Exploratory user research: how to do it and exploit it efficiently?

Question

Really need your advice and thoughts here.

I'm working on user / market research for a mobile self-help app which helps users self-diagnose their mental wellness and seek external help if they need.

The organization of the project is peculiar in a way, in that the app has already been developed and distributed without any "preliminary" user research. The PMs wanted a product really fast, so they they started out with assumptions of what users would need and developed from there. (Note however that we DID do several user tests as design & development advanced, however these tests were more focused on usability and less on product desirability or relevance to user needs.)

At this point PMs want to add a big new feature. So my team (who all have design background, not research) and I decided it was a good time to do a long-delayed user research on how people deal with their own mental health perception and seek for help.

We interviewed a dozen random people. We didn't show them the product because we weren't focusing on product usability or desirability. Rather we wanted to discuss their goals, strategies and obstacles on dealing with their mental health (Interviews aren't probably the best method owing to people's selective memory & bias but it was the cheapest and most accessible way).

Interviews were really interesting, I have tons of info. But my challenge is that a lot of this info is by nature very unstructured and it's quite challenging to put it into a structured form. It definitely has to be structured in order to make sense out of it (otherwise we'll have a hard time using it for development) but I'm resisting pressures from my team to structure everything in a persona + user journey form. "User journeys" the way the team expects are very clean, chronologically consistent. Whereas what I found out is pretty chaotic and with a lot of individual differences. I can make up a sort of formal system, just it's not in the formalism they expect. For me it's becoming "voodoo science" to try to force results into a special format, whereas I want to be honest in my approach (i.e user research is useful to have a contextual picture of people's complexities so we can service them better, but that doesn't mean that we can extrapolate and build user journeys that reflect every situation. Rather I want to describe the different "processes" in which people evolve, in a free-form I have yet to work out.)

For the record I've been working in UX research for the past 7 years, but mainly doing user tests for validating and improving apps. It's the first time I've done high-level exploratory research for product definition. I'm wondering at this point if this research was a waste of time, and if we should have stuck to designing something (anything) from our initial assumptions, test it and start from there.

Sorry if this post feels messy and vague, that's because probably I have to explain better what my team expects by "user journeys" :)

Thanks a lot for reading,

Answer 1

Your business has goals. Someone in your business has decided that connecting your users with mental health care providers is profitable, and users will download your app to do that vs. going to a website or using other online resources.

I think your first step is going to be to identify which customers should use your app, and which should not. You can hone in on people who:

• Are likely to use apps
• Are comfortable doing mental health screenings with your unknown app and might even prefer to do them that way
• Are likely to connect with a health care provider through your app, resulting in your business making money
• Are likely to find value in the TBD feature being developed

This is the benefit of personas - separating valuable customers from everyone else. It seems like your app was developed with a very broad audience in mind, but it's not for everyone. A 10-year-old in Sweden might need mental health care, but can't book an appointment through your app, as an example. If demographic personas don't seem useful, you can focus on personas that identify motivations and behaviors.

Through your research, you might end up with a large amount of specific, individualized personas, and that's okay. I think you're going to need a large pool of respondents here to find patterns, so a survey might be your best bet. Your prior round of research can help you develop good survey questions.

Use your business goals to screen out people who would not be desired customers and find patterns with the respondents who would derive value from your app. Build your personas based on those patterns. Then, prioritize those personas to focus on the ones who would benefit the most from the Big New Feature. Continue doing research by talking to those people. You might not need journey maps, but you can do other exploratory discovery exercises with them.

Answer 2

It definitely has to be structured in order to make sense out of it (otherwise we'll have a hard time using it for development) but I'm resisting pressures from my team to structure everything in a persona + user journey form. "User journeys" the way the team expects are very clean, chronologically consistent. Whereas what I found out is pretty chaotic and with a lot of individual differences.

Its hard to give a detailed answer without actually looking at your dataset, but have you tried taking a look at Card Sorting as a method to derive insights? I have found that its a much more suitable tool to use when you have unstructured qualitative data, such as those you describe. Basically, any key point or insight from an interview can be summarized and turned into a card. You can usually get 1-5 cards from each interview, and then from there you try to see the patterns that emerge in terms of similarities, common grievances, categories etc.