Firing a fractional CTO who is not working out feels close to free. There is usually a short notice period, no severance, no layoff paperwork. But the real cost of switching is not the fee. It is the context that walks out the door, the handover that never quite happens, and the month or two your technical decisions stall while a new person learns what the last one already knew. Plan for that and switching is survivable. Ignore it and you pay for the same ramp-up twice.
The bill you see is the small one
The visible cost of switching is the contractual one: a notice period, maybe a final month you pay while barely using them, and any unused prepaid time. For most engagements that is a few thousand dollars, and it is not the part that hurts.
The expensive part is invisible on the invoice. A fractional CTO who has been with you for six months holds a large amount of undocumented context: why the architecture is the way it is, which vendor conversations are half-finished, what your last agency did badly, which of your engineers needs watching, and what you already decided and rejected. When they leave, most of that leaves with them unless you force it out into writing. The replacement does not start where the last one finished. They start closer to zero.
What the ramp-up actually costs
A new fractional CTO needs four to eight weeks to be genuinely useful, and you are paying full rate the whole time. At $8,000 to $12,000 a month, a two-month ramp is $16,000 to $24,000 spent mostly on someone getting oriented rather than moving your product forward. That is the true price of a switch, and it is roughly the same whether the switch was your idea or theirs.
Worse is the decision stall. While the new person ramps, hard technical calls tend to wait, because nobody wants to make an irreversible bet in their first two weeks. If you switch during a critical build or a fundraise, that pause can cost far more than the ramp itself. The architecture and decision risks that are hard to undo do not politely wait for your new hire to settle in.
When switching is still the right call
None of this means you should tolerate a bad engagement. A fractional CTO who is disengaged, spread across too many clients, or simply wrong for your stage will cost you more every month than a clean switch will. The point is to switch deliberately, not reactively, and to make the outgoing person earn their final invoice by doing a real handover.
The signals that justify eating the switching cost: repeated missed or vague commitments, decisions that never get made, an inability to explain technical risk in terms you understand, or a scope that has quietly drifted far from what you agreed. If you scoped the engagement well at the start, drift is easy to see. If you did not, this is a good moment to read how to scope a fractional engagement so the next one is measurable.
How to switch without paying twice
Make the handover a deliverable, not a courtesy. Before the outgoing fractional CTO leaves, get a written architecture overview, a list of open decisions and their state, vendor and access inventory, and their honest read on your team. Pay for that time; it is the cheapest money you will spend all quarter. Overlap the two people by a week or two if you can, so context transfers person to person rather than through documents alone.
Then scope the new engagement tighter than the last one, using what you just learned about what went wrong. A clean switch with a real handover costs you a ramp period once. A panicked switch with no handover costs you the ramp plus months of rediscovery, and often a second switch a few months later when the new person hits the same undocumented walls. If you are weighing a switch right now, book a call and talk it through before you give notice.
FAQ
How much does it cost to replace a fractional CTO?
The contract cost is usually small, a notice period and any unused time. The real cost is a four-to-eight-week ramp at full rate, roughly $16,000 to $24,000, plus the value of decisions that stall while the new person learns. Budget for the ramp, not the fee.
Should I overlap the outgoing and incoming fractional CTO?
If you can afford a week or two of overlap, yes. Direct handover transfers context that documents miss, and it shortens the new person's ramp. It is almost always cheaper than the rediscovery you pay for when they leave cold.
How do I know it is the engagement and not the model that is failing?
Look at whether decisions are being made and explained. If the person is engaged but wrong for your stage, switch to a better-fit fractional CTO. If no fractional arrangement is getting you decisions, the problem may be scope or your own availability, not the person. The comparison of engagement models can help you tell which.
Can I avoid switching costs by keeping better documentation?
Partly. If you require running documentation from day one, the handover is faster and cheaper and you are less hostage to any one person's memory. It does not eliminate the ramp, but it can cut it in half.