Fractional CTO vs full-time: count the decisions
You are running a profitable company. A couple of hundred people, maybe more. Technology is in everything you do, but nobody senior owns it — and now there is a decision on your desk, possibly a job posting already out the door: do we hire a CTO?
Before you spend nine months and three hundred thousand pounds finding out, it is worth asking a sharper question: do you need a CTO in the building, or do you need CTO-level decisions made well? Those are different problems with different answers. I have been the CTO of a financial services firm for seven years — a few days a month, not five days a week — so here is how I would think about it from your side of the desk.
The decision-count test
Strip away the job title and ask what a CTO actually produces. Not code — your engineers and vendors produce code. A CTO produces decisions. Which vendor to trust. Whether to build or buy. Which security trade-off is acceptable and which one ends up in front of a regulator. Which architecture bet you will still be living with in five years.
So count the technology decisions your company made last quarter. Not tickets, not “the printer server fell over” — decisions where getting it wrong costs real money or real time twelve months from now. A 200-person logistics firm might choose a transport management system, weigh a warehouse software renewal, and debate an AI pilot on customer service email. Three decisions — big ones, but three. A software company shipping to ten thousand customers makes architecture, hiring, tooling and platform calls weekly. Those are two completely different jobs, and it is a mistake to solve them with the same hire.
My rough line: fewer than about ten genuine technology decisions a quarter, sustained over time, and you do not need a three-hundred-thousand-pound seat filled five days a week — you need those ten decisions made well by someone with the scars to make them. Consistently over ten, and you are heading towards a genuine full-time need, which I will come back to honestly. The exercise: with your leadership team, list every technology decision from the last two quarters, strictly defined — if a competent IT manager could resolve it, it does not count. The length of that list, and whether it is growing, tells you most of what you need to know before you speak to a recruiter.
The real cost maths
The salary line is the smallest part of it. A credible full-time CTO for a mid-market company runs £250,000 to £350,000-plus in total compensation, with a recruiter’s fee of 25 to 30 percent of first-year salary on top. The search itself takes six to nine months, during which every decision you just counted is stalled or being made by whoever is nearest.
And the number nobody puts in the business case: the cost of the wrong hire. Executive hiring at this level fails more often than anyone admits, and you do not find out in month two — you find out in month ten. By then you have paid a year of a very large package, lost a year of decisions made badly or not at all, and you pay the recruiting fee again. Realistically, a failed CTO hire costs a year of momentum and north of half a million pounds all-in.
Compare the fractional arrangement: a few days a month of genuinely senior attention at somewhere between a tenth and a quarter of a full-time package, starting in weeks because there is no search. The part boards tend to like most is the risk profile — if it is not working, you are out one month’s fee, not one year’s package. The arrangement either earns its keep every month or it ends.
What the days actually buy
“A few days a month” sounds thin until you look at what fills them. Vendor evaluations with someone on your side of the table asking what the salesperson hopes nobody asks — one properly run negotiation frequently pays for the year. AI adoption calls: sorting the genuine proposals from last year’s product with a new label, and defending what you are not doing yet. Security posture — owning the questions, not running the firewall. Architecture review before you sign the ERP contract or the platform migration. Hiring guidance for technical roles, because non-technical executives are easy to impress in a technical interview. And board translation — turning “we need to re-platform” into risk, cost and time a board can weigh — which is quietly the most valuable of the lot.
When full-time is genuinely right
This test has to cut both ways or it is a sales pitch. Four cases where you should hire full-time without apology: technology is the product, in which case the CTO is a product leader and always a five-day job; you are genuinely and sustainably over the decision threshold, with the trend upward; you have a growing engineering team that needs daily coaching and unblocking, which does not compress into a few days a month; or you have an M&A-heavy roadmap, where technical due diligence and integration is sustained, intensive work.
One caution if you do hire: for a non-technical CEO, a CTO interview is the highest-stakes interview you will ever run with the least ability to score it. Get someone senior and independent — with no stake in which candidate wins — into the final round and the technical references. Do not run that interview alone.
So: count the decisions, do the honest maths on both columns, and if you are truly over the threshold, hire properly.
If you are weighing this decision and want a second pair of eyes on your specific numbers, book a 20-minute call. Either answer is fine — if the count says hire full-time, I will tell you that, and what to look for when you do.