DATA AS A CREATIVE MATERIAL

The Design + Data Lab is an ongoing experiment aiming to establish an open forum for innovation. We try to challenge our understanding of and approach to the design practice , and to create tools needed for designers to become leaders in a future driven by data.

 

As part of the our Design+Data Lab initiative, Topp hosted curious designers and developers from around Europe in an explorative workshop. 

The goal was to explore how we can use data, not only as a way to gather insights and feedback, but as a design material, possessing qualities such as malleability, sketchability, story telling, variety of fidelity found in other contemporary design materials like prototypes, wireframes, video scenarios, software, etc.

Over the past few of years at Topp we’ve been developing methods for incorporating real data into the process of designing new products and services for our clients. Developing new perspectives on how to capitalise on sensors, big data, and other sources of data has been significant as more and more data is created. Traditionally, when considering design and data together people think of practices such as A/B testing, demographic insights, usage monitoring, etc. Organisations at Netflix, Spotify, Facebook, or Google should be given credit for developing robust practices where data has been used to incorporate reflections of actual usage and sentiment into iterations of their designs.

However, we have observed that data has a another powerful opportunity: When data is used not only as a reflective tool to gather insights, but as material that can be incorporated into early design and product processes we open up creative and strategic paths which are otherwise left untapped. 

This design + data workshop is one of a series of initiatives to enhance the process and results of design-based ideation and sketching with data, and opening up new strategic areas for product development.

 

The workshop setup

Over the course of a day, participants were asked to develop and pitch a concept, but with the specific parameter that concepts must be evaluated based on how well data was incorporated in their creative process and the concept itself. 

At the end of the workshop teams selected the concepts which best satisfied the data criteria. This approach helped reveal which methods worked well, and which didn’t - each team of participants chose a different approach to their concept development, and the results resulted in varying results. Additionally, we provided attendants the opportunity to not only speculate about data theoretically, but actually develop hands-on experience with it as a creative medium for developing new product ideas.

Below are some of the insights that we gained from the workshop:

Look for questions, as well as answers.

 Whenever a question is answered it seems to spark more questions. Pursuing these new questions in turn generate even more questions. With every new iteration of this question-insight cycle the understanding of the underlying data increases. This helps to map out the problem space and to solidify the problem description. 

If you are using data in the ideation phase of a design iteration, the simple act of asking questions and start to look for answers in the data is enough to generate ideas. Asking the questions is more important than the answers themselves, which means that you don’t need any statistical significance. Small sets of data are better, in this case, as they are easier to work with. Later on you can validate directions and hypotheses with bigger and additional data sets.

Looking at the bicycle data sets we used in the workshop, we might for example ask what the average duration of a ride is. It turns out it's almost always lower than 40 minutes in New York City. We also notice that some are shorter than two minutes. How can they be that short? The answer is that the distance covered is just a few blocks. Now we know that there's a pretty big user group that just travels a few blocks per trip. Of course, we could ask many follow up questions, like how long are longest trips, are the trips made by returning customers or casual users, do the weather affect the rate of short trips etc.

 

Turn numbers into stories

By applying your human perspective and the amazing capacity of the brain to construct narratives, you can start to read stories from the data. It doesn’t really matter if the stories are true or if they hold up to statistical scrutiny, as long as they spark ideas that drive you forward. Don’t be afraid to adopt a view based on these stories and work from that perspective for a while. 

To create an experience that is useful and enjoyable you can’t rely solely on the realities that the data describe. Without the human interpretation it is just that; a description of some reality.

Looking at the bike sharing data again, we can speculate about why a handful of trips are longer than most. The bikes aren’t returned for a few hours, sometimes for a few days. Maybe the customers keep the bikes at home, or maybe they get stolen. Maybe the stands were full and the bikes couldn’t be returned. Whatever the reason each story we synthesise bring a new perspective.

 

Don’t go for perfect

You’ll probably never find a data set that describes exactly what you are looking for. You might need to combine several sets, or sometimes even make parts of it up. If you can’t find anything helpful, try to skip a step in the question-insight cycle described above by guessing the answers to your current question and repeat the process using the follow up questions. Maybe you can answer these follow up questions instead? It might be helpful to change perspective or to look for marker data, like how the heart rate is a relatively simple marker for many complex conditions inside the body.

You don’t need to be a data scientist to get creative insights from data. Forget confidence intervals and statistical modelling for a moment. A line chart will go a long way to understand some things about the data.

As an example, one of the teams wanted to work with household food waste, but couldn’t find any data with household granularity. If we could find out what the ratio between food waste and normal waste is, we could look to total waste levels for some answers. Maybe the restaurant industry reports levels that can be used as an indication of how household levels are changing? If we could find some smaller community (maybe turn to our own household?) we could extrapolate to the rest of the population from this data. This would of course be totally crazy talk if we were data scientists, but for the ideation process it works!

 

 

What’s next

As mentioned this workshop was part of an ongoing exploration in designing with data, specifically in the early phase of the design and product development process. We invite you to use this approach and share your insights and stories. Don’t hesitate to contact us for more details on our approach and insights. 

We’ll continue to explore other aspects of data being applied to the design process. Findings from workshops like these become incorporated into real world designs, and the methods will be further refined and communicated. We’d love to make this a collaborative effort; if you’re interested in participating in future events or hosting your own, just let us know.

We look forward to your comments and to hear about our findings.