The Economics of One CTO Across a Portfolio

Most funds treat technology leadership gaps as one-off emergencies. A CTO quits, a scramble begins, a provider gets sourced, an engagement happens, everyone moves on. Eighteen months later it happens again at a different portfolio company, and the scramble starts from zero. The firms I work with repeatedly have stopped doing this, and the reason is arithmetic.

What sourcing fresh actually costs

Price the scramble honestly. There's the search itself: weeks of partner time collecting referrals, screening candidates, and negotiating terms, for a role most funds fill a handful of times across a fund's life and never build real muscle for. There's the calendar cost: four to eight weeks between "we need someone" and "someone is in the seat," during which the portfolio company makes no hard technical decisions. On a five-year hold, every idle month is roughly two percent of the value-creation window, spent on procurement.

Then the part nobody prices: every new interim spends their first weeks learning how your firm works. How you report, what your ICs care about, what your deal models assume about technology, how much truth your partners actually want in the room. That onboarding is invisible and you pay for it, at full rate, every single time you start over with someone new.

What compounds when you don't

Now run the alternative: one operator the fund knows, engaged directly whenever a portfolio company needs its cost base reset, its delivery rebuilt around AI, or the seat covered. The second engagement starts faster than the first, and by the third the pattern is genuinely different.

• Context carries. I already know how the firm underwrites deals, what its board decks look like, and which partner wants the bad news first. Day one of company two is what day twenty was at company one.
• Trust is pre-built. When I say a thesis workstream won't make its date, a firm that has watched me be right before acts on it in a week instead of seeking a second opinion for a month.
• The playbook repeats. Delivery metrics, team assessment, the AI-native operating model: I install the same proven system each time, tuned to the company. Repetition makes it faster and better, and the fund can compare companies on the same yardstick.
• Diligence gets teeth. An operator who might personally have to execute the 100-day plan reads a target's technology differently than a report-writer. My due-diligence findings become someone's job, occasionally mine, and that keeps them honest.

That last point deserves a sentence more. The move from technical due diligence before the deal to turnaround CTO after it is the most natural handoff in this business, and it only exists when the fund's provider is a person rather than a bench.

The AI-native multiplier

Here's where the economics stop being about sourcing efficiency and start being about the exit multiple. The biggest value-creation lever available to a software portfolio company right now is becoming genuinely AI-native: engineering throughput that looks like a team twice the size, support and internal ops running on agents, unit economics that make the next buyer's model sing. Most portfolio companies are nowhere near this. Their boards know it, and their eventual acquirers will price it.

Installing that operating model is exactly the kind of work an interim mandate is shaped for: intense, bounded, transformative, and not requiring a permanent hire to maintain once it's running. I've written about what becoming AI-native actually involves, and the portfolio angle is straightforward. A fund that runs the same AI-native transformation through one operator across four companies isn't buying four consulting projects; it's buying one playbook, amortized, with each run better than the last. The first company pays for the learning. The fourth gets it nearly free.

What the math misses

I'll argue against my own spreadsheet for a paragraph. The one-operator model has a real constraint: one person is full-time in one company at a time. If three seats empty in the same quarter, the fund is sourcing for two of them anyway, so the relationship is a first call, never a guarantee of infinite capacity. And a standing relationship needs the same honesty as a new one; the day a fund keeps an operator around out of habit rather than results, the compounding runs in reverse. Familiarity should sharpen scrutiny of the work, not replace it.

But held to that standard, the model wins on every line: cheaper sourcing, faster starts, honest diligence, a repeating value-creation playbook, and a technology story at exit that was built deliberately instead of patched between emergencies.

If your fund is on its second or third technology-leadership scramble, that's the signal to stop treating this as an emergency category. Let's talk about being your first call →