A part-time CTO who makes the technical calls with you — architecture, hiring, roadmap, technical debt — and a team to execute them. Starting from a few days a month.
Response within 24 hours.
A fractional CTO — also called a part-time CTO or CTO as a service — is a technical director who genuinely holds the role in your company without being on your payroll. The distinction matters: they do not hand over a report and leave. They occupy the post, with the obligation to actually decide that comes with it, for an agreed amount of time — most often two to eight days a month.
In practice they own the structural decisions: which architecture, which stack, what you build versus what you buy, who to hire and in what order, which technical debt to repay now, what level of security your market demands. These decisions commit months of development and tens of thousands of euros. A non-technical founder has no way to arbitrate them alone; an excellent developer, alone, does not always have the business context to arbitrate them well.
The model spread for an arithmetic reason. The function becomes essential very early — as soon as there is a product to design and a first euro to spend on development — but it does not fill a full-time role while the engineering team still fits in one room. Hiring a salaried CTO to lead two developers means paying a full package for a few days of real work each month.
So the right question is not "can I afford a CTO?" but "which technical decisions am I currently making without the competence to make them?". If the answer involves the word architecture, the word hiring or the word budget, you already need the function.
They execute a defined scope: you supply specifications, they deliver code. Excellent when the decisions are already made and you are short of hands.
A fractional CTO comes in earlier, when the specifications do not exist yet. Their deliverable is not code: it is judgement calls, and responsibility for those calls over time.
An IT director runs the internal information system: workstations, network, ERP, business tools, operational security. Their customers are your employees.
A CTO runs the product you sell. Their customers are your users. In a startup, it is almost always this second function that is missing.
An agency takes a specification and builds it. The model works when the specification is right — and it rarely is when nobody on the client side was able to write it.
We start by challenging the need, including when the honest conclusion is that you should not build what you asked for. Then we deliver. That order matters more than anything else.
A technical co-founder takes equity, commits for five to ten years and carries the risk with you. No service provider can match that level of exposure.
A fractional CTO does not replace that role. It lets you move without giving away 15% of your equity in the first six months — and often lets you choose that future partner better, because you finally know what to ask of them.
None of these signs is technical. They are all commercial — which is exactly why they go unnoticed until they get expensive.
€18,000, €47,000 and €92,000 for what you believe is the same product. The gap is not margin: the three of them did not understand the same brief. Without someone able to read between the lines, you are not choosing a proposal, you are choosing a price.
Every request is feasible, every deadline achievable, and yet deliveries keep slipping. This is not dishonesty, it is structural: nobody in the loop is tasked with telling you no. The CTO is the only role whose mandate explicitly includes refusing a feature.
You do not know how to write the job description, assess a technical test, or place the salary. Yet this hire determines the next five: it sets the technical bar, the practices, and your ability to attract the profiles that follow.
Each feature takes longer than the last, fixed bugs come back, and the team asks for "time to refactor" without you being able to judge. That is the signature of technical debt nobody is managing — and it is not solved by pushing the team to go faster.
A serious investor will ask who owns the code, how the platform will hold ten times its current traffic, where personal data is hosted, and how dependent you are on a single provider. Discovering these topics during due diligence means discovering them too late.
The real content of the mandate. Not a list of adjectives: identifiable decisions, each with a deliverable and a trace.
Settling what is blocking: rebuild or repair, ship now with less or wait, accept this shortcut or refuse it. Every call comes with its cost, its timeline, and what it will cost to reverse.
Defining the structure before the first line of code: monolith or separate services, which database, which boundaries between modules, which multi-tenant model. Cheap at the start, very expensive to change — so never improvised.
Job description, screening, technical interview, designing and marking the test, calibrating salary against the market. And where a team or provider is already in place, objectively assessing what is being produced rather than judging on a demo.
Reviewing what ships, from our team as much as from yours or a third party's. It is the only mechanism that stops a codebase degrading silently, and the first one dropped when nobody formally owns it.
Turning a wish list into an executable sequence: what gets built now, what waits, what will never be built. With the same question every iteration: what gets us to the next milestone fastest?
Inventorying it, costing it in days, deciding what to repay and what to live with. Not all debt is bad: some shortcuts are rational when validating a market. The dangerous debt is the debt nobody has written down.
Deciding what you develop and what you buy: authentication, payments, invoicing, search, email, analytics. Rebuilding an existing component almost always costs more than the licence — except when that component is precisely what differentiates you.
Access and secrets, backups that are actually tested, hosting and data location, GDPR mapping, preparing for the security questionnaires your enterprise customers send. Never urgent — until the day it blocks a contract.
The three options answer different situations, and two of them are not us. This is the comparison as we present it in meetings, unfavourable rows included.
| Criterion | Hiring a salaried CTO | Fractional CTO | Senior freelance developer |
|---|---|---|---|
| Cost | Market data: €90,000 to €130,000 a year in total package in Paris, recruitment cost included. | Market data: from around €3,900 a month for five days, a reported saving of 30-50% against a hire. | Billed by time, at a day rate. Proportional to volume, with no leverage on decisions. |
| Time to start | Three to six months of search, plus notice period. A failed hire resets the clock. | One to three weeks. An audit or a scoping day can begin before any long commitment. | A few days to two weeks, if the scope is already written. |
| Commitment | A permanent contract, a four to eight month probation, and a long, costly separation if it goes wrong. | Monthly or quarterly, with short notice. The financial risk stays reversible. | Per engagement. Flexible, but with no guaranteed continuity. |
| Seniority available | Potentially the highest — provided you succeed in hiring. A good CTO chooses their project as much as the reverse. | High immediately: you get a level your cash flow could not fund full time. | Variable, and different in kind: an excellent developer is not a technical director. |
| Main risk | Hiring the wrong person and finding out six months, several tens of thousands of euros and one committed architecture later. | Shared availability: they are not there five days out of five. If what you actually need is daily presence, this model will frustrate you. | Nobody is steering the whole. The knowledge leaves with the engagement. |
| What you give up | Nothing technically. But cash, and recruitment time you may not have. | The daily presence and the equity commitment of a partner. | The judgement and the ownership of decisions: you keep the burden of deciding without the means to do it. |
| When it is the right call | Five developers or more, a product at the heart of the business model, funding secured for eighteen months. | Zero to five developers, structural decisions to make now, not yet the means for a full-time leadership hire. | A defined scope, decisions already made, a need for extra delivery capacity. |
The figures in this table are orders of magnitude from the French market, given to place the options relative to one another. They are not our rates: ours appear below and depend on scope.
There comes a point where hiring in-house becomes the better answer, and we would rather say so before than after. It usually arrives around five full-time developers: day-to-day management, rituals, continuous hiring and looking after people then fill an entire role. Stretching a part-time arrangement beyond that produces a CTO chasing their own team.
You should also hire in-house when the technology is your competitive advantage in itself — an optimisation engine, a proprietary model, algorithmic know-how. The value sits in the head of the person building it, and that head needs to stay with you. We can help you hire them, and then step back.
Finally, if you have raised several million, your investors will expect internal technical leadership, and they will be right. The fractional CTO is a model for before and during: it takes you to a product that works and a team that holds, in the best possible conditions to then hire the right CTO — not under pressure, and knowing what to ask of them.
Loïc Guillebeau — Founder & Fractional CTO — Beyond The Brackets
Developer since 14, entrepreneur since 18. I started Beyond The Brackets in 2019 with one belief: technology should never hold back ambition.
In 2021, I co-founded Accelerio, a motorsport events startup. I lived every stage — from ideation to incubation, from product development to being on-site at 6am on event day. That experience taught me that building a great product is as much about business decisions as it is about code.
That is the difference I claim. When I tell you not to build a feature, or to ship with a smaller scope, it is not a developer's theoretical advice: those are calls I had to make for my own company, with my own money at stake and an event date that would not move.
Today, I bring that dual perspective — founder and developer — to other entrepreneurs. Since 2022, we have helped 30+ startups build their products. My role: be the CTO you haven't hired yet.
In practice: I am in your Slack, I know your roadmap, and I tell you no when it is warranted. And if the right decision is that you hire in-house, I will help you do that rather than keep myself on.
« My job is to make sure tech never holds back your ambition. »
Loïc challenged us on our architecture choice from the very first call. He saved us three months of development.
We treat him as our CTO. He is in our Slack, he knows our roadmap, he anticipates our technical needs.
I know nothing about tech. Loïc translates everything into business decisions I can understand.
Three formats, depending on whether you need technical leadership over time, a product delivered, or what already runs to keep running.
Most common
We are your tech team
Founders whose product is the core of the business, with no internal technical leadership.
We build your product
Companies with a defined project and a deadline, who need CTO-level thinking on the scoping rather than over time.
We keep things running
Companies with a product in production that need reliability and a technical contact who answers.
Hourly
One specific decision to settle: a stack choice, a quote to decode, an architecture to validate, a candidate to assess.
One day
A review of an existing product: architecture, code quality, security, debt, dependencies. Delivered with a prioritised action plan.
One day
One day to come out with a scope, an architecture and a budget that hold up. The deliverable stays usable even with a different provider.
One session
A session with your teams to identify the AI use cases that are genuinely profitable for you — and rule out the rest.
We do not publish rates: a price quoted out of context commits nobody and says nothing about your situation. These engagements are described by their deliverable and their length; the budget is scoped on the call, and the number we commit to comes out of the scoping day, once the scope is written.
Five steps, the first two of which leave you free to walk away with what you have learned.
Step 1
A 30-minute call, free and with no commitment. We look at your technical situation and identify the most important opportunity at this stage. You leave with a useful answer, even if we never work together.
Step 2
A scoping day or a technical audit, both paid. We produce a documented scope, architecture and budget. That deliverable is yours and remains usable with any provider.
Step 3
You choose a format. We open the shared channels, set the rhythm of the reviews, and define what we decide alone and what always comes back to you.
Step 4
Short cycles, a permanent staging environment, a monthly review of progress, budget and technical debt. You watch the product advance every week instead of discovering it at handover.
Step 5
The product runs. We move to ongoing support, prepare the hiring of your internal team, or start the next piece of work. In every case, the code, the accounts and the documentation are yours.
We turn down engagements for these reasons regularly. Better to read it here than discover it in month three.
In the French market, part-time CTO offerings start at around €3,900 a month for roughly five days of work, against the €90,000 to €130,000 annual total package of a salaried CTO in Paris. With us, the least committing entry point is an hourly consultation or a scoping day. What that comes to depends on the number of days a month and on whether a dedicated development team is involved — it is the first thing we scope on the call.
Generally between two and eight days a month. Two days is enough to steer an existing team and hold the judgement calls. Beyond eight, the real question becomes an internal hire, and we will tell you so. The volume is set at the start and then adjusted: scoping and launch phases need more presence than steady-state phases.
We write code. Beyond The Brackets is a development and design team first, and the fractional CTO role sits on top of that. Depending on the format, we deliver the product with our team, or we lead yours — code review, standards, architecture — without writing the features ourselves. Advice with no delivery capability is the main flaw of outsourced technical leadership offerings: a judgement call that is never implemented is worth nothing.
That is a normal outcome, and often the goal. We help write the job description, assess candidates technically and calibrate the package. Then we organise the handover: architecture documentation, the history of decisions and the reasons behind them, access, accounts and code. After that we reduce our presence to an occasional role, or to nothing. We never make the product's continuity depend on us being there.
That is the most common case. The work is translation in both directions: turning your business objectives into technical decisions, and returning every technical decision as cost, timeline and commercial consequence. You do not need to understand the code. You do need to understand the trade-offs, because they are yours — and it is our job to make them legible.
Yes, without difficulty, and before the first detailed conversation if you prefer. We also sign intellectual property terms: the code, hosting accounts, domain names and store accounts are created in your name and belong to you. An application published under its provider's account is an asset you do not control — we do not compromise on that, even when the client does not ask.
Beyond The Brackets is based in Paris, with a presence in New York. We meet Paris-region clients in person for scoping and structural reviews; the rest runs remotely in your channels, on shared Slack and a weekly video call. We work with clients across France and internationally: distance has never been the limiting factor — the frequency of contact is.
A conventional agency answers a specification. We start by questioning it, including when that reduces the size of the project. Concretely: an agency will price the twelve features you asked for; we will tell you which ones actually serve your objective at this stage, even if that removes half of them. The difference shows up at the moment someone has to tell you no — structurally difficult for a provider paid by scope delivered.
The first call is free and commits you to nothing. A technical audit or a scoping day are one-off pieces of work with no obligation to continue. For ongoing support we work quarter by quarter with short notice: the idea is that you stay because the value is there, not because a contract obliges you to. The most useful partnerships settle into six to twelve months, but that duration is observed rather than imposed.
30 minutes to audit your technical situation and identify your biggest opportunity. No strings attached — and you leave with a useful answer even if we take it no further.
Response within 24 hours.
