Milestone-based billing and the cash flow it fixes
How to structure milestone billing so payment tracks delivery, why the gap between approval and invoice is the most common cash problem in services, and what to do about late payment across HK, SG and the US.
A profitable services business can still fail, and the usual mechanism is timing. You pay salaries monthly and get paid on terms that are nominally thirty days and functionally fifty-five. The gap between those two facts is what kills otherwise healthy firms.
Milestone billing is the main structural lever available. Done properly it ties payment to verifiable delivery, shortens the gap between work and cash, and removes most of the ambiguity that makes invoices easy to defer.
Why a completion-based invoice is easy to defer
Invoicing at the end of a project concentrates all of your commercial risk at the point where your leverage is lowest. The work is delivered, the client has what they needed, and the only thing left is a payment that competes with every other demand on their finance team.
It also makes the invoice a large number arriving at once, which is exactly the kind of number that gets routed for additional approval. A single fifty thousand dollar invoice attracts scrutiny that five ten thousand dollar invoices, each tied to a signed-off deliverable, do not.
Milestone billing spreads that risk and, more importantly, attaches each payment to something the client has already confirmed they accepted.
Structure the milestones so they can carry a payment
For a milestone to support an invoice it needs to be independently verifiable. Discovery complete is not verifiable. Requirements document delivered and approved is. The distinction is the same one that makes acceptance criteria work, which is why scoping and billing are really one problem.
Front-load slightly. An initiation payment before work starts is standard in professional services and filters out clients who were never going to pay. Something in the range of a quarter to a third of the total is normal, and asking for it is not aggressive.
Keep the intervals short enough that no single unpaid milestone threatens you. Monthly, or every two to three weeks on a fast project, is a reasonable target. Long gaps between billing points reintroduce the concentration problem you were trying to remove.
- Initiation payment on signature, before work begins.
- Milestones sized so each represents a verifiable, approvable deliverable.
- Billing intervals of two to four weeks rather than quarterly.
- A final payment small enough that it is not worth disputing, but present enough to matter.
- Explicit terms for what happens if the client's review stalls the schedule.
The gap nobody owns
Ask most agencies where their cash is and they will talk about late-paying clients. Then look at the actual position and a surprising amount of it is sitting somewhere else: milestones that were completed and approved weeks ago and never invoiced.
This happens because the hand-off between delivery and finance is usually nobody's explicit job. The project manager knows the milestone was approved. The person who raises invoices does not, unless someone tells them. On a busy month, nobody tells them.
It is the most fixable cash flow problem in a services business, because the money is already earned and already agreed. All that is missing is a trigger. Approval should move a billable milestone into an invoicing queue automatically, and someone should own clearing that queue on a schedule.
Before chasing late payers, check how much you have delivered, had approved, and never invoiced. It is often the larger number.
Payment behaviour differs by market
If you work across Hong Kong, Singapore and the United States you are dealing with genuinely different payment cultures, and treating them identically produces avoidable surprises.
Hong Kong and Singapore corporate payers commonly run structured payment runs, often twice monthly. Missing a cut-off by a day can cost two weeks, so knowing the client's run dates is worth more than a polite reminder. Invoicing to hit the run is a real skill.
US clients, particularly larger ones, frequently require a purchase order before an invoice can be processed at all. An invoice without a valid PO number is not late, it is invisible. Establish this during onboarding, not when payment is overdue.
Across all three, larger organisations often impose their own terms regardless of what your engagement letter says. Price that reality rather than arguing with it: if a client pays on sixty days, either the rate reflects the financing you are providing or the milestone structure compensates for it.
Make the invoice hard to question
Most payment delays are not refusals. They are queries, and every query resets the clock. So the goal is an invoice that answers its own questions before anyone has to ask.
That means the invoice references the milestone by name, states the date it was approved and by whom, and where relevant carries the purchase order number and the agreed amount from the statement of work. A finance team receiving that has nothing to check and no reason to route it for clarification.
This is why approval records matter commercially rather than just administratively. An approval captured as a timestamped record against a named deliverable is the difference between an invoice that clears in one pass and one that sits in someone's queue awaiting confirmation from a colleague who is on leave.
Write down what happens when the client stalls
Milestone billing has one failure mode worth planning for: your payment depends on approval, and approval depends on the client doing something. If their review takes a month, you have financed a month of work for free.
Handle it in the engagement terms. A deemed approval clause — where a deliverable is treated as accepted if no response arrives within a stated window — is common and reasonable when the window is generous and clearly communicated. Pair it with a provision that dependency delays on the client side can shift the schedule and trigger a re-baselining.
None of this needs to be adversarial. Presented as how we keep the project on track, it usually reads as competence rather than suspicion, and it protects both sides from a stalled project nobody knows how to restart.
Related guides
Scope and commercials
Scoping client work so it does not drift
Why scope creep is a documentation failure rather than a client behaviour problem, and how to write acceptance criteria that survive a review meeting.
Read itTooling
Choosing a project management tool for client work
Why tools built for internal teams break down on client projects, the specific capabilities that matter when a client sits in the plan, and how to run an evaluation that predicts real adoption.
Read it