Skip to content
Lessons #Startups#Hiring#Consulting#MVP#Founders

Where to Find Developers for Your Startup

· 6 min read

Founders ask this as a question about places. Upwork or Toptal? An agency or a freelancer? Should I just hire someone? Does my cousin’s friend who builds websites count?

Where to look is the last decision you make, not the first. And when the hire goes wrong, it almost never goes wrong because someone picked the wrong platform.

The two questions that actually decide it

Is it clear what needs to be built? Not the idea. The thing. Screen by screen, rule by rule, what happens when the payment fails.

Is there anyone inside your company who can tell good technical work from bad? Not someone who can write code. Someone who can look at what came back and know whether it will still stand up in a year.

Those two answers lead you to one of four situations, and each one has its own right answer.

A decision tree. First: is it clear what needs building? Then, on each branch: can anyone in your company judge the work? The four combinations lead to four different answers. Is it clear what needs building? No Yes Can anyone in your company judge the work? Can anyone in your company judge the work? No Yes No Yes Someone senior to set the scope Most founders land here An agency or a team Review the spec before you build A freelancer or a marketplace
Answer both and the tree leaves you at your answer.

Clear scope, and someone who can judge the work

Use a marketplace. Genuinely.

This is the one case where Upwork and Toptal do what they say on the tin. The work is specified, somebody on your side can check what comes back, and what you want is capable hands at a good price. Pay the premium for a platform that screens, or do your own screening and pay less.

Nothing in this article should talk you out of the cheapest answer when the cheapest answer is correct.

Clear scope, nobody who can judge the work

This is the case that costs the most money, and almost nobody warns you about it.

The scope is already written down, so the quotes come back fast and look comparable. You pick one. Six months later you own something that does what the document said and cannot be grown, because the decisions underneath it were made by someone with no reason to argue with you.

A specification is a list of answers. It is not proof that anyone asked the right questions.

The fix is not a better developer. It is buying a few days of someone senior to put that document under pressure before you commission it, and then using whatever you like for the build. That is a small expense against the size of the mistake it prevents.

Scope still vague

Then it is not time to hire whoever will build it, whatever your budget says.

Most founders arrive here and try to skip it, because it feels like paying for talk. What you are actually buying is the decision about what gets built, which is the decision that determines everything the money does afterwards.

Expect whoever helps you to spend the first conversation on your business and not on the technology. If they open on languages and tools, they are answering a question you did not ask.

How to tell a good one before you pay

Four questions, and they work on an agency, a freelancer or an individual.

“What would you tell me not to build?” Anyone good has an answer ready. The useful ones have talked a client out of something.

“Show me something you shipped that is still running.” Still running is the whole question. Shipped and dead proves you can start.

“Who owns the code and the accounts while the work is happening?” They should be yours from day one, and whoever you ask should not have to think about the answer. If they do, read what an MVP handover should include before you sign anything.

“What happens when you go away?” Everyone goes away eventually. The ones worth hiring have thought about it and will tell you the plan without being pushed.

One thing that is not on the list

Price, as the first filter.

The cheapest quote and the most expensive quote are both telling you something, and neither is telling you about quality. A low number usually means the work was underestimated, and that bill arrives later with interest. A high number can mean a team with the overhead of a team, which is a fine thing to pay for when you need a team and a waste when you do not.

Work out which branch you land on. The budget conversation gets much shorter afterwards.


I work in the branch marked on the diagram: independent technology consulting and custom software, usually for companies where the technical decision has not been made yet. If the tree left you at the freelancer or the marketplace, that is the right answer for your case. And if it did not leave you anywhere clear, write to me: that usually takes one conversation to settle.

Related Posts