Tips for Running a Lean Research Process

You’ve got a great idea for something new. Before you go spend all your cash on building that thing, you need confidence that what you do is going to matter to users (and importantly: buyers). 

Interview Users to Learn What Matters

No matter what you already know, and you know a lot, you want clarity on which things matter. Start with User Interviews. Those lead to Themes, and those lead to Features.

• Schedule at least 12 interviews. Or as many as  you can. Too few and you’ll have trouble really seeing patterns. From a data perspective there’s never really “too many”, but you only have so much time. More than ten interviews and you should start to see strong(ish) patterns emerge. More is better, so aim for twelve plus a handful more. Some folks will bail. Err on the high side.
• Write a script in advance. You’re good at talking to people. You can wing it. Your questions will be similar for consecutive calls, but like the telephone game, they’ll slowly change, and your last interview will end up vastly different from the first. 
• Wait to synthesize. Don’t let recency bias take over. Leave some time between your last interview and your analysis. Review all the notes again before you identify themes.

Move Faster by Prioritizing User Stories

Once you have themes you can collect them into vague ideas of features. You’ll have immediate thoughts on where to go here, and that’s a great starting point. But do not design them yet. All this functionality is too much to build. Your users should tell you what matters and what doesn’t.

• Pick your top one or two features, and for each one, bullet-list possible user stories in the classic form “I want to ___ so that ___”. 
• Put together a survey with your user stories listed (organized by feature). Don’t ask “which user story matters”. Everything matters to some degree. Instead, ask people to rank the stories in order of priority. 
• Get this survey out to as many of your users as possible. Include people you interviewed and those you didn’t, if you can. 
• Synthesize the results, paying close attention to every person’s top three or four functions. Ignore the bottom of the lists, because those bits of functionality don’t really matter to people. 

If you get this far, you’re in good shape. You should now have a couple features with their most important functionality prioritized. 

Going Further with Your User Research

Can you go further with this? Absolutely: you can go deeper on personas and segmentation, user flows, wireframing, prototyping, conducting user validation… If you have the resources, these are incredibly valuable processes, the more of which you do, the stronger your likelihood of success. 

But if you’re running lean, this process will get you to a spec you can work with—fast. Depending on your schedule, you could get there within as little as a couple weeks.

Good luck!

James Barr founded Strata Research, and has led the creation of successful products many times over. He’s passionate about the interaction between design, product management, and development in creating lovable software.