Founders ask this as if one option is correct and the other is a compromise. It isn't. Both are ways of buying technical judgement, and they fail differently. The useful question is which failure you can survive.
What each one actually protects you from
A CTO protects you from building the wrong thing repeatedly. Someone with equity and a long horizon will argue with you, refuse work, and care about decisions whose cost lands in eighteen months.
An outside partner protects you from paying for capacity you don't need yet. You get senior work on demand, without a salary, equity, or the six-month cost of a bad hire.
Notice these are different risks. If you don't know which one is currently killing you, you can't answer the question.
Hire, when these are true
- Technology is the product, not the delivery mechanism. If your defensibility is a model, an algorithm, or a data asset, that knowledge cannot live outside the company.
- You're raising, and investors will ask. A pre-seed round with no technical founder is a harder conversation. Sometimes an unfair one, but you'll still have it.
- The roadmap is continuous, not a project. If there's always a next thing, you're paying agency margin forever.
- You can genuinely evaluate the hire. A non-technical founder hiring their first CTO alone is the single most expensive coin-flip in early-stage startups.
Outsource, when these are true
- You're still testing whether the thing works. Buying eighteen months of salary to validate a three-week hypothesis is backwards.
- The work is a bounded project. A mobile app, a migration, a rebuild. Bounded work has a natural end; employment doesn't.
- You need a skill for eight weeks, not eight years. Nobody hires a full-time iOS engineer for one app release. They just live with the wrong hire afterwards.
- Speed matters more than ownership right now. Hiring a good senior engineer in Europe takes three to five months. A partner starts next week.
What a CTO actually costs
The salary is the part everyone budgets for and the smallest part of the number.
- Salary. €80,000–140,000 in most of Europe for someone who has genuinely shipped and led. Higher in London, Zurich, Amsterdam, and the Nordics.
- Equity. 1–10% depending on stage and whether they're a co-founder. At seed, this is usually worth more than the cash.
- Employer cost. Add 30–35% in Spain and most of the EU for social contributions.
- Recruiting. Three to five months of your time, or 20–25% of first-year salary to an agency.
- The cost of getting it wrong. Six to nine months of salary, plus a codebase built to a judgement you've now decided you don't trust. This is the real number, and it's the one nobody models.
All in, a first CTO hire is a €150,000+ first-year decision with a meaningful failure rate. That doesn't make it wrong — for the right company it's the highest-return hire there is. It makes it a decision worth taking as seriously as a funding round.
Signals you hired wrong (and how early you can tell)
Usually visible by month three if you know what to watch for:
- Everything is a rewrite. A good engineer inherits pragmatically. Someone who wants to rebuild everything they didn't write is telling you about their preferences, not your codebase.
- You never hear about trade-offs. "Yes, we can do that" to everything means either they aren't thinking about cost or they aren't telling you about it.
- Nobody else can deploy. If knowledge concentrates rather than spreading, you've hired an individual contributor with a title.
- Estimates are always right at the task level and always wrong at the release level. That's a translation problem, and it's fixable — but only if you name it.
The option nobody names
There's a third path that gets used constantly and discussed rarely: outsource the build, hire the judgement.
A fractional or part-time technical lead who owns the architecture, reviews the work, and tells you when a supplier is wrong — combined with a delivery partner who does the volume. You get an advocate on your side of the table without a full-time salary, and you get throughput without hiring a team you'd have to unwind.
This is what I do for most of the founders I work with, and it exists because the binary is false.
The failure modes, stated plainly
Hiring goes wrong when you hire someone who has never shipped the kind of thing you're building, because they interviewed well. Twelve months later you have a codebase only they understand and a roadmap that keeps slipping — see the CEO↔CTO translation problem.
Outsourcing goes wrong when nobody on your side can evaluate the work. You end up trusting a status report you can't verify, and by the time you can, the money is spent. If that's your situation, the fix isn't a better supplier — it's technical due diligence by someone who owes you nothing.
The question to ask yourself tonight
Not "can I afford a CTO?" but: if my current technical decisions are wrong, how long until I find out?
If the answer is weeks, outsource and stay close. If the answer is years, you need someone whose own outcome depends on being right — and that means equity, not an invoice.