Funding Agile Work When Finance Thinks in Waterfalls
There is a conversation every technical PM in an agile organization eventually has with finance, and it goes like this. Finance: "What exactly will we get for this budget?" You: "We'll know more after the first two increments." Finance: visible distress. Neither side is wrong. Finance is accountable for capital allocated against outcomes; agile is honest that outcomes get discovered, not predicted. Funding models are how mature organizations resolve this standoff, and getting the model wrong quietly reintroduces every waterfall pathology that agile was hired to remove.
How predictive funding works, and what it silently buys
Classic project funding approves a full budget against a full scope upfront. Its virtue is legibility: one decision, one number, accountability attached. Its hidden cost is that it pays for certainty theater. To get funded, someone must specify scope in detail before the learning that would make the specification accurate, so estimates inflate into promises, and change becomes an exceptions process. The budget is protected; the honesty about what's actually known is the thing that gets spent.
Incremental funding: pay per learning cycle
The agile-native alternative funds in slices: enough for the next increment or quarter, released against demonstrated value, with the option to redirect or stop at each checkpoint. This is venture logic applied internally, and its real product is not cost control. It is the option to stop. Predictive funding makes stopping a failure, money was committed against the full vision. Incremental funding makes stopping a routine portfolio decision: this stream demonstrated less value than that one, capital moves. Teams under incremental funding also behave differently: when the next tranche depends on demonstrated value, increments get ruthlessly outcome-shaped, which is precisely the pressure a healthy product team wants.
The transitional pattern I have found workable in organizations mid-journey: stable funding for the team's existence, capacity funding, with incremental decisions about what that capacity aims at. Finance gets predictable spend; the work keeps its adaptability. It also sidesteps the worst artifact of project-based funding, teams dissolved and re-formed per project, which Tuckman explains the cost of better than any finance model does.
Measure value, or incremental funding becomes theater too
Incremental funding only disciplines anything if the checkpoints measure value, and value means outcomes: orders processed, hours eliminated, revenue enabled, risk retired, the same year-after numbers I argued belong in any definition of success. If checkpoints only inspect output, features shipped, points burned, the organization has rebuilt waterfall reporting with extra meetings.
Which brings up the mistake worth its own warning label: story points are not business value, and must never appear in a funding conversation. Points measure effort and complexity, a team-internal planning currency. The moment finance sees points, points become a productivity metric; the moment points are a metric, estimates inflate to match, and the team's planning instrument is destroyed for a reporting purpose it was never fit for. A high-effort epic can deliver marginal value; a two-point configuration change can unlock a market. Effort in, value out, different units, different conversations, and the PM's job includes keeping the currency exchange closed.
The PM as translator, again
Practically, funding conversations are another boundary-layer job: translating between finance's legitimate need for accountability and the team's legitimate need for adaptability. The translation artifacts are simple: a value hypothesis per increment, in outcome units finance recognizes; honest ranges instead of false point estimates; and checkpoint reviews that show measured outcomes against the hypothesis. Organizations that learn this rhythm stop asking "what exactly will we get", and start asking the better question: "what did the last slice teach us, and is the next slice still the best use of the money?" When finance asks that unprompted, the transition is real.
Nguyễn Hải Nam
Project Management Lead. 16+ years from code to delivery. PMP®. Writing here about project management and engineering.