Working alone makes you fast and fragile

I have spent a long time alternating between working in a team and pushing projects forward on my own, and both have a cost that rarely gets named clearly.

Working alone you are faster, which is true and obvious from the first day. You decide in the morning and by the afternoon it is done. No alignment meetings, nobody to convince, no waiting for somebody to free up.

What you pay for that speed is fragility, and it does not show until late.

The mistake nobody corrects

The risk of working alone is not getting something wrong, it is getting it wrong for a long time. In a team, somebody walking behind your screen can ask why you are doing that, and the question saves you weeks. On your own, the question does not exist.

And it does not turn up by itself, because once you are three weeks into something your own invested effort becomes an argument for continuing. You become the worst possible judge at exactly the moment an outside judgement is most needed.

How I compensate

The only thing that has worked for me is having somebody outside I can show things to without asking permission and without preparing anything. Somebody from the trade, who understands the context well enough, and who I have explicitly told that their job is to make me uncomfortable.

That last part matters more than it sounds. If you do not say it, the person will be kind, because kind is the default social behaviour. You have to give them permission not to be, and you have to give it before you show them anything.

The other thing that helps is writing decisions down. Not to document them, but to be able to argue with myself a few days later, which is the closest thing to a second opinion I have when there is nobody.

What a team is actually for

With a team you gain friction and lose agility, and the trade is worth it for a reason that took me a while to understand: what a team brings is not execution capacity, it is disagreement.

Execution can be bought. Informed disagreement cannot: it comes from somebody who knows the context, has judgement of their own and is willing to argue. That is what makes a team worth what it costs, and it is exactly what is lost when people are hired to execute rather than to have opinions.

That said, every decision needs an owner and not a committee. A team that argues well and decides badly is slower than one person and no more right.

When each one suits

In my experience alone works better when the problem is clear and execution is the expensive part. A team works better when the problem is unclear and choosing wrong is the expensive part.

The common mistake is assembling a team to go faster. A team does not make you faster; it makes it harder to be wrong for a long time, which is a different thing and usually more valuable.

Questions people ask about this

Can I build a product on my own?

Yes, and with less friction every year. What you cannot assemble on your own is outside judgement, so get hold of it even informally: one person, every fortnight, who you show what you have.

When should I bring somebody in?

When the cost of being wrong exceeds the cost of coordinating. While decisions are reversible and cheap, going alone pays. As soon as they get expensive to undo, you need somebody who can contradict you.

How do I get somebody to criticise me properly?

By asking explicitly and defending nothing while they talk. If you start explaining why you did it that way at the first comment, that person has learned that criticising you has a cost, and they will not do it again.