Time and Materials vs Fixed Price

Every services contract is a theory about the future, and every pricing structure is an answer to the same question: who pays when the theory is wrong? Time and materials and fixed price are usually presented as a menu. They are better understood as two different allocations of the same risk, attached to the same underlying object: the estimate.

Two structures, one estimate

Under time and materials, the client pays for the effort actually spent, at agreed rates, as it is spent. The estimate exists, but it stays visible and provisional; the invoice tracks reality, whatever reality turns out to be. Under fixed price, the supplier commits to deliver a defined scope for a single sum. The same estimate is still there, now with a contingency added for the risk of being wrong, folded into one number and presented as a promise.

With invented numbers, purely as illustration: a piece of work is estimated at a hundred days at a given day rate. Sold as T&M, the client is invoiced for the days that actually occur, whether that is ninety or a hundred and thirty. Sold fixed, the supplier quotes the hundred days plus a cushion as one certain-sounding figure, and the difference between that figure and reality lands on the supplier’s side of the table. The estimate did not change between the two offers. Only the address of the risk did.

Who carries which risk

T&M places the estimation and scope risk with the client: when the work grows, the bill grows. The supplier is not carrying nothing, though. It carries utilisation risk, rate pressure at every renewal, and the reputational slow burn of a running meter, because an inflating T&M bill erodes trust invoice by invoice even when every hour was honest.

Fixed price reverses the allocation. The supplier now carries the estimation risk, the delivery risk, and the interpretation risk of every ambiguous line in the statement of work. The client’s risks are quieter but real: it pays the contingency whether or not it is needed, and it buys rigidity, because under a fixed price every change becomes a commercial negotiation (this is where the change order earns its keep, and where scope creep becomes margin erosion on the supplier’s side). And one risk never moves at all: if the project fails, it fails the client, whichever structure was signed.

An estimate sold as a certainty

A fixed price does not remove the estimate; it sells the estimate as a certainty, which makes the estimate itself the product. Once that is seen, the real questions about any fixed price become questions about evidence. What sits under the estimate: measured delivery history on similar work, or a planning session’s optimism? Which assumptions were written into the scope, and which were left implicit for the two parties to discover they disagreed about? How will the contingency be governed: consumed silently until it is gone, or tracked as the risk budget it actually is? A firm that sells fixed prices is, whether it says so or not, in the business of manufacturing reliable estimates. Delivery merely tests them, and the results eventually surface in delivery margin.

One concrete example

Clearly illustrative, with no customer implied. The same programme is priced both ways in parallel universes, and in both it runs over. Under T&M, the overrun surfaces in the third weekly invoice. The conversation is uncomfortable and early: scope is trimmed, priorities are re-cut, the client is annoyed but informed. Under fixed price, the same overrun is silent. The supplier burns its contingency, then protects its margin the only ways it privately can: the senior architect quietly comes off the account, the hardest module slips to a later phase, testing is compressed. The problem finally surfaces months later as a quality dispute and a contested change order. The overrun was identical in both universes. The pricing structure decided when it became visible, and who had an incentive to raise it.

Hybrids, and the decision the choice should trigger

Real contracts increasingly sit between the poles, and every hybrid is a deliberate reallocation of the same risk. Capped T&M keeps effort transparent but hands the client a ceiling, a cap the supplier underwrites. Discovery then fix spends a short, paid T&M phase buying evidence before any promise is priced. Phase-by-phase fixed pricing re-prices at gates as evidence accumulates, so certainty is sold only as fast as it is earned. Fees linked to outcomes move part of the reward onto results. None of these removes the estimate. Each prices its uncertainty differently.

Which is the decision-intelligence point: the choice of structure is a decision about evidence, and it deserves to be made as one. An estimate resting on measured history from similar engagements can honestly be sold as a promise; an estimate resting on inference and hope should be sold as discovery, whatever the client would prefer to hear (the evidence hierarchy gives that distinction a vocabulary). And the estimate itself deserves a record: its assumptions, its evidence, the structure chosen and why, scored later against what delivery actually took. A firm that does this learns its own estimation error and prices its uncertainty knowingly. A firm that does not keeps discovering, one contested closeout at a time, that its real product was always the estimate, and nobody was in charge of making it.