We get a version of this question on almost every discovery call: should this be a fractional CTO engagement or a consulting project? Most people writing about this topic answer it with a comparison table, a definition of ownership versus advice, and a cost range. That is useful shorthand, but it is not actually how the decision gets made. The decision gets made in the first thirty minutes of a scoping call, before either label matters, based on what is actually broken.
The diagnostic question, not the definitional one
The mistake we see most often is not hiring the wrong person - it is answering the wrong question. Founders ask "fractional CTO or consultant" as if it were a menu choice. The better question is narrower: is what is missing a decision, or a document? A team that needs someone to make the call and live with the consequences needs a different engagement than a team that needs a well-reasoned recommendation handed to leadership that already exists.
This question usually shows up after a company has already answered an earlier one - whether to bring in fractional leadership instead of a full-time hire at all. The fractional-versus-consultant call is the second fork, not the first.
This lines up with the distinction others in the category draw between an interim, a fractional, and a consulting engagement - the fractional model sits in the middle, more embedded than a consultant, more flexible than a full-time executive brought in during a transition. The label matters less than an honest answer to what is actually missing.
What we actually ask before recommending anything
Before we propose a structure, we run a short scoping conversation, usually thirty to forty-five minutes, built around a handful of specific questions rather than a generic intake form.
We ask what decision, made in the last quarter, did not get made or got made badly, and why. We ask whether the gap is a person problem, nobody with the authority to decide, or a knowledge problem, where the authority exists, but the depth does not. We ask what a successful ninety days looks like in terms a board would recognize, not just an engineering team. And we ask the uncomfortable one directly: is there already a capable technical lead in the room who just needs backup rather than replacement?.
The clearest live example right now is AI adoption inside the business. Tooling choices, governance, and what any of it does to the shape of the team are frequently owned by nobody, which is exactly why AI adoption tends to pull fractional CTOs into the room specifically rather than a consultant brought in to evaluate a single tool.
The answers sort the conversation fast. If the gap is a specific, bounded deliverable and someone internal is ready to receive and act on it, we are talking about a consulting engagement. If the gap is ongoing judgment with nobody accountable for the outcome, we are talking about a fractional CTO.
When we tell a prospect they don't need us
Turning down a fractional engagement is a normal part of this call, not an exception. A few situations come up often enough that we treat them as a standing checklist, in line with the same short list other operators in the category use to talk themselves out of an engagement.
| Situation | What it actually looks like | What we recommend instead |
|---|
| Capable leader already in place | An internal technical lead just needs backup, not replacement | Coaching or mentoring support, not a fractional executive |
| Delivery is healthy | The real gap is a budget conversation or a board narrative | Help with that specific conversation |
| No direction problem, just no hands | The team needs code written, not decisions made | Engineering capacity, not executive judgment |
Saying no here is not a courtesy. It is the same logic that runs through the rest of how we contract: the credibility of taking an engagement comes from being equally willing to decline one.
How we structure the engagement when it is the right call
When a fractional CTO is the right call, we do not start with a twelve-month retainer. We start with a short assessment, covering the architecture, the delivery process, the team's actual capability, and the specific risks that could stop the company - which matches the pattern most of the category has converged on: assess first, then move into an ongoing operating retainer rather than committing to open-ended leadership on day one.
| Phase | Typical length | What happens |
|---|
| Assessment | 2-3 weeks | Architecture, delivery process, team capability, and risk review, ending in a clear action plan |
| Operating retainer | Ongoing, reviewed monthly | Weekly hours, communication cadence, decision rights, and success metrics fixed up front |
| Advisory transition | First 1-3 months of a full-time hire | A light retainer for continuity once the company outgrows the model |
The retainer has a defined shape rather than an open-ended promise of availability. Weekly hours, communication cadence, decision rights, and success metrics are written down before the engagement starts and revisited monthly, not because we do not trust the relationship, but because vague scope is one of the most common reasons fractional CTO engagements quietly underperform. "Fifteen hours a week" is not a scope. Fifteen hours a week covering architecture review, technical hiring, and a weekly leadership meeting is a scope, and it is the difference between an engagement that compounds and one that drifts.
The blended model, done deliberately
Most engagements end up using both models at once, and that is fine as long as one of them holds the authority. A fractional CTO who owns the roadmap can commission a narrow consulting engagement - a security audit ahead of a raise, a framework evaluation, a one-off migration - and receive it the way an internal executive would receive a specialist's report. What does not work is running both without deciding who is accountable for the outcome, because then the specialist's findings sit in a deck with nobody positioned to act on them.
The evaluation questions we ask of a fractional candidate reflect this directly. We want a story about a company they said no to and why, a decision they got wrong and what it cost, and a reference willing to speak two years after the engagement ended rather than two weeks - roughly the same bar other operators in the space use to separate pattern recognition from a good pitch. A person who cannot describe a time they told a founder not to build something they wanted is usually operating as an advisor, not an owner, whatever the title on the contract says.
The exit is part of the plan.
A fractional CTO engagement that works eventually makes itself smaller. The clearest signal that the model is working is that it becomes obviously too small for the company - usually once the engineering team crosses roughly a dozen people, or technology has become the primary competitive lever rather than supporting infrastructure. At that point, the better fractional engagements taper into a light advisory retainer during the first months of a full-time hire rather than ending abruptly, so the new CTO gets continuity and the company does not lose the context that took months to build.
That exit path is written into the engagement from the start, not negotiated after the fact. It is also the clearest difference from a pure consulting or dev agency relationship, where the output is deliverables rather than an internal capability, and the institutional knowledge leaves when the invoice is paid.
Start with the diagnostic, not the label
If you are trying to decide between a fractional CTO and a consultant, the fastest way through it is not another comparison table. It is answering, honestly, whether what is missing from your technology function is a decision or a document. The structure, the pricing, and the exit plan all follow from that answer.
If you are earlier in this decision and want the classic signal-based version of the question, we have also written directly about the signs that point toward a fractional CTO.
If you want to walk through that scoping conversation directly, our CTO services page is the right place to start.
