The CTO You Can Actually Hire

An experienced part-time CTO and a capable engineering manager may cover your needs better than a full-time executive you can barely afford. But if “capable engineering manager” describes someone you still have to recruit, you’ve priced half an arrangement.

The case for fractional leadership starts with a specific advantage: access to someone’s relevant experience without buying all of their working time. It holds only if the remaining work has an owner. Otherwise, the founder gets a new adviser and keeps the old job.

What happens after the advice

Suppose your team is choosing whether to build subscription billing or use a vendor. An experienced CTO can examine the contract, the cost of maintaining a custom system, and the ways either option could fail. That decision can be prepared for a scheduled review, with the evidence available beforehand.

Choosing the vendor leaves plenty of work. Someone must lead the integration and handle questions the review didn’t anticipate. If your engineering manager can own that work and knows when to bring the CTO back in, fractional leadership has a workable boundary.

Now suppose sales proposes a billing arrangement the vendor can’t handle. An engineer spots the limitation. The manager can assess a workaround, but changing the commercial promise may require the founder’s agreement. Before the team builds around it, someone has to bring that new constraint to the people who can change the decision.

The boundary breaks if those people are unavailable and the manager lacks authority to act. The team either waits, proceeds on an assumption, or pulls the founder back into engineering. In this example, the useful measure is how long the decision can safely wait and who can act before that point. Counting decisions on a roadmap won’t tell you that.

Writing down the reasoning helps. In his 2011 essay Documenting Architecture Decisions, Michael Nygard proposes short records that preserve a decision’s context, status, and consequences. He also keeps superseded decisions so a new person can understand why the team made the old choice.

For the billing example, such a record could explain which requirements the vendor meets and which limitations the team accepted. Add who can approve an exception and what would trigger another review. The document gives people something better than a remembered instruction from a meeting; someone still has to notice when reality changes.

A fractional CTO can hold executive authority and stay involved between meetings if the engagement provides for it. Ask which decisions they would own and what response commitment the fee includes. There must be capacity to investigate an unexpected problem as well as acknowledge the message. A full-time CTO also delegates and becomes unavailable; their team needs a way to act in their absence, too.

Experience still needs local context

Experience can tell a CTO which questions to ask. Your company supplies the answers.

Will Larson’s 2024 essay Layers of context examines proposals to introduce a programming language. A project team sees tools suited to its problem. Infrastructure has to support the language, while engineering leadership has to consider the hiring and training implications. Larson describes thinking a new language was a good idea earlier in his career, then pushing back on a later proposal as his understanding of those wider obligations grew.

The hiring implication is narrower than “an experienced person will decide faster.” Relevant experience can expose a cost your team missed. Applying it still requires learning who will support the system, what else those people are responsible for, and whether your budget can sustain the choice.

A fractional CTO needs direct access to engineers and business constraints, including room to investigate beyond the founder’s summary. That time belongs in the engagement. A leader who sees only a polished status report can give a polished answer to the wrong problem.

Full-time presence has value here: it gives the leader more opportunity to observe changing requirements and hear concerns before someone turns them into an agenda item. Presence alone guarantees little, but treating all of that time as executive overhead makes the comparison dishonest.

When you assess a fractional candidate’s track record, ask how much involvement it took to make the decisions they’re describing. A good call made while fully embedded tells you little about how they’d stay informed on the schedule you’re buying.

Compare the whole job

Daily management deserves its own place in the calculation. In Who runs Engineering processes?, Larson recalls running engineering performance reviews out of a spreadsheet during his first year as Calm’s CTO because there were no shared people processes. Someone had to do the work, and he did it.

That account illustrates a limit of buying only the high-stakes calls. If hiring, performance feedback, and delivery coordination need an owner, removing them from the CTO’s scope leaves them on somebody else’s calendar. Count that person’s capacity and cost.

Then compare the experienced fractional leader against the full-time person you can actually hire at compensation the company can sustain. A candidate who needs support in architecture may be excellent at building your team. A fractional candidate with deep architecture experience may have little capacity for the management work you need. Titles hide those differences; actual responsibilities expose them.

Put recurring costs on the same basis, such as a month, before comparing quotes. For the fractional arrangement, add the fee, any extra management capacity you must hire, and outside help still needed beyond the quoted scope. For the full-time arrangement, add salary, employer-paid benefits and taxes, plus any management or specialist help the candidate would still need. Keep one-off recruitment costs separate from recurring spend. Record any equity commitment alongside cash rather than treating it as free.

If the same engineering manager stays under either option, put that salary on both sides. It cancels out of the difference. Charging it only to the fractional option makes full-time look artificially cheap.

That manager’s available time still matters. If coordinating with the CTO displaces delivery work, identify what gets delayed or who will take it over. List the founder’s remaining decisions, too, along with the work they’ll displace. Keep this capacity question beside the cash total rather than manufacturing an hourly price for it.

If a trusted engineering lead already runs delivery and the unresolved question is confined to a security review or a database migration, a bounded specialist engagement may be enough. If a strong internal manager needs occasional help with unfamiliar decisions, coaching may serve better than adding another person with authority over the team.

Full-time leadership has a stronger case when the work requires more sustained involvement than the fractional candidate can commit to. The billing question could keep reopening as customer promises change, forcing the CTO to revisit technical choices and staffing together. A small team can have that need. A larger team with capable managers can leave more room for scheduled strategic work. The number of engineers alone tells you too little.

Before signing either agreement, take a consequential decision your team has actually faced and walk the candidate through what happened after it. Ask how they would have learned that the assumptions had changed, what authority they would have needed, and who would have acted in their absence.

Who owns the next morning?