Solo Founder or Co-Founder?

A founder I work with showed up to our call already exhausted, phone face-down on the desk.

He had a co-founder. That was the problem.

They had stopped agreeing on anything. Product. Hiring. What day it was, practically. And now he wanted to know if he should buy the guy out or just walk.

Here is what I told him. The question he was asking was the wrong one.

Everyone asks "how many" when they should ask "what"

People treat this like a headcount decision. One founder or two. Pick a number, go.

That framing is comfortable because it feels like a rule. Rules are easier than judgment.

But a company is not built by a number. It gets built by a set of things a person needs around them to survive the early years without quietly falling apart.

A second brain when yours is fried. A skill you do not have. Someone who tells you no. Someone who picks up the rope when your arms give out at 11pm on a Sunday.

That is what a co-founder is supposed to provide. Not a title. A function.

So the real question is not "solo or co-founder." It is "do I have these things, and if not, can I get them another way."

The research leans toward teams, and it is not wrong

I will be honest about what the evidence says. The research broadly leans toward co-founded companies. More resilience. Complementary skills. A load that gets carried by two backs instead of one.

That tracks with what I have watched for years across the Henson Group, Donald Ventures, and now Henson Venture Partners.

The lonely stretches break people. Founding is a grind that eats confidence for breakfast, and a good partner is a mirror that keeps you honest and a hand that keeps you moving.

I do not argue with the data. I just refuse to stop reading at the headline.

A bad co-founder is worse than no co-founder

Here is the part the averages hide.

The same partnership that doubles your strength can also cut your speed in half.

I have seen founders spend more energy managing a co-founder relationship than building the actual product. Every decision becomes a negotiation. Equity is locked up with someone who checked out in month four. Trust erodes, and trust is the whole thing.

When a marriage like that goes bad, it does not just slow the company. It can kill it, and it takes the friendship down with it.

So no, two is not automatically safer. Two of the wrong kind is a slow leak you cannot see until the boat is already low in the water.

The founder with the cold coffee learned that the hard way.

Solo does not mean alone

Now flip it.

I left the Microsoft ecosystem years ago and built from the ground up more than once. Some of those runs I was carrying more weight solo than a title chart would suggest. I am a veteran, so I already knew something about doing hard things without waiting for perfect conditions.

What kept me upright was never a co-founder box on a cap table. It was a support network. Advisors who had scars. Early hires who owned real pieces of the work. Peers I could call at ugly hours who would tell me the truth.

A solo founder with that around them can absolutely win. I have watched it happen, and I have lived pieces of it.

The trick is being ruthless about the gaps. If you go solo, you have to name what a co-founder would have given you and go build it on purpose. A strong first hire. A tight advisory bench. A therapist, honestly. Fund the loneliness problem before it funds itself.

Solo is a viable path. Solo and unsupported is not a path, it is a countdown.

Ask the harder version of the question

So sit with the real one.

Do you have the complementary skills, or a plan to hire them? Do you have someone who will disagree with you and mean it? Do you have people who share the emotional weight so it does not all land on you?

If yes, the structure matters less than founders think.

If no, adding a co-founder just to check a box is how you end up with the wrong partner and worse odds than if you had stayed solo.

Do not marry someone to avoid being single. Same logic. Higher stakes.

Pick the path where the support is real, not just labeled.