you do only one thing, put founder stock on a standard four-year schedule with a one-year cliff and file the 83(b) election within thirty days. Our guide to founder vesting and acceleration covers the mechanics, and the thirty-day deadline is statutory with no relief for missing it.

With vesting in place, the split conversation carries far less risk, which is precisely why it should happen after you have agreed to vest rather than before.

2. What Actually Deserves Weight

Founders systematically over-weight some contributions and under-weight others.

Over-weighted: the idea

Whoever had the idea usually believes it deserves a large premium. It rarely does. Ideas are widely distributed, most change substantially, and the value is created in execution. A meaningful but modest premium is defensible; a majority stake on the basis of an idea alone is not, and it will cause resentment in the people who build it.

Under-weighted: opportunity cost

A founder who left a well-paid job to work full time for nothing is making a large, quantifiable investment. A founder who kept their salary for the first eighteen months is not. This is the difference founders are most reluctant to discuss and the one that most reliably causes damage later.

Legitimately significant

• Full-time commitment, and from when. The single strongest factor.
• Cash invested. Real money at genuine risk.
• Prior work that transferred in — existing intellectual property, a working product, an established customer base.
• Domain expertise or relationships that make the company possible.
• Who is taking the CEO role, which carries responsibility, exposure and usually the fundraising burden.
• Realistic replaceability. Uncomfortable but relevant: how hard would it be to hire someone to do this?

Not a factor

• Who is older, or who introduced whom
• Who registered the domain
• Anything an advisor or contractor did — those people get advisor grants, not founder equity

3. Fixed Splits Versus Dynamic Models

The fixed split

Agree percentages at formation, apply vesting, and leave it. Simple, clean, and understood by every investor and lawyer you will meet.

The weakness is that it is decided at the point of maximum ignorance. You are guessing about contributions nobody has made yet.

Dynamic equity models

Several frameworks allocate equity continuously in proportion to contributions as they occur — time, cash, IP, relationships — each converted into units, with the split recalculated as the company progresses.

The logic is sound and the practice is difficult. Someone must track everything honestly, disagreements about relative value shift from one conversation to a permanent argument, and — decisively — institutional investors are generally uncomfortable with a cap table that is still moving. They want a fixed, known capitalisation.

Where dynamic thinking genuinely helps is before formation. Run the model for six months informally, then convert to a fixed split with vesting once you know something.

A workable middle path

1. Work together for a period before deciding. Three to six months tells you more than any framework.
2. Then set a fixed split, informed by what you have actually observed.
3. Apply four-year vesting with a cliff, and give credit for time already served.
4. Reserve a small pool for adjustment — a few percent the board can grant later if contributions diverge sharply from expectation.

4. A Framework for the Conversation

If you want structure rather than negotiation by instinct, score each founder on weighted factors. A defensible starting set:

• Full-time commitment and start date — weight this most heavily
• Opportunity cost — salary and security given up
• Cash contributed
• Pre-existing IP or product
• Domain expertise and network
• Role and accountability
• Idea origination — include it, weight it lightly

Score independently, then compare. The value is not the arithmetic — it is that the discussion becomes about a shared model rather than about whether one of you is being greedy. Where scores diverge sharply, that divergence is the conversation you needed to have.

Do not treat the output as binding. Treat it as a starting point that makes the negotiation about factors instead of feelings.

5. How to Actually Have the Conversation

The framework is the easy part. The reason splits go wrong is almost never analytical — it is that two people who like each other avoid a discussion that feels like it could damage the relationship, and the avoidance does the damage instead.

Schedule it properly. Not at the end of a working session, not over drinks. Put two hours in the calendar with a stated purpose, so nobody is ambushed and both of you have thought about it. Treating it as a real meeting signals that it is a normal business decision rather than an accusation.

Both write your proposal down first, separately. Each person writes what they think the split should be and why, before either sees the other's. This prevents anchoring, and it surfaces the actual gap immediately. If you both wrote the same thing, the meeting is short. If you did not, you now know precisely where the disagreement is instead of circling it.

Talk about the future, not the past. Equity pays for what people will contribute over the next four years far more than for what happened in the last four months. Framing it forward makes it easier for a founder who has contributed less so far to argue honestly about what they are about to commit to — and it makes the commitment explicit, which is exactly what you want on the record.

Say the uncomfortable thing about full-time. “Are you going to leave your job, and when?” is the question that most determines the right answer and the one people most reliably avoid. Ask it directly and get a date. A co-founder who cannot give one is telling you something important.

Write down what you agreed, the same day. Not the legal documents — those come later — but a plain summary: the percentages, the reasoning, the commitments each person made, and what happens if those commitments change. Memories of this conversation diverge remarkably within a year, and a contemporaneous note prevents a great deal of it.

If you cannot have this conversation, that is the finding. Two people who cannot discuss equity candidly will not be able to discuss a missed forecast, a firing, or an acquisition offer. Better to learn it in month two than in year three.

6. Specific Situations

The part-time co-founder

Someone keeping their job “until we raise” is contributing meaningfully less risk and less time. Options: a smaller initial stake, a delayed vesting start, or a milestone-based increase when they go full time. What does not work is treating them as equal and hoping resentment does not build.

The late-joining co-founder

Someone joining a year in, when real progress exists, is taking substantially less risk. Between 5% and 20% is a common range depending on seniority and stage — not an equal share.

The technical-versus-commercial debate

An old and generally unproductive argument. Both are necessary; neither succeeds alone. The relevant questions are commitment, opportunity cost and replaceability, which are answerable, rather than which discipline is intrinsically more valuable, which is not.

Three or more founders

Equal splits get harder as numbers grow, because contribution variance widens. A four-way equal split among people contributing very differently is a common source of failure. Also consider control: no majority holder can complicate decisions, and investors will ask.

“But it was my idea”

Handle this directly rather than conceding to avoid conflict. A modest premium — a few percentage points — acknowledges origination. A majority stake on that basis alone will be questioned by every investor and resented by every co-builder.

7. Documenting It Properly

An agreed split that exists only in conversation is not an agreement.

• Founder stock purchase agreements, with each founder buying shares at a nominal price
• Vesting terms with the repurchase right, cliff and any acceleration
• 83(b) elections filed within thirty days, with proof of mailing retained
• Invention assignment agreements from every founder, covering work done before incorporation. Missing assignments are the most common serious diligence finding — see our guide to preparing for diligence.
• A founders' agreement covering roles, decision-making, what happens on departure, and how disputes are resolved
• Board consents approving the issuances

This is exactly the work a startup lawyer does cheaply at formation and expensively later — see our guide to choosing counsel.

8. When It Is Already Wrong

Most founders reading this have already split. If it no longer reflects reality:

• Say something. Silent resentment destroys more companies than unequal splits do. The conversation is uncomfortable once and corrosive indefinitely.
• Adjust forward, not backward. Trying to claw back issued shares is legally messy and relationally worse. Additional grants to the under-compensated founder are cleaner and are board-approvable.
• Use a financing as the moment. Investors frequently support rebalancing and can act as a neutral party. Refresh grants to founders are routine at Series A and B.
• If someone has disengaged, address it directly. A founder who has effectively stopped contributing but continues vesting is a problem that compounds. This is what the repurchase right exists for, and it must be exercised within its window.
• Get separate advice if it becomes contentious. Company counsel represents the company, not either of you individually — a distinction that becomes very sharp very quickly.

Frequently Asked Questions

Is 50/50 ever right?

Frequently, when two founders genuinely start together, commit equally and take equal risk. The problem is not equality — it is equality adopted to avoid a conversation. If you can each articulate why it is fair, it is fine.

Do investors care about the split?

They care about three things: that no founder has a demotivatingly small stake, that vesting is in place, and that there is no dead equity from a departed founder. The precise percentages matter far less than founders assume.

What about equity for the person who introduced us?

An advisor grant, not founder equity. Typically a fraction of a percent vesting over one to two years. Our guide to option pools covers advisor ranges.

Should the CEO get more?

Often a modest premium, reflecting accountability, exposure and the fundraising burden. A large premium purely for the title is harder to justify and is frequently resented. Some teams handle this with salary differentials instead once cash allows.

What if a co-founder wants to leave amicably?

Follow the documents. They keep vested shares, the company repurchases the unvested portion within the exercise window, and everyone signs a separation and release. Amicable departures still need paperwork — handshake arrangements resurface in diligence years later, at the worst possible moment.

What if we cannot agree at all?

Take it seriously rather than deferring it. A genuine deadlock about who contributed what is usually a proxy for a disagreement about roles, authority or commitment, and none of those improve with time. Options: bring in a neutral third party — an experienced founder or an advisor with no stake — to facilitate one session; agree a smaller gap now with an explicit review at a defined milestone; or conclude that this is not the right partnership. The last option is painful and is occasionally correct, and it is far cheaper before incorporation than after.

Should salary differences affect the split?

They should be considered together rather than separately, because they are two forms of the same compensation. A founder taking a salary while another takes nothing is receiving value the other is not, and that belongs in the equity conversation. The cleanest approach is to decide both at once: agree the split on long-run contribution, then use salary to correct short-run imbalances once there is cash. What causes trouble is deciding equity in month one and salary in month fourteen without reference to each other.

Do we need a founders' agreement if we have stock purchase agreements?

They do different jobs. The stock purchase agreements handle ownership, vesting and the repurchase right — the legal machinery. A founders' agreement handles the human questions those documents do not: who decides what, how disagreements get resolved, what counts as leaving, and what everyone owes the company in time and exclusivity. Plenty of companies operate without one. The ones that most needed it are the ones that discover the gap during a dispute.

How does a split work if one founder is in a different country?

The percentages are a separate question from the mechanics, and the mechanics get harder. Share issuance, vesting and the equivalent of an 83(b) election all have local tax consequences that differ by jurisdiction, and a structure that is efficient for a US founder may create a charge for a co-founder elsewhere. Take advice in both places before issuing anything, and if a US parent is likely, read our guide to the Delaware flip first — sequencing the structure before the issuance saves considerable cost.

The Bottom Line

Put vesting in place first; it is worth more than getting the split exactly right. Weight full-time commitment and opportunity cost heavily, weight the idea lightly, and have the conversation properly rather than defaulting to equal to avoid discomfort.

Then document it — stock purchase agreements, 83(b) elections, IP assignment — while it is cheap and everyone is still friendly.

Global Capital Network connects founders with investors and the counsel who help them start properly. See upcoming events or get in touch.

This article is general information, not legal or tax advice. Founder equity arrangements are jurisdiction-specific. Work with qualified startup counsel.