So there is Lean Product Development — when we are making small steps towards solving a problem and checking after each step, whether we actually got closer.
This process helps minimize risks, including:
- problem risk: the problem (aka a customer or business outcome) may not be something worth of solving, or it is not the most important problem we could tackle.
- solution risk: our solution doesn’t really solve the problem or may even introduce more pain than it relieves.
But there is another angle for combining “Lean” and “UX Research”. We can think of a research project as a product/service. Thus we can apply lean principles to the research itself. What do I mean?
Research, being a service, can run into the same obstacles as any service.
- problem risk: we are not solving the right puzzle. E.g. we realize that the question was not the most important one; it turns out that there was a misunderstanding between the researcher and the PM about the question; the user reveals something that fundamentally changes how we think about the problem space and what questions should we be asking, etc.
- solution risk: no answers are coming from the research. E.g. the chosen methods are not working out for some reason; survey response rates are unexpectedly low; interview sample turns out to be too heterogeneous to discover any patterns; behavioral data is contaminated, etc.
For these challenges, the answer is the same as in lean product development. Make small steps toward solving the problem (I mean answering the research question) and adjust course accordingly after each step.
In my practice, this looks like the following:
- when I start a new research project, I create a prototype (or MVP if you like): quick and dirty research. 2 days tops. Just grab some of the closest available pieces of information. E.g. a couple of interviews, a good look at earlier research results or just running a simple query. Then I try to answer the research question based on that data.
- Show the results to the PM and the designer. See if we think the research questions are still spot on. See if the answers are relevant.
- If not: change the questions or adjust the research method.
- If yes: do another iteration to make the answer better: go deeper, make it more representative, more accurate. Add other methods in the mix.
Because no research plan survives the first contact with reality. (Just as in Steve Blank’s “No business plan survives the first contact with customers”).
That’s what I call Double Lean UX Research.