Similarity and close existing ties shaped who founded companies together more than any other factor a 2003 study of US founding teams tested.¹ So knowing how to choose a co-founder means working against that pull: look for what you lack, try the partnership on real work, and settle the awkward questions while everyone still gets along.

The steps, in the order they are best taken:

1. Decide whether you need a partner or just some help
2. Write down the skills and contacts you lack
3. Work together on a real project for a few weeks
4. Put roles and decision rights on one page
5. Settle equity, vesting and exits with a local lawyer

Many teams rush the last step. In Thomas Hellmann and Noam Wasserman’s 2017 study of North American technology start-ups, teams that split their shares equally were less likely to raise money from outside investors, though the authors judge the split a sign of the kind of team rather than the cause.² If you have not yet decided to leave a salary, begin with whether starting a business makes sense for you.

Do you need a co-founder at all?

Not necessarily. Many start-up investors treat a second founder as close to essential, but the direct evidence comparing solo and team founders is thin and points both ways, so the honest answer depends on what your business needs that you cannot supply yourself.

The case for a partner is mostly practitioner advice. In a 2006 essay, Paul Graham, a co-founder of the US accelerator Y Combinator, which invests in start-ups, put a single founder first on his list of mistakes that kill start-ups: the work is too hard for one person, and you need colleagues to test ideas on, to talk you out of bad decisions and to lift morale when things go wrong.³ That is experience from one investor’s portfolio of fast-growth technology firms, not a study.

The research that exists complicates it. A 2018 working paper by Jason Greenberg and Ethan Mollick, not yet peer-reviewed and read here only in abstract, found that companies started by solo founders survived longer than those started by teams, in a dataset of crowdfunded firms.⁴ A team brings more hands, and also a relationship that can fail.

Myth

Every start-up needs a co-founder to succeed.

Fact

Many investors prefer teams, but the direct evidence is mixed, and many gaps can be filled by hiring, contracting or borrowing help.

A small 2019 interview study by Travis Howell and Christopher Bingham of 59 technology ventures in one US city suggests how solo founders cope. The successful ones got much of what a co-founder brings through other routes: early employees, alliances with other organizations, and benefactors who lent equipment, staff or advice without asking for anything back. The authors’ summary is that successful solo founders are not really alone.⁵

In practice, a designer opening a small studio may need a bookkeeper and a lawyer, not a partner. A software product that needs building and selling from the first month is a different case. If what you are missing can be hired, contracted or borrowed, you may not need to give away part of the company to get it.

How to choose a co-founder for the skills you lack

Choose for difference in skills, work history and contacts, and for knowledge of the industry you are entering. Most people do the opposite: in a 2003 analysis of founding teams drawn to represent the US population, Martin Ruef and colleagues found that similarity and strong existing ties shaped who teamed up more than the other factors tested, including the functions a business needed.¹

The pull is easy to understand. People like you are easy to find, easy to trust and already in your phone. The catch is that two people with the same background tend to know the same things and the same people, so the gaps in the business stay gaps.

One study of fast-growth firms favors range. In 161 young technology firms in Silicon Valley, Christine Beckman, Diane Burton and Charles O’Reilly found that firms whose founding and early management teams had worked in different functions, and for many different employers, were more likely to raise venture capital and to go public.⁶ It covers one region and one industry and is an observational studyobservational study: A study in which researchers record what people already do or are exposed to, rather than assigning anyone to anything. It can show that two things go together, not that one causes the other, because the groups being compared may differ in other ways as well.Full entry in the glossary, so it describes who did well, not why.

Age and industry know-how count too. In US administrative records on firms and their owners, Pierre Azoulay and three co-authors reported that founders of the fastest-growing new firms were about 45 on average, and that experience in the same industry predicted much greater success.⁷

45mean age at founding of the 1-in-1,000 fastest-growing new US firms, in US administrative recordsSource: Azoulay, Jones, Kim and Miranda, 2020

Picture two software developers from the same course building a booking tool for dental clinics. Neither has ever sold anything to a clinic. A former practice manager who knows how clinics buy would add more than a third developer would. A founder in their twenties, or new to a trade, may gain more from a partner with years inside it than from one who shares their age and enthusiasm.

Founding with a friend: easy to start, easy to skip the hard talk

Starting with a friend or relative is common and comfortable, and research on founding teams of relatives links that choice to faster, more equal deals and weaker growth later. Noam Wasserman’s 2012 book The Founder’s Dilemmas, built on data about almost ten thousand founders, weighs whether to found with friends or relatives and, its publisher says, shows why the easy short-term choice is often the riskiest in the long run.⁸

His later study with Thomas Hellmann gives one reason. Founding teams made up of family members were more likely to agree their split quickly and equally, and less likely to raise money from outside investors. The authors’ explanation is that relatives may see unequal rewards as unfair in themselves, so they skip the discussion that would reveal who brings what.² In Danish data on two-person founding teams, Alex Coad and Bram Timmermans reported that family firms grew their workforce more slowly, an abstract-only finding here.⁹

Take two old friends who agree to go 50/50 over dinner. One quits her job; the other keeps his and helps at weekends. Nobody raises it, because raising it feels like distrust. A year later, the unspoken gap has become a grievance.

A friend can make a fine co-founder. The risk lies in skipping the conversation you would insist on with a stranger.

How to split the work between co-founders

Split the work by the results each person owns, name who makes which decisions, and agree how a deadlock gets broken. Y Combinator’s Michael Seibel, in 2024 advice aimed at venture-funded technology start-ups, argues that co-founders should be only the people essential to getting a first product into customers’ hands, and that one person, the chief executive, must be ultimately accountable.¹⁰

Clear ownership matters because early work overlaps and nobody is anyone’s boss. When two founders both “do product”, decisions stall or get made twice; when nobody owns sales, the sales do not happen. Writing it down forces the choices that goodwill postpones.

A workable split for two founders of that dental booking tool might look like this. The developer owns the product, its reliability and hiring engineers. The former practice manager owns sales, pricing and customer support. Both must agree before raising money, adding a founder or changing the company’s direction, and if they cannot agree, a named adviser they both trust has the casting view.

Seibel also separates pay from ownership: a founder who needs a salary to cover rent should not get less equity for it, because salary is what lets someone work and equity is what motivates them for the years ahead.¹⁰ That is advice from one accelerator, not research, but it keeps two different conversations from blurring. If the two of you cannot agree a division like this in an afternoon, you have found your first disagreement while it is cheap to settle.

Splitting equity: why a quick 50/50 deserves a second look

No study can tell you the right split for your team, but a large North American study of founder equity shows how often teams settle it in a hurry. Hellmann and Wasserman drew on a survey of technology and life-sciences start-ups that Wasserman runs.²

The study

Limited evidence

Equal splits were common, and often agreed within a day, in a survey of North American founders

32% of teams split the equity equally, and 42% settled their split within a day. Equal splitters were less likely to have raised money from outside investors, but once the analysis tried to separate cause from selection, the split itself showed no clear effect. In a supplementary 2013 survey of 289 respondents, 86% said all founders were satisfied with the deal when it was signed; looking back, 66% did.²

The lesson is not “never split 50/50”. Equal splits were, if anything, linked to more satisfaction in hindsight. The warning is about speed: a split agreed in an evening means the team has not yet discussed who brings what, and a team that avoids that talk may avoid harder ones later. A deal that pleased everyone on signing day did not always stay that way. Say one founder has worked on the idea for six months and the other joins now: talking it through may still end at 50/50, but both will know why.

Vesting is the safeguard Y Combinator recommends. In the arrangement Seibel calls typical, each founder earns their shares over four years, with a one-year cliff: anyone who leaves before the first year is up has earned none of them, though he says early leavers in YC’s experience often get a small token stake. He also urges founders to agree in advance what happens if someone leaves or is let go.¹⁰

1. Earned over time: in the arrangement Y Combinator calls typical for US tech start-ups, each founder earns their shares over four years
2. The one-year cliff: a founder who leaves before the first year is up has earned none of the shares
3. Roles and decisions: who owns which results, which decisions each person makes, and how a deadlock is broken
4. The split: how the shares are divided, talked through rather than assumed
5. Exits: what happens to shares and roles if someone leaves or is let go

These are US start-up conventions, not rules. How shares, vesting and founder agreements are written, taxed and enforced differs by country and by type of company, so have a lawyer where the company is registered draft or review the documents.

From first conversation to signed agreement

Taken in order, and before any shares change hands, the advice above turns into five steps.

1. Decide whether you need a partner or just some help

List what the business needs in its first year. For each gap, ask whether it could be hired, contracted or borrowed. Only a gap that needs someone fully committed for years, and that you cannot fill another way, calls for giving away ownership.

2. Write down the skills and contacts you lack

List the jobs you would do badly and the customers you cannot reach. Look for candidates among former colleagues and industry contacts as well as friends, and weigh industry experience heavily.

3. Work together on a real project for a few weeks

Before any equity changes hands, build a prototype, pitch real customers or deliver a paid project together over several weeks. You learn how the other person handles deadlines, disagreement and bad news, which no interview reveals. No study has tested trial projects, so treat this as practical sense, not a proven method.

4. Put roles and decision rights on one page

Cover who owns which results, which decisions each person makes alone, which need both, how a deadlock is broken, how much time each commits and what each will be paid.

5. Settle equity, vesting and exits with a local lawyer

Discuss the split openly, even if you end at 50/50, and agree vesting and what happens if someone leaves. Then have a lawyer turn it into a founder or shareholders’ agreement valid where you are registered.

WiserHours

Before you sign with a co-founder

• List the gaps and whether each could be hired or borrowed
• Look beyond friends to colleagues and industry contacts
• Complete a real project together first
• Write one page on roles, decisions, time and pay
• Talk through the split and what happens if someone leaves
• Have a local lawyer draft or review the agreement

Find this checklist online:
https://wiserhours.com/entrepreneurship/choose-a-co-founder/#checklist-choose-a-co-founder-before-you-sign

What the founding-team studies measured, and what they missed

Research on founding teams is almost all observational, mostly from US and Canadian technology start-ups plus one Danish study; nobody has randomly assigned founders to partners, so it shows which teams did well, not why.

The outcome research also leans toward fast-growth technology firms: a two-person bakery or consultancy faces the same questions about roles and exits, but few of these studies looked at firms like it. Treat the findings as patterns worth weighing, not rules.

Lawyer, accountant or both: three moments to get advice

Your circumstances, and your country’s company and tax law, decide the details, so professional advice matters at three points.

• Before anyone signs anything about shares, vesting or who owns the work already done: a lawyer qualified where the company is, or will be, registered.
• Before shares are issued or money comes in from outside: an accountant or tax adviser, because how founders’ shares are taxed differs by country.
• Routinely, once a year or when roles change: review the agreement together. If a partnership is already under strain, co-founder conflict is a separate subject within the entrepreneurship section.

The bottom line

In the tech start-ups studied so far, the co-founder who fills your gaps and knows the industry tends to be a better bet than the friend who thinks like you, and some businesses need no co-founder at all. Whoever you choose, work together on something real first, then write down roles, decisions and what happens if someone leaves. Talk the split through rather than settling it in an evening, and have a local lawyer put it on paper.