· Bridgera
Start with the decision you can make today
A fixed-scope project fits a problem with a useful boundary: a website release, an identity system or one operational workflow. An ongoing programme fits a need that changes as evidence arrives, such as a prioritised series of improvements. Neither arrangement removes the need to decide what work is authorised. Before choosing a label, ask what you can define now and what you expect to learn through delivery.
Write the desired output and the conditions under which it is useful. A website is too broad a description on its own. Name the pages, publishing needs, integrations and journeys that belong in the release. If important unknowns make those decisions impossible, scope an investigation first. Buying a smaller discovery deliverable can produce a more defensible production brief than asking a supplier to hide uncertainty inside a confident quote.
Make a fixed scope specific enough to review
A scope should name deliverables, inputs, review points and acceptance criteria. It should also state assumptions that affect delivery: content availability, access to existing systems and the behaviour supported by an external platform. These details are not administrative decoration. They make it possible to tell whether a requested change is part of the original work and whether a finished deliverable answers the agreed brief.
Use examples to make boundaries concrete. If a product catalogue is included, identify the content structure and who supplies the records. If an integration is included, specify the data and direction of travel. If a motion sequence is included, agree its destinations and formats. A fixed quote becomes useful when both sides understand what it buys. It becomes brittle when one side sees a finished business outcome and the other sees only a narrow production task.
Define the discovery output, not just the activity
Discovery should end with decisions or material that someone can inspect. That might be a workflow map, a tested prototype, a content plan or a comparison of technical approaches. State what uncertainty the work will reduce and what it cannot establish. A prototype can show whether a connection is feasible; it may not demonstrate that a production system is ready for every failure condition.
Agree the next decision before discovery starts. The buyer might choose a production scope, revise the approach or stop the idea. All are legitimate outcomes if the investigation was designed to answer a real question. Keep the production proposal separate enough to reflect what was learned. Do not treat a discovery payment as automatic permission for a wider build, and do not describe exploratory work as a released system simply because it produced a convincing demonstration.
An ongoing programme still needs boundaries
A retainer can become vague when the buyer expects every new request to fit and the supplier expects a limited capacity that was never described. Define how priorities are chosen, what kinds of work are covered and how new commitments are approved. The programme should have a visible queue and a review rhythm. That gives both sides a way to explain why one task moves forward while another waits.
For creative and technical work, scope each approved batch around a concrete result. An ongoing relationship does not need an open-ended production promise. Bridgera’s working model agrees fixed scope in writing before work starts, so a continuing programme can be organised as successive agreed pieces of work. Support obligations and response expectations require their own terms. A delivery queue is different from a service commitment to respond to every operational issue.
Changes need a decision, not an argument
New information is normal. A stakeholder may identify a missing user role or an external API may behave differently from the assumption. When that happens, describe the change and the dependency it creates. Then compare practical choices: add it to the agreement, exchange another priority where feasible or plan a later release. A short written decision prevents a reasonable new idea from becoming a hidden obligation.
Consolidated feedback helps keep this process usable. One person should resolve contradictory comments and distinguish a factual correction from a new direction. Reviews also need a clear question. Approving an interface concept is different from accepting an implemented workflow. If the team reopens an earlier decision, identify the effect on later work rather than pretend it is a minor edit. Scope control should make choices visible without discouraging useful feedback.
Buy a handover as well as an output
Ownership should be described before delivery. Name the code, design files, rendered assets and accounts that belong to the client. Identify third-party licences and operating dependencies separately. A finished page or application is not the whole handover if the buyer cannot edit content, deploy a change or find the account that controls a critical service. Documentation and an operating walkthrough should match the team that will take responsibility.
A good engagement document lets you answer simple questions: what will exist, how will we inspect it, who supplies the inputs and what happens next? It also makes the limits clear without hiding them in general language. Choose a defined project when the output is clear, discovery when a decision is missing and a bounded programme when priorities need to evolve. The commercial arrangement should follow those needs, with the authorised work recorded before production begins.
Before signing, ask someone outside the project discussion to read the scope and describe what they think will be delivered. If their account differs from yours, rewrite the ambiguous part. Pay particular attention to words such as launch, integration and support, which can describe very different obligations. A short concrete example often resolves the confusion. The aim is not to anticipate every future request; it is to give both sides a reliable basis for making the next decision when circumstances change.