How We Use Data to Inspire Design


Original article:

When most people imagine good design, numbers probably don’t come to mind. In fact, anything quantitative might feel completely at odds with the concept of beautiful design. But at IDEO, in addition to connecting with people and learning their stories, designers use quantitative data as a tool to gain empathy and inspiration. We learn from numbers the same way we learn from people, because we see numbers as a representation of people.

In our traditional human-centered design process, we empathize by going where people live and work. We talk with extreme users. We immerse ourselves in their lives. We explore the tension between what people say versus what they do. We prototype. As the world becomes increasingly digital, data becomes a natural byproduct of people’s lives. We’ve learned that the qualitative process we traditionally use cannot only be strengthened by this quantitative data, but can also uncover insights that qualitative data alone cannot. Quantitative data is a rich, ripe source for design research that IDEO is using (and you should too!) to get inspired by users.

How might we use quantitative data to inspire design?

Talk to extreme users

If we wanted to learn how to improve a product, would it be better to talk to someone who feels indifferent towards the product or someone who hates it? At IDEO, we prefer to talk to the extreme users, the “haters” and “super users.” Why? We’ve learned that the needs of extreme users are amplified. They really need something to be one way or another. There’s no in-between. They can clearly articulate what is amazing or awful about the product and show us the workarounds they use to make up for the product’s weaknesses. This helps us pull out meaningful needs that may not emerge when engaging with the average user who represents the mean.

Quantitative data is perfect for helping designers determine who the extreme users are to better understand what makes them stand out. We can analyze the behaviors of people in the bottom or top quartile of the dataset. We can also use the data to guide our qualitative research. For example, one startup we were working with was questioning whether or not to incorporate a feature into their product. The data showed that most people used the feature only ~1.5 times, so their first thought was to get rid of it. Rather than focusing on the “average” user, we looked at the whole distribution. We spoke to the two or three people at the extreme who maxed out the use of the feature to learn what value the feature was providing and how to tweak the feature (and the messaging) to make it more helpful to others.

Immerse yourself in people’s lives

There’s no better way to understand the people we’re designing for than by immersing ourselves in their lives. Why? Because observing what people do and how they interact with their environment gives us clues about what they think, feel, and need. This helps us uncover insights and inspire new design solutions.

Digital products enable designers to collect an endless stream of quantitative data to immerse themselves in the lives of users. With permission, users can take us along with them wherever they go — even when we’re not physically present — and share their needs and desires in real time with us. For example, IDEO worked with an automobile company to understand how people move through cities. To immerse ourselves in users’ lives, we asked them to download the Moves app which tracked how they moved through their city. We also asked users to take photos and videos of moments when their trips went wrong, and moments when their trips made them smile. For every major travel incident, we were able to connect the quantitative data with the photos taken in the field and the human stories. This way, we understood the insight wasn’t just the numbers: it was grounded in human experience and connected to a larger system.

Explore the tension between what people say versus what they do

Good design is often built on a solid understanding of both people’s explicit and latent needs. One way we uncover the needs and values that may not be obvious to the people who hold them is by listening to the stories people tell and comparing them to observations of what they actually do. The differences often indicate a latent need that might not otherwise be expressed.

In addition to traditional ethnographic observation, designers can capture the tension between what people say and do from both survey data and behavioral data. Survey data lets us understand people’s thoughts and attitudes about a subject. Live data (i.e., from a website or app) helps us observe actual behavior in a real-world context. For example, an IDEO team was developing an app to identify levers of behavior change to coach safe driving. In surveys and interviews, the team heard people claim they were good drivers. However, the team observed through tracking behavioral data that those drivers were actually riskier. If the team designed opportunities based on what they heard, they’d design products to reward good drivers. If they designed based on what they observed, they’d design products to help people improve bad driving. However, the people who needed their driving improved wouldn’t have used the product because they believed they were good drivers. This insight helped us design for that tension.

Create a prototype and see how it resonates

Prototyping gets ideas out of designers’ heads and into the world. By building rough prototypes that participants can see, touch, feel, and react to, we can rapidly elicit feedback and test the functionality of early design ideas. This is done overtly, through tactics like questioning and role playing; and tacitly, through observation. After asking for feedback from a number of people who represent different types of users, we synthesize it to find strategic and directional themes to guide our teams in further development.

Traditionally when we built prototypes (anything from paper wireframes to physical retail spaces), we relied on our own observations and conversations with users to understand how the concept resonated. Now, digital tools enable us to collect and leverage quantitative data in our prototype as well. We use live data and integrate it with tools like Slack to prototype digital experiences that feel real for users, but are actually very rough and rapid to create on the backend. These “Wizard of Oz” prototypes enable us to learn quickly and investigate a lot of different possibilities.

For example, an IDEO team was working on a behavior change project and wanted to design a messaging system. Rather than spending time coding a system that would automate responses, we created a ‘magic’ automated experience to quickly prototype effective messaging statements. First, we hooked up the Slack interface to import user data so designers could quickly understand it. Then, we used that data to create, send, and receive personalized text messages to help improve users’ behavior. Based on the the user’s qualitative responses via text messages and behaviors captured via quantitative data, we were able to figure out messaging statements that resonated with users. By creating prototypes that collect both qualitative and quantitative feedback, designers are able to get a better, more holistic understanding of how a concept resonates with users.

Inspiration comes from a variety of sources including observations, conversations, and quantitative data. Quantitative data can be used for direction and inspiration throughout the design process, as it helps us understand where to focus and how and why people behave the way they do. It also serves as an input for prototyping. And together, with qualitative data, it provides us with a complete, meaningful story. The combination of both heightens insights to ensure that the weakness of one is balanced by the other. So the next time you turn to data, consider how you can merge both the people and the numbers to get to the heart of your design challenge.