Authors:
Jon Kolko
Twenty-five years ago, I learned a right way to conduct generative, qualitative research: Go to places where people work and live, establish rapport, become an apprentice, and ask open-ended questions.
Recently, I've found a passion for a different approach: a selfie study. This is an entirely remote-based method that combines cultural probe-style interventions with participant-created content. Participants receive daily text messages and respond by recording themselves doing things. The method runs contrary to a lot of what I've held true about research, primarily that being physically present is an axiomatic part of being empathetic during and after research.
Here are some examples of how we've used selfie studies.
Employing Selfie Studies
We recently worked with a very old and conservative life insurance company. The company was hoping to better understand how families made decisions around purchasing complex products like whole life insurance. The company has older, traditional executives, who were convinced that families get together around a dinner table and have serious conversations around life, death, and wealth. If such meetings actually occurred, it would be impossible to attend them, so we recommended and successfully deployed a selfie study.
It turns out that purchasing life insurance is not the result of serious decision making. Rather, it is treated extraordinarily casually, driven in many ways by the simplicity that newcomers in the space, like Lemonade, have introduced to the purchasing process, and by the changing norms of society at large. One of our participants researched and made the decision to purchase life insurance on the treadmill at Gold's Gym. Watching a video that captured this process completely blew the minds of our stakeholders, paving the way for a fundamental shift toward new, casual, digital-only purchasing.
We also recently deployed a selfie study with parents of young children to understand what the idea of early childhood education means to them. In the past, our study might have entailed a three-hour visit with parents, a seat at their kitchen table, and a guided interview. In the selfie study, we sent them two prompts every day via text and asked them to record themselves completing a task or answering a question. An example task was asking parents to show us their child's most educational toy and explain why it was valuable.
The responses were candid and provided insight not only into their toy selection but also into a wide variety of perspectives on education. We learned about which sources they trust to teach their children ("This is the television, where we watch Miss Rachel to whom we owe our lives"); how their socioeconomic circumstances affect their children's learning at home ("This is a Speak and Spell that we were given at the church; it's the most expensive toy we have"); and their values about technology ("This is our backyard. It's the most educational toy we have. I want my kids as far away from electronics as possible.").
The selfie study is effective because it provides a raw, intimate view into people's private lives.
We learn a lot from what our participants say. We learn more from the context of each video. Affluent product managers often have beautiful, clean, structured homes with partners and at-home caregivers to help provide their children with a strong foundation for learning. They have expensive toys and devices and read countless books on the most effective ways to structure educational experience.
When the study was over, the product managers watched videos of everyday people, living in what one of our clients described as "perpetual anarchy." Children, toys, television, noise, and chaos were everywhere. It's one thing for our stakeholders to read the transcript of an interview with a single parent describing living in a one-bedroom apartment with their four children. It's entirely different to watch that parent in real time breastfeeding one child, disciplining another, calming a third, and teaching a fourth with homemade flash cards, because they couldn't afford store-bought ones.
Texts to Prompt Behavior
Selfie prompts are the texts we send to our participants to provoke their behavior and recordings. These are short messages that describe what we want our participants to do and how we want them to do it.
Here's an example prompt that we used in the education study described above:
Good morning, Sarah! This morning, I would like you to think about the place in your house where you feel your child learns the most. Take me on a short tour of the space, and, when you're done, upload your video here: [link]. Please complete this activity by 2pm and text if you have any questions. Have a great day•
There are a few parts of the prompt that are important.
First, the prompt is casual, using pleasantries like "good morning" and "have a great day," and emojis. It's important that the participant views us as part of the regular cadence of their life, not as an inconvenience or imposition, and these types of normal text elements are familiar. We change our tone for younger audiences, occasionally using shorthand text that might be more familiar to them.
We're consistent in our tone throughout the study, so each participant understands they are interacting with one researcher, and we reinforce that personal tone by using first-person references ("I would like…," "Take me on a short tour…"). And we offer ourselves for help through text.
We have some pragmatic elements in the message, like a deadline and a place to upload the video. We already described deadlines and the file-sharing tools we use to our participants during a preinterview, but we reinforce them in every text.
The most important part of the prompt is the request for action. This request needs to be extremely short and directed, without unnecessary language or qualifiers, and focused on action rather than conversation. Phrases like show us or demonstrate or let us see an example lead to rich, powerful, and thorough views into participants' lives, while more-introspective instructions like discuss or explain can lead to thinner, shorter, and less-engaging responses. Participants respond well to two prompts per day (one in the morning and one in the afternoon) and remain engaged in the process for as long as a week with the correct incentives and prompts.
The Acceptance of the Selfie
In the past, my team would have depended on retrospective accounts of insurance purchasing, and on photographs, audio recordings, and transcripts of our educational sessions in people's homes. This would have generated solid, actionable data. But I've found that it's nothing compared to video, in particular the blurry, shaky, first-person, interrupted, messy, poorly lit, and real selfie.
We first tried video diaries about a decade ago, and it was rough. Participants were nervous about taking videos of themselves, and the results were cold, impersonal, and lacking the expansiveness of a contextual interview. Participants struggled with saving and uploading videos; many participants didn't text on a regular basis and so they had difficulty receiving and acknowledging our text message prompts. Additionally, our clients themselves were less familiar with the medium and skeptical of the process.
But a lot changes in 10 years. Most of us have welcomed video into our lives in such an integrated, open way that nervousness around selfies isn't an issue any longer. We've gotten comfortable sharing intimate parts of ourselves with strangers. Covid made even the most technically avoidant of us somewhat capable of using devices and software. Even those of us who are camera-shy have gotten used to seeing ourselves on-screen. And many of our clients are themselves using platforms like TikTok, sharing intimate videos of their lives and families.
The rawness of the data is the key. Traditional design research generates presentations, spreadsheets, and write-ups. These are tools that sterilize human behavior to fit into a corporate context. Selfies break that mold. The selfie study is effective because it provides a raw, intimate view into people's private lives. I've been surprised to see the most quant-driven clients forget about a sample size of one when they see an out-of-focus, clumsy video of someone's living room. Selfies are a form of acceptable voyeurism, and they give us an emotional tether to our customers that's unlike any traditional market or design research methodology.
For more information. This method has a large operational overhead, primarily related to tracking and managing moving parts. I've written a thorough tutorial of the mechanics of the technique, and you can read it here: www.wonderfulnarrative.com/selfie-tutorial.
Author
Jon Kolko is a partner at Narrative and the founder of Modernist Studio, acquired in 2021, and the Austin Center for Design. He has written a number of books, including Well-Designed: How to Use Empathy to Create Products People Love (Harvard Business Review Press]. [email protected]
Copyright 2025 held by owner/author
The Digital Library is published by the Association for Computing Machinery. Copyright © 2025 ACM, Inc.
Post Comment
No Comments Found