Research meets the conditions it has to work in.

We begin by examining the challenge itself: who experiences it, what is already being done, what has failed before, what people know from experience, and what constraints any response will have to live with.

Those constraints may involve infrastructure, cost, skills, incentives, trust, regulation, accessibility, maintenance, staff turnover, institutional capacity or something entirely specific to the setting.

The important point is not to arrive with a preferred answer and then look for somewhere to put it. The conditions help determine what is worth testing.

The people who would use, maintain, govern or live with an idea are part of understanding whether it makes sense at all. Their involvement is not a final consultation exercise.

What they know can expose assumptions, alter the question, change the design or show that the proposed response does not fit the reality it was meant for.

A useful test therefore asks more than whether people like an idea. It asks what they actually do with it, what prevents them from using it, what they change themselves and what would have to remain true for it to work over time.

Prototypes, pilots and small-scale experiments allow us to learn before a large commitment makes change difficult or expensive.

The size of the test should match the question being asked. A small pilot cannot prove that something will work everywhere, and completing one does not create an entitlement to scale.

Scaling is a separate decision. It requires evidence that the approach still makes sense under different conditions, with more people, greater cost, more responsibility and less direct control from the original team.

Something does not become more useful simply because it can be digitised, automated or made intelligent.

We use technology where it improves what people can actually do or makes something possible that would otherwise be difficult to achieve.

That judgement includes more than technical performance. Access, cost, maintenance, privacy, safety, accessibility, accountability and human control all affect whether a technology deserves to be used.

Sometimes the right decision is to build less. Sometimes it is not to build at all.

A Lab project does not need to end in a product to have succeeded. The result might be a stronger method, a different intervention, a change in practice, a service, a workflow, a product, or evidence that the original idea should not continue.

Stopping can be a valuable result when it happens early enough. It can prevent weak assumptions from becoming expensive programmes, unwanted products or practices that persist simply because too much has already been invested in them.

Not every experiment becomes a product. It should not.

NoviThea's commissioned work responds to defined needs from partners, clients and collaborators.

The Lab gives us room to pursue questions that emerge from that work but need more investigation, as well as questions that matter before anyone has written a brief or committed funding to them.

Research in the field can expose something worth testing further. Lab work can, in turn, change the methods, assumptions and questions we take into future projects. The two inform one another without being the same thing.

Some work eventually reaches a point where the useful result needs to remain in people's hands after the original research or experiment has ended. That is when a product may make sense.

But a product is not the destination of the Lab. The question is whether something useful has been learned, whether the evidence supports continuing, and what form the next step should take.

Sometimes that form is a product. Sometimes it is a method, a different way of working, a decision or simply the knowledge that an idea should go no further.