A contract is a mechanism for allocating risk between a buyer and a supplier. It becomes difficult on adaptive work because the traditional instrument for doing that is a fixed scope, and the whole premise of adaptive delivery is that the scope will change.
The way out is to fix something else. Price can be fixed against a period of capacity, against a defined increment, or against a ceiling with an incentive to come in under it. What no contract can fix simultaneously is scope, cost and date, and an agreement that appears to has simply moved the flexibility somewhere less visible, usually quality.
Why the traditional shape fights the delivery model
A firm fixed price against a full scope statement asks the supplier to price work nobody understands yet. The rational supplier prices the uncertainty into the number, so the buyer pays a premium for a specification that will be wrong.
Worse, the agreement makes learning adversarial. Every discovery becomes a change request, the supplier is incentivised to argue that a request is outside the original scope, and the buyer is incentivised to argue that it was implied. Both sides then spend effort on the boundary rather than on the outcome.
None of that means fixed price is wrong. Where the scope genuinely is knowable and stable, it transfers risk efficiently and is the right instrument. The mismatch appears when it is applied to work that will be understood by building it.
Structures that fit adaptive delivery
| Structure | What is fixed | What flexes | Where it fits |
|---|---|---|---|
| Capped time and materials | An upper limit on spend | Scope, and the actual spend below the cap | The buyer wants a ceiling and the supplier will not price unknowns |
| Fixed capacity | The team, the rate and the period | Everything the team works on | Ongoing product work where the buyer sets priorities continuously |
| Incremental delivery | Price per increment, agreed as each is defined | Whether later increments are commissioned at all | The buyer wants the right to stop without penalty |
| Target cost with shared outcome | A target, with overrun and underrun shared by formula | The final cost | Both parties can influence the result and neither can control it alone |
| Graduated fixed price | A base price, with the rate adjusted by delivery against agreed dates | The supplier's margin | The buyer wants schedule incentive without micromanaging scope |
Two clauses are worth knowing by name because they solve the change problem directly. A change for free arrangement lets the buyer swap an item of equivalent size for one already in scope at no cost, which removes the incentive to argue about boundaries. A money for nothing arrangement lets the buyer terminate early and pay a fraction of the remaining value, which means the supplier is not punished for finishing sooner and the buyer is not forced to consume work it no longer wants.
The two tier structure
The common arrangement separates the relationship from the work. A master agreement holds the terms that do not change, covering rates, intellectual property, liability, confidentiality, dispute resolution and how work is authorised. Individual statements of work then commission specific increments under it.
This works because the negotiation that takes months happens once. Commissioning the next increment is then a short document referencing an agreement already in place, which is what makes it possible to fund work in pieces without a legal cycle between each one.
Writing scope when the scope will change
The instinct is to describe the deliverables. On adaptive work the more durable description is of the outcome and the constraints.
State what the work is for and how success will be judged. State the boundaries, including what is out of scope, since that ages better than a list of what is in. State the quality requirements and the definition of completeness, because those genuinely should not move. Then describe the way of working, including the cadence, who attends reviews, how priorities are set and how either side raises a problem.
An initial backlog belongs in the agreement as an indication of scale rather than as a promise of contents. Naming it as such in the document itself prevents the argument later, and the argument is otherwise certain.
What the buyer gives up and what they gain
A buyer moving away from fixed price and fixed scope loses the ability to point at a document and demand it. That is a real loss and it should be acknowledged rather than argued away.
What replaces it is a set of controls that operate continuously. Working software or an equivalent deliverable at the end of every increment, the right to reprioritise what comes next, and the right to stop. Those are stronger than a specification for anything genuinely uncertain, because they act while the money is still being spent rather than at the end when it has been.
The buyer's obligation grows in return. Adaptive contracts require somebody on the buyer's side who is available, informed and able to decide. Where that person does not exist, the arrangement produces a supplier waiting for answers, and the buyer will conclude that adaptive delivery does not work.
Public sector and regulated procurement
Rules that require competitive tendering against a specification constrain what is possible, and they are not negotiable by the project.
What is usually available is competing on capability and rate rather than on a priced scope, awarding a framework with call offs against it, or structuring the first increment as a small competitively awarded piece that establishes the supplier before the larger commitment. These are procurement design decisions that have to be made before the tender is issued, which means the delivery approach has to be settled earlier on this kind of work than on any other.