Before the Contract: What a Paid Discovery Phase Should Deliver for a UK SME

For a UK SME, commissioning custom software is rarely a casual purchase. It involves budget, internal reputation and a business process that may already be strained. A paid discovery phase is meant to make that commitment safer. Yet discovery is also one of the easiest stages to disguise: a series of workshops can look rigorous while producing little that a buyer can use.

The distinction matters because discovery sits before the main contract. If it is genuine, it should reduce uncertainty and give a board or management team a defensible decision. If it is a sales exercise, it will generate warm feelings and a proposal rather than evidence.

Discovery should produce decisions, not atmosphere

A useful discovery phase ends with a recommendation that could be no. That is not a failure. A well-run discovery may conclude that the current process should be simplified, that an off-the-shelf product is adequate, or that the organisation is not ready to build. A supplier that cannot imagine reaching that conclusion is selling, not advising.

The phase should also be bounded. The buyer should know the time commitment, the people required, the questions in scope and the artefacts that will be delivered. A fixed fee is helpful, but clarity matters more than price. A low-cost discovery that produces a sales document is expensive in the long run.

The deliverables a UK SME should expect

Genuine discovery outputs vary with the problem, but most credible engagements should provide the following.

Validated user needs and priority map

This is more than a list of requirements. It should show which users were spoken to, what tasks they need to complete, where current processes fail and which needs are commercially important. For an SME, that may mean speaking to operational staff, customers and finance. The output should distinguish evidence from assumption.

Technical risk notes

The supplier should identify the parts of the build that are uncertain: integrations with existing systems, data quality, security boundaries, third-party APIs, hosting constraints or performance requirements. Each risk should have an indication of likelihood, impact and a proposed way to test or reduce it. A discovery that contains no technical risk is usually not looking hard enough.

A costed delivery plan

The plan should set out phases, dependencies, assumptions and indicative costs. It need not be a fixed quote for every detail, but it should show how the supplier arrived at the number. Assumptions are particularly important: if the price depends on access to a legacy database or a client product owner, that should be explicit.

An operating cost estimate

Software has a second budget after launch. Discovery should provide at least a first view of hosting, support, licensing and maintenance. This is where many business cases become unrealistic, so a supplier that avoids the topic is leaving the buyer with a blind spot.

A clear go/no-go recommendation

The final report should state whether the project should proceed, what conditions apply and what alternatives were considered. It should also record open questions. A recommendation that everything is ready, with no conditions, is rarely credible.

Red flags in a discovery proposal

Buyers should be cautious when discovery is priced as a credit against a future build. That can create pressure to proceed regardless of findings. The same applies when the output is described only as a requirements document or a workshop series, with no named deliverables.

Other warning signs include no access to the people who will do the work, no senior technical input, and no mechanism for the buyer to speak privately with users. A discovery phase that is run entirely by sales and design, without engineering, may produce an attractive concept that is expensive or difficult to build.

A supplier should also be willing to share the raw notes and research materials. The buyer is paying for evidence, not just a polished presentation.

Questions to ask before signing

Before committing, a UK SME should ask who will conduct the discovery, how many users will be consulted, what happens if the recommendation is not to build, and whether the supplier will provide a costed plan with assumptions. It is also reasonable to ask how conflicts of interest are managed when the same firm may bid for the build.

Finally, ask what the buyer will own. The research, decision log and risk register should be usable even if the build goes elsewhere. That is a practical test of whether discovery is a genuine product or a sales device.

A small phase with a large effect

Discovery is not a guarantee of project success. It is a way to make the next decision with better information. For a UK SME, the value lies in reducing the chance of an expensive build that solves the wrong problem. A paid phase that produces evidence, risks, costs and a real recommendation is worth the money. One that produces only confidence is not.