Cofounder Agreement Template: 12 Decisions to Make Before You Build

A cofounder agreement is the written answer to a simple question: what deal are we actually making with each other?

The useful version is not a 16-page template downloaded five minutes before signatures. It is the output of a direct conversation about ownership, work, control, money, intellectual property, and the ways the relationship could change. The document matters because it preserves those decisions. The decisions matter because they expose disagreements while the company is still cheap to change.

This guide gives you the founder-side template: the 12 decisions to make, the evidence to bring into the room, and the document stack your lawyer should produce afterward. It is part of our practical Starting Up library. It is general US-oriented educational information, not legal or tax advice. Entity law, employment restrictions, securities rules, marital-property rules, and tax treatment vary by founder and jurisdiction. Use this to arrive at counsel prepared—not to replace counsel.

The short answer: what belongs in a cofounder agreement?

A complete founder agreement should answer 12 questions:

Decision	The question you must answer	Where it usually lands after formation
1. Parties and scope	Who is a founder, what venture is covered, and when does the agreement start?	Founder memo, formation documents, cap table
2. Goals and commitment	What is each founder trying to build, on what timeline, and with how much time?	Founder agreement, employment or service documents
3. Equity	Who owns what, and why?	Charter or operating agreement, stock ledger, cap table
4. Vesting	How is founder ownership earned, and what can the company repurchase?	Restricted stock purchase agreements, board consent
5. Roles and pay	Who owns each operating domain, and when do salaries begin?	Bylaws or operating agreement, board consent, employment documents
6. Decisions	What can one founder decide, and what requires joint approval?	Bylaws, operating agreement, board and stockholder consents
7. Deadlock	What happens when the required decision-makers cannot agree?	Founder agreement, bylaws or operating agreement
8. Intellectual property	What existing and future IP goes to the company?	PIIA/CIIAA, IP assignment, contractor agreements
9. Money and assets	Is founder cash equity, a loan, or a reimbursable expense?	Stock purchase records, promissory note, expense policy, books
10. Outside work and confidentiality	What may founders do elsewhere, and what must remain protected?	PIIA/CIIAA, employment or service agreement
11. Departure and removal	What happens to equity, titles, board seats, access, and work when someone leaves?	Stock agreements, bylaws, board consents, separation documents
12. Sale, shutdown, and amendments	Who can approve a sale or dissolution, and how can the deal change?	Charter, bylaws or operating agreement, voting agreements

That last column is the part most templates blur. A pre-incorporation founder agreement can record the deal, but once the company exists, its terms need to match the entity’s actual governing and equity documents. A sentence saying “Alex owns 40%” does not itself issue stock, approve a grant, update a ledger, assign code, or file a tax election.

The Penn Carey Law Entrepreneurship Legal Clinic’s founders-agreement materials make the same foundational point: the agreement is the product of conversations that should happen early. Start there.

When to make the agreement

Do it before the facts become expensive.

The cleanest moment is after the founders have worked together enough to evaluate the relationship, but before the team has accumulated significant code, design, customer promises, cash, or informal equity expectations. If you are still testing whether the partnership works, run a small project together first. Our validation playbook applies to cofounder fit too: observe behavior under a real deadline instead of relying on enthusiasm.

Use this sequence:

1. Run a real working test. Ship a prototype, conduct customer interviews, or complete a short sprint together.
2. Hold the founder alignment meeting. Decide the 12 issues in this guide without drafting contract language in the room.
3. Choose the entity path. The correct documents differ for an LLC and a corporation; use our LLC vs. C corporation guide to understand the tradeoffs.
4. Send a written deal memo to counsel. Give the lawyer decisions, context, and unresolved questions—not a mystery template with blanks filled in.
5. Execute the complete stack. Formation, equity issuance, vesting, IP assignments, consents, tax filings, and cap-table updates should close as one coordinated process.

If you have already been building for months, do not wait for a cleaner moment. Create an inventory of contributions and IP, write down where the founders disagree, and bring both to counsel. The longer ambiguity sits, the more each person builds a different private history of what was promised.

Decision 1: define the parties and the venture

Write down every person who believes they are a founder. That can include the obvious full-time team, but also someone who contributed the original prototype, domain, customer list, patent work, or early cash and was promised a future role.

For each person, record:

• legal name and contact details;
• current employer and any relevant invention-assignment obligations;
• proposed founder role;
• full-time, part-time, or future start date;
• cash, property, relationships, data, code, designs, or other assets contributed;
• prior work that remains personally owned;
• agreements already made in email, chat, decks, or cap-table tools.

Then define the venture narrowly enough to distinguish company work from unrelated prior projects. “Any software idea we ever discuss” is not a useful scope. Describe the product, customer, problem, and relevant technology or market.

This is also where you identify contributors who are not founders. Advisors, agencies, freelancers, and early employees need appropriate service and IP documents, not honorary founder status used as a substitute for deciding what the relationship is.

Decision 2: align on goals, risk, and time commitment

A contract cannot repair incompatible goals. Put the differences on the table before negotiating percentages.

Each founder should answer, separately and in writing:

• Are we building for venture-scale growth, a durable bootstrapped business, or an experiment?
• What outcome would make this worth five years?
• When does each person become full-time?
• How long can each person work without salary?
• What personal cash can each person risk?
• Are relocation, travel, or regulated-industry constraints likely?
• What would make each person leave?
• May a founder consult, study, maintain open-source projects, or run another business?
• What does “working hard” mean in observable terms?

Do not turn lifestyle differences into moral judgments. A founder supporting a family may have a different cash threshold than a founder with savings. The agreement’s job is to make those constraints visible and design around them.

Decision 3: choose the equity split—and write the rationale

A 50/50 split can be right. It can also be an avoidance mechanism.

Evaluate equity against expected future contribution, not only who had the idea. Useful inputs include:

• full-time start date and expected time commitment;
• role scope and responsibility;
• relevant product, technical, distribution, or regulatory work already completed;
• cash or property contributed;
• IP the company needs and can legally receive;
• opportunity cost and foregone compensation;
• difficulty of replacing the contribution;
• continuing obligations after launch.

Avoid converting every input into a fake-precision score. The purpose of a framework is to surface assumptions. If one founder believes the original idea is half the company’s value and the other believes execution is everything, a spreadsheet average will hide the disagreement rather than resolve it.

Record a one-paragraph rationale with the split. Six months later, memory will rewrite the negotiation. A rationale gives the team something concrete to revisit if a promised full-time start never happens or a major contribution changes.

Also separate three concepts founders routinely collapse:

• Economic ownership: who participates in the value of the company.
• Voting power: who can approve corporate actions.
• Operating authority: who makes day-to-day decisions in a domain.

They do not need to be identical. Equal economics does not require every product, hiring, or spending decision to need two signatures.

Decision 4: design founder vesting, repurchase, and acceleration

Founder vesting protects the company from dead equity: a large ownership block held by someone who leaves before doing the work the grant assumed.

A common venture-backed startup pattern is four years of vesting with a one-year cliff, then monthly vesting. The Penn clinic’s annotated model uses that structure and explains the company-return mechanics for unvested interests. Treat it as a convention, not a law.

Your actual decision has at least seven parts:

1. Total term. Often four years, but context matters.
2. Cliff. Often one year; decide whether zero vests before that date.
3. Vesting start date. Formation date, service start, or an earlier date that credits meaningful prior work.
4. Frequency. Usually monthly after the cliff.
5. Company repurchase right. Define what happens to unvested shares when service ends and at what price.
6. Treatment of vested shares. Usually retained, but transfer restrictions and company rights may still apply.
7. Acceleration. Decide whether an acquisition alone accelerates vesting (single trigger) or whether acquisition plus a qualifying termination is required (double trigger).

Do not write “four years, one-year cliff” and assume the rest will sort itself out. The repurchase language and the date from which vesting runs determine the real economics.

The 83(b) deadline belongs on the closing checklist

When a founder receives substantially nonvested stock for services, a Section 83(b) election may allow the founder to include the current spread between fair market value and purchase price in income at transfer instead of waiting until the property vests. That choice has real tax consequences and is not automatically right for every security or founder.

What is not flexible is the filing window. The IRS’s current Form 15620 and instructions state that an 83(b) election must be filed no later than 30 days after the property is transferred. The instructions also require a copy for the service recipient and, when different, the transferee.

Operationally:

• identify at closing whether an election is relevant;
• confirm the transfer date from the executed documents;
• have tax or startup counsel review the election;
• file within the IRS window using the current instructions;
• retain proof of timely filing and the required copies with the closing records.

Do not bury this in a “tax tasks later” list. Later may be too late.

Decision 5: assign roles, time, compensation, and expenses

Titles are weak. Outcomes and authority are useful.

For each founder, define:

• primary domain: product, engineering, sales, operations, finance, or another area;
• the outcomes they own for the next 6–12 months;
• what they may decide without a vote;
• expected time commitment and full-time start date;
• who evaluates whether the commitment is being met;
• salary now, salary trigger later, and who approves changes;
• expense limits and reimbursement rules;
• reporting cadence and the metrics each founder maintains.

“CTO” does not answer whether that founder owns security, infrastructure spend, product delivery, hiring engineers, or technical sales. Write the operating boundary.

Founder salaries should be governed, not improvised. Agree whether pay begins at revenue, financing, a cash threshold, or board approval. If one founder takes salary while another defers it, decide whether the difference is simply compensation timing, a loan, or something else. Do not let unpaid amounts silently become assumed extra equity.

Decision 6: separate domain authority from reserved decisions

A startup cannot vote on every action. It also should not let one founder unilaterally issue equity or sell the company.

Create two lists.

Domain decisions belong to the accountable founder within an approved plan. Examples:

• product sequencing within the agreed roadmap;
• vendor selection below a spending limit;
• routine customer terms within an approved policy;
• hiring within an approved headcount plan;
• technical architecture within security and budget constraints.

Reserved decisions require a specified higher approval. Common candidates include:

• issuing equity or changing the cap table;
• raising debt or equity financing;
• changing founder compensation;
• hiring or firing an executive;
• approving spending above a threshold;
• entering a new line of business;
• selling material IP;
• changing the charter or operating agreement;
• selling, merging, or dissolving the company;
• entering a related-party transaction.

For every reserved matter, specify who approves and by what threshold. “Mutual agreement” fails as soon as the board changes or a third founder joins.

Decision 7: build a deadlock ladder before you need one

Two founders with equal votes can block each other indefinitely. The solution is not necessarily giving one person total control. It is designing a sequence that distinguishes a hard conversation from a company-ending dispute.

A practical ladder might be:

1. written statement of the decision, available options, evidence, and deadline;
2. a dedicated founder meeting after a cooling-off period;
3. input from a mutually trusted independent director or adviser;
4. mediation for disputes suited to mediation;
5. a defined governance, buyout, or separation process drafted by counsel.

Do not casually paste a “shotgun” buy-sell clause into a template. A mechanism where one founder names a price and the other must buy or sell can favor the founder with more liquidity, even if both own the same percentage. Deadlock provisions need to fit the cap table, financing plans, jurisdiction, and founders’ financial reality.

Also distinguish business judgment from breach. Disagreeing about the roadmap is different from diverting company funds, withholding credentials, or violating an IP obligation. The escalation path and remedies should reflect that difference.

Decision 8: make the company own the intellectual property it needs

Investors and acquirers will ask a basic diligence question: does the company own the product it is selling?

Create an IP schedule with four buckets:

1. Company IP already created: code, designs, data sets, content, domains, trademarks, inventions, customer materials.
2. Prior founder IP contributed to the company: list it and document the transfer or license.
3. Prior founder IP excluded from the company: identify it clearly so the boundary is visible.
4. Third-party IP: open-source software, employer-owned work, university rights, agency deliverables, contractor work, licensed data, fonts, images, and APIs.

Use actual assignment language. Cooley’s guidance on confidential-information and inventions-assignment agreements stresses a present assignment—language that assigns rights now—paired with a promise to assign future rights, plus a prior-inventions carve-out. See Cooley GO’s CIIAA overview.

The contractor “work made for hire” trap

Do not assume that paying a freelancer means the company owns the copyright.

The U.S. Copyright Office explains that commissioned work qualifies as a work made for hire only if it fits one of nine statutory categories and the parties expressly agree in a signed writing. Read Circular 30. A contractor agreement should therefore include a legally effective assignment of relevant rights, not rely only on the phrase “work made for hire.”

For technical founders, run a repository audit:

• Which accounts own the organization and repositories?
• Who made the first commits, and under which employment obligations?
• Did any agency or freelancer contribute?
• Are contributor agreements or assignments signed?
• Which open-source licenses apply?
• Do personal projects or reusable libraries overlap with the product?
• Are production credentials, domains, and cloud accounts controlled by the company?

An IP clause cannot transfer rights a founder never owned. Resolve prior-employer, university, or third-party claims before promising clean ownership.

Decision 9: classify founder cash, property, and expenses

Every contribution should have a label.

If a founder wires $25,000, is it:

• payment for founder shares;
• an additional capital contribution;
• a loan the company must repay;
• an expense advanced for reimbursement;
• or a gift with no repayment or ownership effect?

Do not decide after the money is spent.

Maintain a contribution schedule with the date, amount or asset, agreed treatment, supporting receipt or transfer document, and required approval. For non-cash property—equipment, domains, data, patents, or software—record whether the company owns it or merely has permission to use it.

Agree on future funding too. Are founders required to contribute more cash? What happens if one cannot? Does new money dilute everyone through a formally approved financing, become debt, or have no automatic equity effect? A founder’s emergency wire should not rewrite the cap table through implication.

Decision 10: set boundaries for outside work, confidentiality, and conflicts

Founders often have consulting clients, open-source projects, teaching, investments, or prior businesses. Blanket language can be both unrealistic and legally fragile.

Define:

• permitted outside projects and time limits;
• excluded prior inventions;
• use of company devices, data, staff, and confidential information;
• disclosure and approval of conflicts;
• rules for opportunities related to the company’s business;
• confidentiality obligations and ordinary exceptions;
• customer and employee solicitation restrictions, if lawful and appropriate.

Non-compete and non-solicit enforceability is jurisdiction-specific and changes over time. Cooley’s guidance notes that state law differs and that California, for example, generally does not enforce employee non-competes except in narrow circumstances. Do not copy a nationwide restriction from a generic form and assume it works. Ask counsel to tailor the agreement to where each founder works and where the company operates.

Decision 11: rehearse founder departure before writing the clause

“Founder leaves” is not one scenario.

Run at least these cases:

• voluntary resignation before the cliff;
• voluntary resignation after partial vesting;
• termination for cause;
• termination without cause;
• failure to become full-time by the promised date;
• sustained underperformance without misconduct;
• disability or long-term inability to work;
• death;
• founder bankruptcy or divorce affecting shares;
• a founder who leaves employment but keeps a board seat;
• acquisition followed by role elimination;
• both founders wanting to continue the same idea separately.

For each case, answer:

• What happens to unvested shares?
• What happens to vested shares?
• Does the company or another holder have a purchase right?
• At what price and on what payment terms?
• Does vesting accelerate?
• When do officer roles and board seats end?
• Who controls repositories, domains, banking, customer systems, and credentials?
• What continuing confidentiality, IP, cooperation, or non-disparagement duties apply?

Avoid vague “good leaver / bad leaver” labels without exact definitions. The economics can change dramatically depending on whether vested shares are retained, purchased at fair market value, or exposed to another formula. Counsel should make the trigger and remedy precise.

Your founder-equity decisions should also remain consistent with the company’s broader hiring philosophy. See our guide to equity for first employees for the option-pool and dilution context that follows founder issuance.

Decision 12: define sale, shutdown, governing law, and amendments

Companies end in more ways than success.

Specify the approval required to:

• accept an acquisition offer;
• sell material assets or IP;
• stop operating;
• dissolve the entity;
• distribute remaining assets after liabilities;
• continue or reuse the idea after shutdown;
• amend the founder deal;
• add a new founder;
• move agreed terms into replacement company documents.

Choose governing law and dispute venue with counsel. The correct answer depends on the entity, founder locations, and subject matter. A Delaware corporation does not mean every founder-employment issue is governed only by Delaware law.

Include a written-amendment process. Startups change quickly, but memory is not governance. When a founder changes commitment, a financing changes control, or the board expands, update the relevant documents deliberately.

The 60-minute founder alignment meeting

Do not begin with contract prose. Use one hour to find the disagreements.

Minutes 0–10: the company each person thinks you are building

Each founder answers without interruption:

• What are we building?
• For whom?
• What does success look like in three years?
• Are we trying to raise venture capital?
• What will make me leave?

Minutes 10–20: commitment and money

Write down:

• full-time start dates;
• minimum and target salary timing;
• personal runway constraints;
• existing and future cash contributions;
• outside-work boundaries.

Minutes 20–35: equity and vesting

Agree on:

• split and rationale;
• vesting start dates;
• prior-work credit;
• cliff and schedule;
• acceleration position;
• what happens if a promised commitment never begins.

Minutes 35–45: authority and deadlock

List:

• each founder’s domain;
• routine spending authority;
• reserved decisions;
• approval thresholds;
• first three steps of the deadlock ladder.

Minutes 45–55: IP and departure

Inventory:

• prior employers and obligations;
• existing code, domains, data, designs, and brands;
• contractors and agencies;
• prior inventions to exclude;
• the three most likely departure scenarios.

Minutes 55–60: unresolved list and owner

Do not force fake agreement. Record each unresolved issue, the information needed, the person responsible, and the date for resolution. Send the memo to counsel only when the open questions are visible.

Red flags: do not sign a generic template yet

Pause and get tailored advice if any of these are true:

• a founder is still employed elsewhere and has signed an invention-assignment agreement;
• code or designs were created by contractors without signed assignments;
• a university, lab, accelerator, grant, or prior company may have IP rights;
• founders live or work in different countries;
• one founder is contributing patents, regulated data, or licensed technology;
• the split changes based on future milestones or dynamic contributions;
• one founder is a minor, married in a jurisdiction where spousal rights matter, or subject to a divorce order;
• the company is issuing restricted stock and the 83(b) window is running;
• the team wants unusual acceleration, forced-sale, buy-sell, or bad-leaver terms;
• founders disagree about whether money already spent was a loan or equity;
• the entity has already accepted investment or issued securities;
• a founder wants to keep company-related IP personally and license it in.

A template is most dangerous when it makes an unresolved issue look resolved.

The lawyer handoff packet

A prepared founder team makes legal work faster and more useful. Send counsel one folder containing:

1. the one-page founder deal memo;
2. a cap-table draft with the equity rationale;
3. vesting start dates and prior-work credit;
4. founder legal names, addresses, citizenship or tax-residency notes requested by counsel, and current employers;
5. the IP inventory and prior-inventions schedule;
6. contractor, agency, employment, accelerator, grant, and university agreements;
7. the contribution schedule for cash and property;
8. the reserved-decisions list and proposed deadlock ladder;
9. departure scenarios and agreed economics;
10. intended entity, financing path, and state or country of operation;
11. unresolved questions;
12. target signing and stock-transfer dates so tax deadlines can be planned.

Ask counsel to confirm which decisions belong in:

• the certificate or charter;
• bylaws or LLC operating agreement;
• founder restricted stock purchase agreements;
• board and stockholder consents;
• voting or transfer-right agreements;
• confidential-information and inventions-assignment agreements;
• contractor assignments;
• cap table and stock ledger;
• tax filings, including any 83(b) election.

This is the real cofounder agreement template: a complete set of decisions, implemented in documents that actually perform the intended legal and operational work.

Final founder checklist

Before anyone signs or transfers money, confirm:

• Every founder and significant early contributor is identified.
• The venture and excluded prior projects are defined.
• Goals, full-time dates, and outside-work limits are written down.
• The equity split has a rationale.
• Founder equity has a defined vesting start, cliff, schedule, and repurchase mechanism.
• Acceleration is explicitly included or excluded.
• The team knows whether an 83(b) election is relevant and who owns the deadline.
• Roles describe outcomes and authority, not only titles.
• Founder salary, expenses, cash contributions, and loans are classified.
• Reserved decisions have approval thresholds.
• A realistic deadlock process exists.
• Existing and future IP is inventoried and properly assigned.
• Contractors and agencies have signed IP documents.
• Prior-employer and third-party obligations have been reviewed.
• Departure, disability, death, sale, and shutdown scenarios have been tested.
• The founder deal matches the charter, bylaws or operating agreement, equity documents, consents, and cap table.
• Qualified startup and tax counsel has reviewed the final package.

The goal is not to predict every conflict. It is to remove the ambiguity that turns ordinary change into betrayal. If you can make these decisions clearly while the stakes are low, you have done more than draft an agreement—you have tested whether the founding team can govern a company together.