Revolving Door Method: Rapid User Research at Crunch-Time

I went to the local UPA meeting last night. It was a good presentation by the user research team at Revolution Health (a startup founded by Steve Case, former founder of AOL), Leah Rader and Beth Toland. They were a lot of fun, and had a cool “rapid user research” method they used that they call the “Revolving Door Method.”

Key ideas:

  • This is basically a compression of their standard research model to serve the need of providing just-in-time research results in a matter of days (rather than the normal weeks) as a crunch-time like a site launch approaches.
  • They lined up a recruiter to get them 5 people a day, 3 days a week, for 5 weeks, without specifying exactly what the study would be about, but ensuring that they recruited from target populations for their overall product line.
  • They internally advertised to all projects and product managers that they could get user research results as a part of this “blitz.” They had a one-page form for people to fill out to get into the research queue.
  • When the 5 weeks started, they would run each recruit through 5-7 different subject areas, for a total of only 60 minutes for each recruit.
  • They set up topic stations and moved the recruit between them, so all the materials for each topic were at hand and they didn’t have to shuffle them.
  • At idle times for the recruits (such as if they arrived early), they had a survey for them to fill out.
  • They would run the recruits through against the prioritized list of topics, juggling the topics so each topic got data from about 7-9 recruits.
  • They ran the recruits through the topics Monday through Wednesday and reported out to clients on Thursday.
  • The internal client was required to sit in and take notes with the first recruit for their topic. They could choose to stay for others, but this really helped their engagement.
  • They did a quick analysis of the data each week for each topic that they called “Brady Bunch Style.” Basically, the results from each recruit for a specific topic were laid out in a 3×3 grid on a table or sheet and they would look for patterns, themes, and outliers. This is what they would report out on.
  • Every week, they held a 2 hour meeting on Thursdays where everyone in the company was invited to hear the weeks results. This became very popular and well attended.
  • They also posted advertisements for upcoming research and recent results in public areas around the company.
  • After five weeks of this intensive work, everyone was exhausted. This is not a sustainable model, but it is great for crunch time. At other times, they follow their normal model of providing results within 4 weeks (rather than a few days in this one).
  • This model doesn’t provide statistically significant results, but on all the follow up, in depth studies they’ve done since, they’ve never had a finding contradicted – just elaborated and detailed. As the one researcher said, “I haven’t had to hang my head in shame yet.”
  • There was lots of discussion relating this “rapid” method to agile software techniques.
  • This entry was posted in Interaction Design. Bookmark the permalink.

    Leave a Reply

    Your email address will not be published. Required fields are marked *