When I joined 8fit in 2016, I felt the need to empathise with our customers. There were a couple of challenges that kept the company detached from its users.
First, most of the users were based in the US. Second, since we worked with nutritionists and coaches and were privileged to live in the city obsessed with healthy living, most of us hardly knew what it was like to be unfit. It was vital to help the team relate to and experience this user journey on a very personal level.
Essentially, we wanted to combine our quantitive analysis with our quantitative research. My personal ultimate challenge was to make it data-informed. We ran a survey that aimed to collect quantitive data about users’ lifestyles, and, at the same time, set up the interviews that were meant to educate us about their lifestyles in more depth. In order to make both of them comparable, we made sure that the questions in the survey were aligned with the interview questions.
The process
I’d like to outline a process and share a couple learnings.
- First, select your interviewees. You probably want these users to be representative of your target group. Check the demographics and other metrics relevant to your research.
- At the same time, it’s a good idea to check with your team how they plan to make use of personas. Maybe they want to know something specific about the people they create this product for. Personas don’t play a decorative role only, they can and should be applicable in product development.
- Create an interview questionnaire. Oftentimes, what people say and what people do are two different things. Try to validate what they say through what they do. For example, we wanted to find out if people eat healthily, and we validated their answers by asking them to describe what they ate in the past few days.
- When it comes to technicalities, it’s probably a good idea to run one or two interviews internally — it will help to time and to conduct it properly. I was surprised by how people interpret the questions. We didn’t get it right from the first try. But every single interview improved the next round of question asking.
- Schedule the interviews. We used a combo of ScheduleOnce / Appear.in, but we noticed, that an interviewee is more likely to drop out when they need to perform an action (download an app, confirm time). Make sure that you give them a variety of options, so that they can pick whichever works best for them. Also, when people drop out, make sure to recheck the demographics ratio. You want to keep your data aligned with your target group.
- For the interviews we created a handy spreadsheet that allowed us to take notes directly during the interview. Normally we had two interviewers — one engaging in conversation, another taking notes. That saved us from a lot of manual work of listening and summarising the interviews.
- The hardest part is analysing and grouping the data. It’s easy to be swayed by biases, and you’d normally want all of your team members to counterbalance the process. We used affinity mapping to identify behavioural patterns. When we grouped the interviewees, we were able to lay out their qualities on behavioural axes. That helped us pinpoint their similarities and differences, and essentially understand their personalities. For each of the personas we created a segment fit to make it easier to work with the groups further on.
We then aligned our survey data with interview analysis and saw that some of the patterns match and others we have to iterate on.
Fundamentally, through this research not only we were able to understand people’s motivation, struggles and emphasise with them, but also make use of personas in the product development cycle later on. Having gained this valuable human connection with our users helped us build and ship features that, hopefully, change their lives for better.
Join the product and design professionals reading Noteworthy to understand the ideas and people that shape the products we use every day (and get an early peak at our product!).
If you have thoughts to share, we accept post submissions!
