Recruiting a Co-Founder: Profile, Deal and Mistakes to Avoid

Quick answer. A good co-founder brings a skill you lack, shares the risk and stays aligned over time. Before granting equity, test them on a real project for a few months, write down the split, set vesting and a founders agreement. Disputes between partners remain one of the leading causes of failure in young companies.

The question is not only where to find a co-founder, but how to keep them from becoming, eighteen months later, your company's main blocker. In company-formation files, disputes between partners rank among the most frequent reasons for breakdown: a split frozen in the early excitement, no vesting, departures with no share buy-back. This article helps you choose the right profile, test them before granting equity, and structure a deal that holds when money and fatigue arrive.

If you are still unsure about partnering at all, read our analysis first on whether to go it alone or partner up. Here, the decision is assumed made: you are searching for and selecting a partner.

A co-founder is not an employee paid in shares. It is someone who shares the risk, commits without guaranteed pay, and on whom your ability to execute depends. Before searching, clarify the gap you are filling.

The three profiles most often seen in young companies:

• The technical co-founder: they build the product. Essential if your project relies on technology you cannot master and cannot outsource without losing control.
• The business co-founder: they sell, raise, recruit, structure. Useful when the lead founder is a product or technical profile lacking a commercial or financial network.
• The operations or domain co-founder: they know the sector, the clients, the regulation. Decisive in regulated or operationally intensive markets.

Our reading#

Complementarity is not declared on a CV. What matters is the area you cannot cover yourself and that is critical for the next twelve months. If you are a product profile with no sales skill, a second product profile doubles your comfort but not your ability to generate revenue. The right question is not "who is like me" but "who makes me executable".

There is no magic channel. Lasting partnerships almost always come from an already tested relationship, or from working together before any equity commitment.

1. Your direct network: former colleagues, classmates, people with whom you have already delivered a project. The most reliable pool, because you have seen the person under pressure.
2. Specialised communities and events: hackathons, sector meetups, incubation programmes. You meet motivated profiles there, but the time filter remains essential.
3. Co-founder matching platforms: useful to widen the search, provided you never skip the testing stage.
4. Startup ecosystems and accelerators: they offer a framework, a network and sometimes a first seriousness filter.

Our support for setting up the company often starts with this phase: structuring the project so it is legible to a future partner or investor.

This is the most neglected and most protective stage. Giving equity to someone you have never seen work under pressure means signing a blank cheque for several years.

In practice#

Before any share grant, organise a real collaboration period, from a few weeks to a few months, on a concrete deliverable: a prototype, a first sales phase, a client file. Observe:

• reliability on commitments and deadlines;
• how they handle a disagreement, bad news, a challenge to their view;
• alignment on ambition, pace and acceptable risk;
• transparency about money, available time and other commitments.

The underestimated risk#

The danger is not the obviously bad co-founder, who shows quickly. It is the good profile, badly aligned: highly competent, but wanting a work-life balance incompatible with an early stage, or seeing the company as a side project rather than a life commitment. This misalignment never shows on a CV. It surfaces in joint work, under tension.

Once trust is established, the deal must be written. An oral agreement between co-founders is worthless the day one leaves or steps back.

What percentage to give a co-founder?#

There is no legal scale. The split reflects the real contribution: project seniority, funding, full or part time, value of the skill, risk taken. Two prudence principles:

• Avoid a strict 50/50 out of comfort: it is appealing but creates a risk of total deadlock in case of disagreement, with no exit mechanism. If you choose it anyway, you must include a conflict-resolution clause in the agreement.
• Do not freeze the split on day-one enthusiasm: that is exactly what vesting is for.

To see how a split and a future round affect your capitalisation table, our dilution and cap table simulator gives a first numerical read.

Vesting, the central protection mechanism#

Vesting makes the definitive acquisition of shares conditional on continued presence over time. The most common scheme in the startup ecosystem provides for gradual acquisition over several years, with an initial period (cliff) below which a departure entitles to nothing. Concretely, a co-founder who leaves after a few months does not keep all their equity. It is the best protection against the "ghost co-founder" who holds 30% of the capital after disappearing.

Vesting is built into the shareholders agreement and, where relevant, through transfer undertakings. To go further, read our dedicated article on the founders agreement and vesting before a round and the overview of the 15 vital clauses of a shareholders agreement.

Decision table: which deal for which situation#

When a co-founder joins after incorporation, or when you want to interest a key profile who lacks historical-founder status, founder share warrants (BSPCE) are a common alignment tool in startups.

Their tax and social regime is regulated. The scheme is open, under conditions, to share companies less than fifteen years old. One structural point to know: for BSPCE eligibility, the company's capital must be held directly and continuously, for at least 15%, by individuals (or by legal entities themselves directly held, for at least 75%, by individuals). This threshold was lowered from 25% to 15% by the 2026 Finance Act (LOI n 2026-103 of 19 February 2026); it appears in Article 163 bis G of the French General Tax Code, in its version in force as of 21 February 2026. Any holding structure or entry of corporate investors must be checked against this threshold, on pain of losing eligibility.

We detail the mechanics, advantages and limits in our BSPCE guide. Implementation requires a valuation of the shares and coordination with the statutory auditor where one is appointed: this is a topic to frame with your startup tech accountant.

Most projects with several co-founders and a fundraising path are set up as an SAS, for its statutory flexibility. That flexibility comes at a price: partner-protection rules are not automatic, they must be written.

In an SAS, certain clauses governing the circulation and retention of shares fall under the unanimity regime of the partners, not a simple majority. Article L227-19 of the Commercial Code now reserves unanimity for inalienability clauses (Article L227-13) and change-of-control clauses (Article L227-17) only. Approval clauses (Article L227-14) and exclusion clauses (Article L227-16) can, for their part, be adopted or amended under the company's rules for amending the articles of association.

A useful historical point to avoid a common mistake: this liberalisation did not happen in one step. The approval clause (L227-14) left the unanimity regime as early as Ordinance no. 2017-747 of 4 May 2017, in force on 6 May 2017. Law no. 2019-744 of 19 July 2019, known as the Soilihi law, only freed the exclusion clause (L227-16). Concretely, you can today provide for an exclusion clause and an approval clause adoptable by statutory majority, which changes how you manage a partner who has become a blocker.

These clauses are at the heart of any solid agreement. Have them drafted with a professional: this is the purpose of our service to secure the company's agreement and articles.

In company-formation files, one pattern recurs regularly: two partners launch a company 50/50, with no vesting and no agreement, convinced their good relationship will suffice. Eighteen months later, one steps back, keeps half the capital, and the company is blocked at every important decision. The share buy-back becomes a conflictual, sometimes judicial negotiation that paralyses growth at the worst moment. All of this is preventable upstream, with vesting, an agreement and an exit clause. To anticipate this scenario, see also how to manage a co-founder's departure.

Granting equity is not always the right answer. Before committing capital, weigh the trade-off:

• Co-founder: the person is essential long term, accepts the risk and sees the project as a life commitment. Equity is justified, with vesting.
• First employee or key executive: the skill is crucial but the commitment is that of a job. A BSPCE or deferred grant can interest them in the capital without diluting like a co-founder.
• Contractor or freelance: the need is one-off or specialised. An invoiced service avoids giving up capital for a temporary assignment.

An outsourced CFO can help you cost each option and its impact on the cap table: this is the purpose of our outsourced CFO for startups and SMEs.

• Look for the skill that makes you executable, not a profile that resembles you.
• Test the person on a real project before any share grant: misalignment surfaces under tension, never on a CV.
• Avoid 50/50 without a deadlock clause; never freeze the split without vesting.
• Write everything down: split, vesting, agreement, approval and exclusion clauses.
• For BSPCE, the capital must be held directly and continuously, for at least 15%, by individuals (threshold lowered by the 2026 Finance Act).
• In an SAS, unanimity is required only for inalienability and change-of-control clauses; approval and exclusion are decided by statutory majority.