Every technical engagement forces a billing choice, and founders usually make it on instinct. Milestone billing feels safe: you pay when something ships, so you never pay for nothing. A monthly retainer feels loose: you pay the same whether the month was busy or quiet. The instinct is often backwards. The right structure depends entirely on whether you are buying output you can define in advance or judgment you cannot, and picking the wrong one creates the exact problems it was meant to prevent.
Here is how each one actually behaves once real work starts, and which to use for which kind of work.
What milestone billing is good at, and where it breaks
Milestone billing ties payment to deliverables: define a milestone, hit it, invoice it, repeat. Its strengths are real. You keep budget visibility because you know the milestones and their prices upfront. Cash leaves your account only against something you can see. For a well-defined build with a fixed scope, it aligns everyone on shipping and gives you natural checkpoints to stop if things go wrong.
Where it breaks is scope that moves and value that is subjective. Milestone billing needs milestones that are easy to define and measure. The moment a deliverable is fuzzy, "improve the architecture," "advise on hiring," "get us diligence-ready," you get delayed and contested payments, because the two sides disagree about whether the milestone was actually met. It also carries heavier overhead: documentation, tracking, and approval workflows for every milestone. And it is explicitly a poor fit for advisory work, fluid scopes, or anything whose value is a judgment call rather than a shipped artifact.
There is a subtler failure. Milestone billing quietly pushes both sides toward defining success as "the milestone shipped" rather than "the right thing shipped." That is how you end up paying, on time and in full, for a feature that met the spec and helped no one. It is a cousin of the fixed bid that quietly becomes 60 percent more: the structure rewards hitting the letter of the scope, and every change to the spec becomes a negotiation.
What a retainer is good at, and where it breaks
A monthly retainer is a flat fee for ongoing availability and ownership. Its strengths are the mirror image of milestone billing's weaknesses. It handles fluid scope gracefully, because you are not renegotiating a price every time the work shifts. It removes the friction of pricing each interaction, so you actually ask the small questions that turn out to be the important ones. And it builds the kind of ongoing relationship where the provider acts as a trusted advisor rather than a vendor counting deliverables.
Its failure mode is the opposite one. On a quiet month, you can feel like you are wasting money, and if the scope of what is included is unclear, scope creep runs the other way, with the provider absorbing more and more until the fee stops making sense for them. A retainer only works with a clearly defined scope of what is in and what is out, and with enough real activity that the flat fee maps to real value. That is the same reasoning behind why I don't bill by the hour: a retainer is the right instrument specifically when the value is judgment and access, not a countable pile of output.
The rule: match the structure to what you are buying
Strip it down to one question. Are you buying output you can specify in advance, or judgment you cannot?
If you can write down exactly what "done" looks like, a discrete build with clear deliverables and a fixed scope, milestone billing fits. It gives you checkpoints, budget control, and a clean way to stop. Use it for project work: build this integration, ship this redesign, migrate this data.
If the value is ongoing decisions, direction, and availability, where "done" is not a fixed artifact and the scope will move week to week, a retainer fits. Use it for leadership: someone owning your technology strategy, managing your vendors, keeping you diligence-ready, being reachable when a decision cannot wait. Fractional CTO work is almost always this second kind, which is why serious engagements are priced as retainers and not per deliverable.
The hybrid that usually wins
You do not have to pick one for everything. Many founders run both: a retainer for the ongoing judgment and a separate milestone or project fee for a specific, well-defined build inside that relationship. The retainer keeps a senior person accountable for direction and for challenging scope; the milestone structure keeps a discrete build honest and stoppable. That combination gives you predictability where you need it and checkpoints where you can define them, without forcing advisory work into a deliverable it does not fit.
If you are trying to decide how to structure and price a specific engagement rather than a one-off build, the way I scope and price ongoing work is laid out on the pricing page.
FAQ
Is milestone billing cheaper than a retainer?
Not inherently. It can look cheaper because you only pay against deliverables, but the overhead, the change-order negotiations, and the risk of paying in full for the wrong thing often make it more expensive for anything with a moving scope. Cheaper depends on whether the work is genuinely fixed.
Can advisory work be milestone-billed?
Poorly. Advisory value is a judgment call, not a measurable deliverable, so milestones become contested and payment stalls. Advisory and leadership work belongs on a retainer; save milestones for concrete builds.
How do I stop a retainer from feeling like wasted money in slow months?
Define the scope of what is included clearly, and judge the retainer over a quarter, not a week. The quiet months are paid for by the months where one fast decision saves you far more than the fee. If quiet months persist for a quarter, the scope is wrong and you should resize it.
What if I genuinely need both?
Run a retainer for ongoing judgment and add milestone pricing for a specific defined build. That hybrid is common and usually the most cost-effective structure. If you want help deciding which parts of your work go where, book a call.