Data-driven co-production: Challenging our preconceptions


Ben Hartridge | Researcher | @benhartridge 

Maria Portugal | Communications and Design Specialist

With Crisis Skylight Oxford, Maria Portugal and Ben Hartridge are designing new information and publicity materials to support outreach work in food banks, which the Skylight is pursuing as part of the Hothouses for Innovation initiative. It’s just a small part of our overall work but provides a neat example of our approach.

In this blog, Ben reflects on the importance to the Lab’s design work of listening to both data and users’ voices.



There is a moment of silence in the room as those present digest what I’ve just said. The silence is broken.

“I wasn’t expecting that.”

I’ve just shown a group of staff at Crisis, a national homelessness charity, a chart showing the demographics of people who used food banks in 2016-17. Before I pulled up the graph, I asked the staff which group of people were most commonly using food bank services. The group pretty much agreed. It must be single women with dependent children. When I show the chart, however, silence falls.


In fact, a study by the University of Oxford found that nearly 40% of food bank users are single men. Single mothers, while still the second largest group, only made up 13%.

Using evidence to challenge our own and others’ assumptions in this way is a vital part of the Dartington Service Design Lab’s (the Lab) approach to design. Most of us here are scientists by training; taught to be rigorous, sceptical and curious in equal measure. Ensuring we are not assuming falsehoods to be true, or misunderstanding the problem, is essential for designing something that will work. Had we continued with the assumption that most food bank users are single mothers, we would have been designing with the wrong audience in mind.

But data and evidence are not the only ways we check and challenge our assumptions. The Lab’s approach to design fuses research evidence with user experience and practical knowledge.


User Voice

A few weeks after I showed staff the food bank data, Maria and I are back in Oxford. This time, we’ve brought several copies of a prototype flyer along to the member’s coffee morning in the Skylight building (‘member’ is how Crisis refers to people who use its services).

I love going to these sessions. A relatively small group of members come each week and are hugely generous in offering the benefit of their difficult experience to improve Crisis’ work in Oxford.

I hand the flyers out around the room. “What do you think?”

They’re impressed, but there’s plenty that could be better. After a while, the conversation focuses on the map on the flyer’s reverse. It’s a screengrab from Google Maps.

“Could you show the whole train station on here?” “Yeah, where’s the bus station?” “When I tell people how to get here I say it’s next to the Job Centre.”


Clearly, for the members – who often don’t have access to a smartphone or a computer – Google Maps does not explain to them where Crisis is. Our assumption that people find their way to the Skylight using Google Maps – or are familiar with what the app looks like – has been shaken up. The question that follows is, ‘How are we going to use this new information in our design?’

The map on the flyer is going to be used by Crisis members. So, in the member’s coffee morning the following week, we asked that the members draw their own maps to the Skylight. The maps revealed the landmarks that are important for members when they navigate through the city. On the final flyer, the map will combine elements from all these maps so it is relevant and useful to Crisis’ members.

This short episode captures much of the Lab’s approach to design. It’s about challenging and testing our assumptions with data and evidence, and also user experience and knowledge. It’s about listening. Listening, and not charging ahead, blinkered, based on our own assumptions.

ReflectionBLOGBen, Maria