Product discovery for confident software decisions

Decide what is worth building

A request for a new platform often starts with a symptom: staff re-entering data, customers abandoning a process, or an ageing application becoming difficult to change. Commissioning software before understanding that problem can commit your organisation to the wrong solution.

Discovery gives decision-makers a shared view of the need, the constraints and the available options. It can support a new digital product, a replacement service or a targeted improvement to something already in use. Businesses, non-profits and government agencies may have different approval processes; all need evidence they can explain to the people making the investment.

Shape the work around the uncertainty

We begin by identifying the decisions the discovery must support. The scope might combine interviews with users and staff, analysis of an existing service, a review of integrations and data, or prototypes to explore a difficult interaction.

Our researchers, designers and engineers work together so that desirable ideas are considered alongside delivery and operational realities. When AI is an option, we examine the task, available data and need for human oversight before recommending it.

The useful output is a reasoned choice. We document findings, unresolved assumptions, dependencies and priorities, with recommendations for the next stage. That may be a small experiment, a phased build, changes to the existing system or further investigation before a commitment.

Research into a proposed public service

For the Department for Business & Trade, we conducted 25 in-depth interviews during a one-month discovery into payment practices. The findings identified payment-process issues relevant to the proposed digital service and policy work. The engagement shows why understanding how people actually work can change the questions a project needs to answer.

Choose a focused starting point

If your main uncertainty is what users need, start with user research and testing. To test a proposed interaction, consider digital prototyping. If the proposition is understood and you need a first usable release, MVP development may be the next step.

Do we need discovery if we already have requirements?

Existing requirements are useful input. We review which assumptions have evidence behind them and where uncertainty could materially affect scope or cost. The work can concentrate on those gaps rather than repeat decisions you have already made.