The conversation you don't have in week one is the one that ends the company in year two.
Two co-founders who trust each other completely still need a co-founder agreement. The trust isn't the issue. The issue is that trust doesn't resolve what happens when one person gets sick, when one gets acquired by a former employer, when a potential acquirer surfaces in year three, when one person decides they want to pivot hard into a direction the other finds wrong. Those situations aren't tests of trust — they're tests of what you both agreed to in writing.
What a co-founder agreement actually covers
The equity split is the part that gets discussed. It's also the easy part. Most founding pairs can work out whether a 50/50 or 60/40 split reflects their relative contributions. The harder conversations are the ones nobody initiates until they become urgent:
Who has final authority when co-founders disagree? On product direction? On a key hire the other person vetoes? On whether to take a term sheet?
What happens to equity when one founder leaves in month six — or month eighteen? Does it vest over time, or does the departing founder walk with the full grant?
What intellectual property did each person bring to the company versus what they create for it? If one co-founder built a prototype before incorporation, whose IP is it?
What does each person's committed time look like? What happens if one co-founder returns to a full-time job?
These aren't pessimistic questions. They're structural ones. The answers don't require the relationship to be adversarial — they require it to be defined.
Vesting mechanics — the clause that protects both parties
A four-year vesting schedule with a one-year cliff is the right default for co-founder equity. The cliff means a founder who leaves in the first year takes no equity. After the cliff, shares vest monthly. The remaining founder keeps their shares, the cap table stays clean, and the company doesn't end up with a meaningful stakeholder who stopped contributing in month fourteen.
The one-year cliff protects both sides. It gives the founding pair enough time to discover whether they work well together before either person is locked in. The cliff is uncomfortable to negotiate because it acknowledges that things might not work — but that acknowledgment is the point. A founding relationship that can't survive the conversation about what happens if it fails isn't ready to start.
Standard vesting schedules can include double-trigger acceleration — meaning that if the company gets acquired and the acquiring company terminates you within a year, your remaining unvested shares vest immediately. Get this in writing before the scenario exists.
IP assignment: what trips up seed-round diligence
Every co-founder needs to assign intellectual property they create for the company to the company — including relevant work they created before incorporation. If one co-founder built a prototype on nights and weekends three months before the entity existed, that IP needs to be formally assigned to the company.
Investors' counsel finds this in diligence. A clean IP assignment on day one takes two hours. A messy one discovered mid-seed-round takes two weeks, costs money, and sometimes costs the deal.
The IP assignment clause should also cover work co-founders create on their own time that's related to the company's business. The standard position: it belongs to the company. The exception is prior work that clearly predates the company and isn't related to the product you're building. Get specific in the agreement about where the line sits — vague language gets interpreted by lawyers at the worst possible moment.
Decision rights: get specific before you need to
A co-founder agreement that says "co-founders make decisions jointly" fails before it's signed. Joint decision-making with no tie-breaking mechanism creates gridlock.
Define which domains each co-founder owns. If one runs product and one runs go-to-market, each needs final authority in their domain. The CEO title doesn't solve this — "authority" has to be explicit about what it means in practice. Define the categories that require both to agree: major pivots, acquisition conversations, capital raises, executive hires. Define the categories that belong to each person individually.
The failure mode is that decisions drift into informal consensus territory where neither person owns anything. Resentment builds when one co-founder needs sign-off on everything, or when the other keeps second-guessing decisions in their domain. Define it early and you prevent the dynamic entirely.
The departure clause most founders write wrong
The most important clause in a co-founder agreement is the one that covers what happens when the partnership doesn't work.
Most co-founder agreements address vesting — the departing founder's unvested equity returns to the company. That's necessary but not sufficient. The clause should also define whether the company holds a right of first refusal on the departing founder's vested shares; what buyout mechanisms exist and how a fair price gets determined; and under what conditions a co-founder can be removed from an operating role while retaining equity.
A departing founder who still owns 25% of the company, sits on the board, and has no defined path to resolution can block a fundraise, an acquisition, or a key hire. That's a solvable problem on day one. It becomes an expensive, relationship-destroying problem on day 600.
A startup-focused lawyer can draft the full agreement in five to eight pages for $1,500–$3,000. That's the highest-ROI legal spend you'll make before your first product ships. Use a startup-focused lawyer — not a general commercial attorney, and not a friend who passed the bar. Someone who has seen what goes wrong.
If you've built a founding team and you're working through the structural side of the business, start here before you write a line of code.
Building a company in a vertical you know? Tell us about it. We co-found with operators who have the domain expertise to know what to build — and help structure the founding team from day one.