The 50/50 split is not a decision. It is a postponed decision. Two founders who cannot agree on relative contribution, relative risk, and relative future commitment choose 50/50 because it avoids a conversation they are not ready to have. The problem is that the conversation does not disappear. It resurfaces at the worst possible time: when one founder wants to leave, when an investor asks "who is really driving this," or when one person is working 80-hour weeks and the other has taken a part-time consulting gig on the side.
I have structured or reviewed equity arrangements for over 20 startups. The splits that survive five years share three properties. They reflect actual contribution, they account for future commitment, and they are written down with explicit cliff and vesting terms before the company takes on any external capital.
Why 50/50 Breaks
The most common failure mode is asymmetric contribution that neither co-founder wants to name. One founder quits their job and goes full-time. The other stays employed "until we get traction." Six months later, the full-time founder has built the product, closed the first three customers, and written the investor pitch. The part-time founder has contributed design feedback and a few introductions. Both own 50%.
This is not a hypothetical. I have seen this exact pattern at four companies. In three of them, the split was renegotiated under duress between months 9 and 18, with lawyers involved and at least one founder feeling cheated. In one, the company died because the renegotiation poisoned the relationship.
The solution is not to be ruthless about relative value. It is to have the structured conversation before the disparity develops.
The Framework: Four Dimensions of Equity Justification
I evaluate co-founder equity across four dimensions. Each gets a score from 1 to 5. The relative scores determine the starting split.
DIMENSION 1: Idea Origin
5 - Identified the problem, validated it, owns the IP
3 - Joined after validation, contributed to refinement
1 - Joined after the product was defined
DIMENSION 2: Capital at Risk
5 - Contributing $50k+ or equivalent in deferred salary
3 - Working at below-market rate, no cash in
1 - Market-rate salary from day one
DIMENSION 3: Full-Time Commitment
5 - Full-time from formation, no side income
3 - Full-time but with part-time consulting income
1 - Part-time or employed elsewhere
DIMENSION 4: Unique Capability
5 - The business cannot be built without this person's specific skill
3 - Strong contributor, replaceable with 3 to 6 months of hiring
1 - Generalist, easily replaced in 30 to 60 days
Add the scores for each co-founder. Convert to percentages. This is your starting point, not your final answer. The conversation about the scores is more valuable than the scores themselves.
A Real Example
Two co-founders, call them A and B. A is the technical founder who identified the problem, built the prototype, and is going full-time with no salary for 12 months. B is the commercial founder who joined after the prototype, is working full-time, and has a unique distribution network in the target industry.
Founder A Founder B
Idea Origin 5 2
Capital at Risk 5 3
Commitment 5 5
Unique Capability 4 5
TOTAL 19 15
Raw split: A gets 55.9%, B gets 44.1%. After the conversation, they might negotiate to 55/45 or 57/43 depending on how they value B's distribution network relative to A's technical risk. The point is that both founders can see the reasoning and argue about specific inputs rather than abstract fairness.
Cap Table Structure That Survives Dilution
The split is only the starting point. How you structure the cap table determines whether the split survives two or three rounds of dilution.
A realistic seed-stage cap table:
| Party | Shares | Pre-Seed % | Post-Seed % | Post-Series-A % |
|---|---|---|---|---|
| Founder A | 4,750,000 | 47.5% | 38.0% | 27.4% |
| Founder B | 3,750,000 | 37.5% | 30.0% | 21.6% |
| Option Pool | 1,500,000 | 15.0% | 12.0% | 8.6% |
| Seed Investors | 0 | 0% | 20.0% | 14.4% |
| Series A | 0 | 0% | 0% | 28.0% |
| TOTAL | 10,000,000 | 100% | 100% | 100% |
At Series A, Founder A has 27.4% and Founder B has 21.6%. These are still meaningful stakes. The mistake I see is founders who give away too much at seed, leaving both co-founders at sub-10% by Series B, which can destroy motivation and complicate future hiring of a CEO.
Vesting: The Mechanic That Makes the Split Real
The split percentage is meaningless without a vesting schedule. Vesting is the mechanism that ensures equity is earned over time, not granted on day one.
The standard Silicon Valley schedule is four years with a one-year cliff. This is fine as a default but worth customizing:
# Co-founder vesting parameters
standard:
cliff: 12 months # Nothing vests before month 12
total: 48 months # Full vest at month 48
schedule: monthly # Pro-rata monthly after cliff
modified_for_early_customers:
cliff: 6 months # Shorter cliff if one founder has prior customer relationships
acceleration: single-trigger # Full acceleration on acquisition
modified_for_technical_founder:
cliff: 12 months
total: 48 months
back_weighted: true # 25% vest at cliff, then 6.25% per quarter thereafter
Back-weighted schedules make sense when one co-founder is taking on disproportionate early risk and you want the vesting to reflect that the company's survival in months 1 to 12 depends heavily on their effort.
The Cliff Conversation You Need to Have
The most important conversation is the "what happens if one of us leaves" conversation. Have it before you incorporate, not after.
Three scenarios to discuss explicitly:
Scenario A: Co-founder leaves before cliff. Unvested shares are forfeit. This is standard and both parties should agree to it.
Scenario B: Co-founder leaves after cliff but before full vest. They keep vested shares only. This is also standard but you need to decide whether the company has a right of first refusal on those shares.
Scenario C: Co-founder is asked to leave. This is the uncomfortable one. You need a good-leaver/bad-leaver distinction. Good leaver (health, family, mutual agreement) keeps all vested shares. Bad leaver (misconduct, competitive departure) may forfeit a portion of vested shares depending on jurisdiction and agreement terms.
Not having this conversation means a judge will have it for you, at five-figure-per-day legal fees.
What I Got Wrong
At my second company, I agreed to equal vesting schedules for two co-founders with very different levels of early-stage risk. One of them took on the full operational burden in the first year. The other contributed part-time. The equal schedule became a source of resentment by month 14 that was visible to our first investors. We eventually restructured, but the restructuring cost us three months of management focus at exactly the wrong time.
The lesson: vesting terms should reflect not just ownership but the distribution of early-stage execution risk. Differentiated schedules are not a sign of distrust. They are a sign of precision.
When NOT to Use This Framework
This framework assumes two to four co-founders building a scalable product business. It does not apply well to: solo founders (obvious); consulting or agency businesses where equity often stays with a single operator; businesses where one founder is also the primary investor and the equity reflects investment not labor; or situations where one founder has a non-equity compensation arrangement (salary-only, revenue share) instead of equity.
The Boring Reality
No equity split survives serious conflict. If the co-founder relationship breaks down, lawyers and leverage replace frameworks. What frameworks do is reduce the probability and severity of that breakdown by making expectations explicit before emotion is involved. Sign the agreement, get it in the data room, and move on to building the product.