From Traditional Research to Mobile Ethnography

When we recruit traditional qualitative researchers for our consulting team, they are well-versed in interviews, ethnos, and discussion guides. Their expertise is usually rooted in following participants down wandering cognitive paths. But dscout is recruiting them to be mobile researchers.

And that’s when the fun begins.

Once we’ve locked the doors behind them, we reveal the capacity — and the constraints — of dscout’s in-the-moment methodologies. We interviewed dscout research lead and gravitytank alum Carey Palmer for her take on what makes mobile a different beast.

When did you realize that mobile research would be so different from facilitating focus groups and making field notes?

When I came to dscout, I was used to interviews and ethnographies. When you’ve internalized your objectives and have a discussion guide, you don’t have to think through every single individual question before you begin. To stimulate new avenues for conversation, you’d rely on the participant and environment to lead you.

In a shop-along, we’d talk about the assortment the participant sees, what stands out, and their process for making choices. The challenge with mobile research is that we are not physically there with the participant, so our question prompts can’t come from what we are seeing. We have to anticipate and share our questions with them before they begin shopping. That can be a challenge.

How do you prepare for that?

The strength of mobile research is that you are getting feedback from people as part of a natural process. But that means I have to think more carefully about planning the research, because if I don’t have the questions phrased correctly, I might lose nuances or insights in the scouts’ answers. We work upfront with the client to ask the right questions.

To feel confident that I have a comprehensive research plan, I conduct mini-interviews with friends and colleagues and use those conversations to fuel the mission design. This is a great way to refine your approach and question list. And it’s always a good practice to pilot your mission design before going live with participants.

Have you ever realized that you didn’t ask a question exactly the right way?

Sure. We have our set list of questions, but we also have a messaging system that we rely on heavily. If someone doesn’t naturally expand on a topic, we can ask them why they said something. When we ask them, they are actually ecstatic to hear that we are so interested in what they have to say. Then they really want to have a conversation!

Also, I often ask scouts to complete a secondary reflection mission, where I ask new questions that emerged as a result of seeing data from the first mission. I love doing this because it gives me a day or two of analysis time and allows me to generate some really thoughtful questions to deepen our understanding.

Do you have any advice for other researchers trying to write mobile research questions?

I do. When it’s important to have an open-ended question, I think carefully about the wording of the question, and I’ll specify the length of the response — say, two or three sentences, for example. That helps to ensure the participants know how much detail I want.

Tone is also important. Make the question phrasing feel really affable and personable. That eases people into the process. The “white lab coat” approach is not what people respond to in social research. It’s not natural to them.

Any other ways that your mobile research approach differs from a traditional one?

Mobile research is an unobtrusive way to be with users, and we are getting feedback from people wherever they are. We aren’t following them constantly as we might be in an ethnography — pay no attention to the man behind the camera!

And in traditional research, you’d be trying to capture what you need over two hours, and it might not produce anything. But with in-the-moment research, people are choosing what and when to show you things, so we can actually combine observation and interviews better. When participants are showing you specifically the moments that you are interested in, and answering your questions about why those moments matter to them… that’s really interesting!

Can you give an example where in-the-moment research was a good replacement for a focus group or IDI?

I remember research analyst Francesca de la Fuente had one project where people filmed their reactions after watching TV. She got the coolest snippets of people being so emotional about their Walking Dead episode.

Most people watch TV alone, or with their family — not with strangers. So if you had been there with your clipboard and recording their reactions, their reactions would have changed. Or if they’d been in the focus group room, it would have changed it. Even relying on them to record a diary entry wouldn’t have been the same as getting that in-the-moment video.

Being in their own space, without anyone directly observing, it was natural. And, it’s what made the project.

Any projects where mobile research was a good replacement for traditional ethnos?

We once mailed out prototypes of a movable television to people’s homes, and they had to act out when they would use it. We did that because we know that living with an object for a while inspires ideation — “Oh, I am in the kitchen! I might use it while I’m doing this.”

That scenario would be harder to achieve in a regular interview or even ethnography. When you’re at someone’s house, they won’t really be engaged in their regular routine.

How about analysis? Is that any more complicated than in a typical qualitative study?

The depth of analysis the dscout dashboard enables is really valuable to me. Without any effort on my part, I can see counts and percentages on multiple choice question responses. I can use filters to view the data in many different ways. I can say, “show me moments that were experience high points, and see how that subset of data compares to everything else.” I can add notes to snippets as I’m making observations. I can binge-watch videos. And I can export everything into Excel for further analysis.