The decision you’re actually making
Founders frame this as a hiring question. It’s closer to an equity-and-risk question: how much of the company, and how much calendar time, are you willing to trade for technical judgment you can trust?
There are five realistic paths: give equity to a technical co-founder, hand a spec to an agency (or buy delivery capacity without the leadership layer), retain a fractional CTO for judgment, take both together, or hire your own senior engineer. There’s also a sixth (building it yourself with AI tools) that deserves its own section further down.
None of these is wrong, and each is quietly expensive in its own way. Two of them are written up in full so you can see the shape before you choose: Campaign Finance, where we ended up co-founders, and Systellar Space, where we advised first and built after.
Four options, honestly compared
The equity co-founder.
- You get
- The technical half at co-founder depth, permanently, for 10 to 50% of the company.
- Right when
- The technology is the company: deep-tech IP, a novel system, an unfair technical advantage. That needs an owner, and owners cost equity.
- It fails on
- The search. Months of matching, and a mismatch unwinds slowly and expensively. A co-founder divorce costs more than any invoice.
The dev agency.
- You get
- A quoted price for a described scope. Capacity without commitment.
- Right when
- You can write the spec and defend it. The unknowns are in the execution, and the architecture calls have already been made by someone senior.
- It fails on
- Domain-expert founders rarely have a finished spec. An agency executes the brief it's given. When the brief is wrong, you pay to have the wrong thing built politely.
The fractional CTO, judgment only.
- You get
- Senior judgment for part of the week: architecture, hiring plans, vendor calls, roadmaps. No delivery attached.
- Right when
- You already have engineers who need direction, or you need an experienced voice in the room before the big commitments get made.
- It fails on
- The gap between deciding and shipping. You get decisions and a hiring problem. The team still has to come from somewhere, and assembling it is months of work that a few days a month will not cover.
The fractional CTO with a build team.
- You get
- The judgment and the hands under one retainer. Decisions get made, then shipped by the same people who made them.
- Right when
- You have the domain depth and the funding, but no engineering org and no time to assemble one. You want one accountable party from architecture to production.
- It fails on
- Concentration. One firm holds a lot of context, so the exit has to be designed in: documentation, handover, code your future in-house team can inherit. Ask any studio how engagements end before you ask how they start.
The same options, in numbers
Words compare badly. Here is the same decision as a table, with the numbers people actually plan around: what the money looks like, how long the wait is, and what you still own when the engagement ends.
The market ranges come from published fractional CTO marketplace and agency pricing pages, rounded, because they move. Our own numbers do not move.
| Technical co-founder | Dev agency | Fractional CTO only | Fractional CTO + build team | |
|---|---|---|---|---|
| What you pay | Equity, 10 to 50%, often with a reduced salary | A fixed price per milestone | A cash retainer for part of a senior week | One retainer for judgment and delivery |
| Typical cost | Nothing up front, the whole bill later | $15k to $80k+ for an MVP | $3k to $25k a month | Quoted after a scoping call |
| Time before work starts | Two to six months of searching | Two to six weeks to kick off | Days to weeks | One scoping call |
| Time to a working product | Whatever you lose by not finding the right person | Eight to fourteen weeks | Not the deliverable | Written into the scope |
| What you own | Nothing without vesting and a cliff | The code, if the contract is clean | Decisions, and a hiring problem | Code, docs, and the reasoning behind both |
| Where it breaks | The search, then the divorce | The brief nobody could write | The gap between deciding and shipping | Concentration, so the exit is designed in |
| What exiting costs | Dilution, vesting, and a hard conversation | Handover gaps and a new vendor | Stopping the retainer | A handover we owe you either way |
Two of the columns above are ours: CTO only, and both together. The audit is the only one with a fixed price. Everything else is quoted after we hear what you are building.
The other option:
build it yourself with AI.
In 2026 this is a real option, and for validating demand it’s often the right one. A prototype that proves people want the thing is worth more than a deck.
The limit shows up around month three. AI writes code much faster than it writes architecture, and the prototype that won your first users starts failing your next ones: hallucinated APIs, no types holding the seams, security decided by autocomplete. Rebuilding at that point is normal and plannable. It is a much worse thing to find out halfway through an investor’s technical due diligence.
If you’re there now, an audit answers what to keep, what to rebuild, and whether the codebase can carry the plan. Whoever you end up hiring.
And the last one:
your first senior hire.
Look at the table again and this one seems obviously best. You post a role, you interview, you pay a salary, and the person shows up every day. Compared with a monthly retainer it is a tidy line on a spreadsheet, and in the perfect world that is what you do: hire early, grow the team from one.
Two things have to be true first, and most founders skip both.
The first is that you can direct the work. A senior engineer needs someone to decide what gets built, in what order, and what gets cut when the money runs short. If that someone is you, and you can defend the calls without a technical background, hire. If it is not you yet, you become the technical manager of a person more technical than you are, and that is a job nobody does well from reading about it.
The second is time. Allow three to six months from opening the search to a start date, because good engineers are employed and not in a hurry. You pay runway through the whole search. Then the thing they build still has nobody above them reviewing the architecture, comparing the vendor quotes, or telling you when the plan is wrong.
An employee is the cheapest way to buy hours and the most expensive way to buy a decision.
Which is a strange thing for us to write, because a hire is a thing you keep and a retainer is a thing you stop. Both are true. If you can direct the work, hire, and we will tell you so on the call.
More on the first hire, including who reviews their work and what the handover has to contain.
How to choose
- If the technology is the company, find the equity co-founder and take the months it takes to find them.
- If you can write and defend the full spec, get agency quotes. It's the cheapest path to a known thing.
- If you have engineers who need direction, a fractional CTO for a few days a month is enough.
- If the calls are already made and you only need throughput, buy delivery capacity without the leadership layer on top of it. That is an agency or your own hires; it is not a shape we sell.
- If you can direct the work yourself and the calls have already been made, hire your first senior engineer and pay them a salary.
- If you have domain depth and funding but no engineering org, you want the CTO with the build team attached.
- If AI built your prototype and you're deciding what's next, start with an audit, whoever you end up hiring.
Two of the shapes above are ours: the fractional CTO on its own, and the two together. That’s the lens this guide was written through, and it’s also why the failure modes above are specific: we’ve usually been hired just after one of them. How we build, and the part AI plays in it, is on the stack page.
How to vet whoever you pick
Every one of these models has an ending. Ask what the ending looks like before you sign anything, because that is the part nobody puts on a pricing page.
- Ask to see code from something that shipped 12 months ago. Not the case study page. The repository, and who still understands it.
- Ask who reviews the senior work. On an agency engagement, get the name of the person who signs off architecture. On an advisory retainer, get the name of the person who reads the code.
- Ask what the last engagement that ended badly looked like, and what changed afterwards. A clean answer to that question is a rehearsed one.
- Ask for the handover plan now, in writing: documentation, repository access, and who owns the accounts. A handover discussed at the end is a negotiation.
- Ask what they would refuse to do for you. A partner who never says no is selling hours rather than outcomes, and you will find out which one at the invoice.
The fifth question is the one that hurts, so here is ours. We do not write a fixed-price MVP for a founder who cannot yet describe the product, we do not take delivery work where nobody on our side can make the architecture calls, and we turn down teams that want AI to write the whole codebase unsupervised. Those are the three conversations where we say no and point somewhere else.
One number worth keeping in mind while you compare vendor stories. Stripe found developers spending 42% of the working week on maintenance, technical debt, and bad code, which is most of a week lost to decisions nobody made on purpose. That is the bill this whole comparison is trying to avoid, and it is cheaper to pay for judgment now than for the rewrite in two years.
Questions founders ask
What is the difference between a fractional CTO and a technical co-founder?
A co-founder owns the technical half of the company and is paid in equity, permanently. A fractional CTO is a senior technical leader you retain for part of their week and pay in cash, for as long as the engagement makes sense. The co-founder arrives with the cap table and, eventually, an exit conversation. The fractional CTO arrives with an invoice.
Is an agency better than a technical co-founder?
Only if you can write and defend the spec yourself. An agency executes the brief it is given, at a price you can compare, and it ends at handover. It cannot tell you the brief is wrong, because that is not what you bought. If the architecture calls have already been made by someone senior, an agency is the cheaper path. If they have not, you are paying to have the wrong thing built politely.
Do I have to give up equity to get a technical partner?
No, but the alternative is cash, and cash runs out. We take no equity and no revenue share. A co-founder takes 10 to 50% of the company, which at a modest valuation is the most expensive single hire you will ever make, and the right call when the technology is the company rather than a delivery of it.
How much does a fractional CTO cost?
Published ranges in 2026 sit between $3,000 and $25,000 a month, with day rates around $1,500 to $4,000. The variance is mostly about whether anyone builds. A few days a month of judgment is the cheap end. Judgment plus a team that ships is the expensive end, and it is still less than the salary of one senior engineer.
Can I combine an agency with a fractional CTO?
Yes, and it is a common structure. The fractional CTO writes the spec, picks the vendor, and reviews what comes back. That arrangement works, with one gotcha: you are now paying two invoices for one outcome, and the accountability question does not disappear. When the build is wrong, the review you paid for is part of what failed.
Can I engage a fractional CTO without the build team?
Yes, and it is a smaller engagement. The CTO works part of the week on architecture, hiring, and the calls you should not be making blind, while you keep whatever engineers you already have. The tradeoff is the one on this page: decisions do not ship themselves, so the gap between deciding and building stays yours to close.
Can I hire the build team without the CTO layer?
That is the agency model, and it is not a shape we sell. If the architecture calls are already made and what you need is throughput, an agency or your own first hire covers it, and both are cheaper than a partnership for the same reason they are riskier: nobody is in the room to tell you the direction is wrong.
How long does it take to find a technical co-founder?
Months, usually, and you should plan for it to be longer than the search window you set yourself. Most working co-founder pairs met before they were co-founders. If you are starting the search cold and you have budget, a retained engagement buys you the technical judgment while the search runs, and you can stop it when you find the person.
What if my prototype was built with AI?
That is the normal 2026 case. A prototype that won users is worth more than a prototype that was built correctly. The question is whether the codebase can carry the next stage, and that is a fixed-scope read rather than a rebuild. Ours is €4,900 and takes 7 business days. It is also the cheapest way to test our judgment before you commit to anything bigger.
The 42% figure on this page is from Stripe’s Developer Coefficient report, 2018. The market ranges in the table are rounded from published fractional CTO and agency pricing in 2026, and they move.
Can’t decide?
One question splits the cases the rules above don’t settle. Can you make the architecture calls yourself?
If you can, and you already have engineers who need direction, you are buying judgment, and the CTO on its own covers it. If you can, and what you lack is throughput, that is a hiring problem rather than an advisory one: bring the engineers in and keep directing them yourself. Capacity against a spec you hand over is the agency model, and it is not a shape we sell.
If you can’t, the judgment has to arrive from outside, and it arrives with a hiring problem attached. That is the shape we are usually hired into, and the honest framing is that it costs more than the other two. The reason it still wins is narrow: assembling a team is months of work that a few days a month will not cover.
Start with the audit if you want to test that read against your own codebase before committing to anything longer. It is €4,900, fixed scope, and it comes back in 7 business days.