Fractional CTO vs freelance CTO vs agency: which one for a pre-seed startup?

At pre-seed, the right model depends on what you are actually missing: an agency if you have a fixed scope and a lump-sum budget, a freelance CTO if you need someone coding and leading full time, a fractional CTO if what you lack is senior technical decisions without the salary that comes with them. All three are legitimate choices, and picking the wrong one usually costs more than the model itself.

I work as a fractional CTO, so my bias is obvious. I have also seen enough projects to know that sometimes an agency or a freelancer is exactly what a founder needs. I will tell you when.

Fractional CTO vs freelance CTO vs agency: which one for a pre-seed startup?

Before comparing, it helps to agree on the words, because people use them loosely.

Fractional CTO

A fractional CTO (sometimes called a part-time or outsourced CTO) is a senior engineering leader who works with several companies, usually 1 to 2 days a week each. They own architecture, code review, technical hiring and the roadmap. They sometimes write code, but that is not what you are buying. They are paid a monthly retainer, with a commitment of a few months.

Freelance CTO

A freelance CTO is usually a very senior developer who takes one engagement at a time, 3 to 5 days a week, billed by the day. In practice it is a lead developer who also wears the CTO hat: they write a lot of code, they make the calls, and they are there almost every day. Some do interim leadership between two permanent hires.

Agency

An agency sells a delivered project: a scope, a price, a date. It brings a full team (project manager, design, development, QA) and leaves once the product ships, unless you sign a maintenance contract.

The comparison in one table

These are typical market ranges for France and Western Europe in 2026, not guaranteed prices. I explain where they come from in my breakdown of what a fractional CTO costs.

When an agency is the right call

A good agency is hard to beat in one specific case: you know exactly what you want built, the scope will not move for three months, and you have the budget to pay for it in one go. A demo MVP for a fundraise, an event app, an internal tool with a clear spec: the agency delivers, with a full team you could never have assembled yourself.

It fits less well when the product has to change every week based on customer feedback, which is more or less the definition of pre-seed. Every change becomes a change request, and on delivery day nobody inside the company knows the code.

If you go with an agency, set three conditions: the code in your own repository, the infrastructure on your own accounts, and handover documentation written into the contract.

When a freelance CTO is the right call

A freelance CTO is the right call when you need a lot of code, fast, from someone who also makes the decisions. For example: a non-technical founder who has raised, has no team yet, and has six months to ship a product. A senior freelancer working four days a week moves faster than any part-time setup.

The risk is dependence on one person. When the engagement ends, the knowledge of the system walks out with them. They also cost about as much as a loaded salary, which weighs heavily on a small round. Take a hypothetical: with €300,000 raised, a freelance CTO at €15,000 a month burns half the round in ten months.

When a fractional CTO is the right call

A fractional CTO is the right call when you already have hands (one or two developers, a freelancer, a small nearshore team) but nobody senior to make the calls. They set the architecture, review the code that goes to production, help you hire, and give investors a credible technical story.

The limit is simple: at 1 to 2 days a week, they do not replace a developer. If nobody is writing the code, you need hands first. That is why I also offer a fixed-scope MVP build in about 4 weeks, followed by fractional CTO support once the product is live.

The decision checklist

• Pick an agency if your scope is fixed, you can pay a lump sum, and nobody will need to change the product the month after launch.
• Pick an agency if you need a full team (design, mobile, back end) for a one-off, well-specified project.
• Pick a freelance CTO if you have no developers, you have raised enough to pay the equivalent of a salary, and you need code every day.
• Pick a freelance CTO if you are waiting for a permanent CTO and need someone to hold the team together until then.
• Pick a fractional CTO if you already have developers but nobody senior to make decisions.
• Pick a fractional CTO if you are preparing a round and investors will look closely at your tech.
• Pick a fractional CTO if you want technical leadership without giving 5% of the company to a salaried CTO too early.

Combinations often work better than any pure model: an agency or a freelancer writing the code, and a fractional CTO reviewing, approving and keeping the whole thing coherent.

Three mistakes I see at pre-seed

• Hiring for the title, not the need. A founder hires a "CTO" when what they need for the next six months is someone shipping features. Or the reverse: they hire a fast developer when the real gap is nobody deciding what to build and how.
• Choosing on day rate alone. The cheapest option per day often becomes the most expensive per year once you add a rewrite at month 12.
• Giving away equity too early. Handing 5% to a first technical hire before you know the product works is hard to undo. A retainer can end. Equity cannot.

What to demand, whatever the model

The most important point does not appear in any table: who owns what gets built. Under French law, software written by an outside contractor generally stays the contractor's property unless the rights are assigned in writing. Rules differ across Europe, so do not assume the default works in your favour. Check that clause in the agency or freelance contract.

Guid'Up is a good example. The founder came to us after a previous build had been abandoned, and she wanted a product she could run herself. We set up her own store and infrastructure accounts, which we operate for her, and delivered the Flutter app, the admin back-office and the site from one monorepo. The result: she owns every account and publishes content without a developer.

So, whatever the model, ask for:

• the code in a repository you own,
• hosting, domain names and app store accounts in your name,
• documentation good enough for another team to take over,
• the option to stop at agreed checkpoints, without penalty.

If you have already inherited code you do not trust, a two-week technical audit tells you what you are dealing with before you choose what comes next.

The right choice, in one sentence

Buy hands if you are short of hands, buy decisions if you are short of decisions, and never pay for a product you will not own.

If you are torn between these three options, book a 30-minute call. I reply within 48 hours, and if an agency or a freelancer suits you better, I will say so.