A founder emails asking for a fractional CTO. We talk for twenty minutes, and it turns out they do not need one. They need a four-week project: a hard look at their platform, a build-versus-buy recommendation, and a handover. After that there is nothing for a part-time executive to do, and paying one a monthly retainer would be paying for a relationship when they only needed a decision.
The reverse happens just as often. A founder asks for a one-off review, and the real need is someone who stays for six months to hire a team and make sure the rebuild actually lands. A single report would be a document on a shelf nobody is accountable for.
Knowing which one you are is the whole game, because the two are priced completely differently and solve completely different problems.
What separates a project from an ongoing engagement
The line is not size or difficulty. It is whether the deliverable is a decision or an outcome.
A project ends with a document and a decision
Project work is time-boxed and produces an artifact: a roadmap, a build-versus-buy call, a technical memo for diligence, an architecture review. The deliverable is the thinking, and once it is handed over, the engagement is genuinely complete. Nobody needs to stay. Common examples are a strategy sprint to assess your platform and map dependencies, or diligence support before an acquisition where someone evaluates a target and writes it up. These are usually four to eight weeks and a fixed scope. Pricing typically runs anywhere from $5k to $50k depending on depth, because you are buying a bounded piece of senior thinking, not a standing seat.
An ongoing engagement ends with an outcome someone owned
Retainer work is different in kind. You are not buying a decision; you are buying someone to stay, make the call, and be accountable for whether it worked. Building and shaping a team. Owning architecture as it evolves. Being the technical voice in the room week after week as the company changes around it. The tell is the phrase "and make sure it works." If you need someone to stick around and be on the hook for the result, that is a retainer, not a project. Ongoing fractional engagements commonly run a few thousand to low five figures a month depending on days.
There is a caveat that decides whether ongoing even makes sense: authority. A retainer where every technical decision still needs your sign-off is just expensive advice. If you are not ready to let someone own the calls, you may actually want the project version, where the decision is the product and you stay in control of execution.
How to tell which one you need
Run your situation through three questions before you decide what to buy.
Is there a single decision at the center, or an ongoing stream of them?
If you can name the one thing you need answered -- should we rebuild, is this architecture sound, what should the roadmap be -- that points to a project. If the honest answer is "a new technical decision shows up every week and nobody good is making it," that is the ongoing version. One question is a project. A faucet of questions is a retainer.
Does someone need to be accountable after the decision?
A decision with no execution risk can be handed off and closed. A decision that will go wrong in the implementation -- a migration, a team build, a security overhaul -- needs an owner through the messy middle. If failure happens in the doing rather than the deciding, you want someone who stays. This is the same logic behind scoping a fractional engagement before it quietly becomes full-time: name the outcome someone is accountable for, or you will not know when the work is done.
Is the need temporary by nature or permanent by nature?
Some needs have a natural end: a raise, a diligence process, a one-time platform choice, a handoff from an agency. Others do not end until you hire a full-time leader: ongoing architectural judgment, team management, being the technical half of the leadership table. Temporary needs are projects. Permanent needs are ongoing engagements, and at some point they become the case for a full-time hire instead. If you want a clearer map of what these arrangements cost and how they are structured, our pricing page lays out the difference between a scoped project and a monthly engagement.
The expensive mistake in either direction is real. Buy a project when you needed an owner, and the report gathers dust while the actual work drifts. Buy a retainer when you needed a decision, and you pay monthly for a person with nothing left to lead.
Frequently asked questions
Is project-based or retainer fractional CTO work cheaper?
It depends on duration. A bounded project can run from a few thousand to tens of thousands as a one-time cost, while a retainer is a recurring monthly fee. For a single decision, the project is far cheaper. But a project stretched into months of ongoing involvement usually ends up more expensive than just running a retainer, because you keep re-scoping and re-buying the same person. Match the model to the shape of the need, not to which number looks smaller today.
Can a project turn into an ongoing engagement?
Often, and that is healthy when it is deliberate. A diligence project or a strategy sprint frequently surfaces work that genuinely needs an owner, and converting to a retainer at that point is a real decision based on what you learned. The trap is drifting into it by accident -- a four-week project that quietly becomes a permanent presence nobody re-scoped or re-priced.
How do I avoid overpaying for the wrong model?
Start by writing down the single sentence that describes what you need. If it is a decision or a document, buy a project and stop there. If it is "someone to own X and make sure it works," buy the ongoing version. When you genuinely cannot tell, a short scoped conversation will sort it -- book a call and we can name the deliverable together before any money changes hands.