How to improve someone’s diet — a UX case study
Our ideas, experiments and learnings.
Research tells us that improving someone’s diet is the single biggest change that can lead them to successfully reversing their type 2 diabetes.
“You can’t exercise your way out of a bad diet”
— Our Head of Product
So we set ourselves the challenge to see if we could improve the diet of someone who’s been recently diagnosed with diabetes.
Here’s the approach we took and some of our ideas, experiments and learnings.
Stage 1: Empathise
We conducted interviews with people recently diagnosed with diabetes to find out how they felt at the moment of diagnosis, and what was standing in the way of them making improvements to their diet.
We found that:
- People are overwhelmed at diagnosis — they want answers
- People have misconceptions about healthy food — they believe it’s not tasty, too expensive and too difficult to prepare
- Successful people have support when implementing changes to their diet
Stage 2: Define
We translated our research insights into three How might we… statements that defined the problems we knew we needed to solve.
- How might we… expand someone’s knowledge about healthy eating?
- How might we… change someone’s misconceptions about healthy eating?
- How might we… support someone in implementing changes to their diet?
Described using the How might we… format, our problems became immediately actionable.
“The ‘how’ part assumes there are solutions out there — it provides creative confidence. ‘Might’ says we can put ideas out there that might work or might not — either way, it’s OK. And the ‘we’ part says we’re going to do it together and build on each other’s ideas.”
— Tim Brown, founder of IDEO
Stage 3: Ideate
Ideation sessions should be well-planned, quick and fun. The goal of them is always to generate as many ideas as possible – because the more ideas you have, the more likely it is you’ll be able to converge on one that works well.
We decided to tackle How might we… statements 1 and 2 first. We held a separate ideation session for each using our own flavour of the Google Design Sprint Crazy8s exercise: Musical8s (it’s like Crazy8s but lasts 16 minutes instead of 8 and uses a carefully curated playlist to increase inspiration).
Once we had lots of ideas, we used a behavioural design framework to analyse and combine them, and then converge on a couple of solutions that both solve the problem and are built on the principles of behavioural science.
Our ideas
Here are the two ideas we converged on, described using hypotheses taken from the Hypothesis-Driven Designer Canvas we filled out for each.
The canvas is great for describing ideas – it ensures the whole team focus on the reasoning behind why they might work.
Idea #1: “Lessons”
Our hypothesis: people are motivated to learn, so if we give them a structured framework to learn about what they want, at their own pace, then they will expand their knowledge.
Idea #2: “Learn By Doing”
Our hypothesis: people’s misconceptions block them from even trying to eat healthily, so if we send them free ingredients for healthy recipes then they will give them a try and discover that healthy food can be simple and tasty.
We’ll leave the third How might we… statement for now as we decided to iterate on these ideas to tackle it later.
Stage 4: Prototype
The other great thing about the Hypothesis-Driven Designer Canvas is that it ensures you know what you need to test. And if you know this, then you already know what the scope of your prototype needs to be.
“Lessons” Prototype #1
For the “Lessons” concept, our hypothesis was centred around providing a structured experience that people would use to learn. This meant our prototype needed to be something people could interact with over the course of a number of days, and get real value from.
First we built a 7 day email course using MailChimp. Each day we sent subscribers a lesson about a particular aspect of adopting and maintaining a healthy diet.
This performed well. People learned new information and, importantly, retained it.
Some people questioned their previous beliefs. And some people felt empowered to go on and make changes.
“Lessons” Prototype #2
Next step was to put this experience into an app — to test our hypothesis that having a choice of subject and learning at your own pace were key to retaining more knowledge.
We built a simple Android and responsive web app, with a choice of interactive lessons, each with a simple test at the end. We promoted it to specific target users using Facebook ads and then monitored how people used it and how many test questions they answered correctly.
“Learn By Doing” Prototype
To test our “Learn By Doing” concept, we sent a small recipe box (thanks Gousto!) to each participant accompanied by an email that explained what it was about the recipe that made it healthy.
We encouraged people to share a photo of the recipes they cooked to see the impact this would have on their motivation.
After sharing, we then suggested people could buy the same ingredients from their regular supermarket and cook the recipes again.
We followed up with them a week later to see if they did.
Stage 5: Test
“Lessons” Concept
We learned that our “Lessons” concept — a service that gave people a structured framework to learn about what they want, at their own pace — did help people to expand their knowledge.
We also learned:
- Some participants had previously tried to learn by themselves, unsuccessfully, but found the way we structured and presented information to be more useful.
- Participants tended to refer back to the lessons after completing the study, helping them to maintain diet changes after 2 weeks. Some of these people had made unhealthy diet choices but were aware of them, which they weren’t before, indicating that knowledge had been retained after the study.
“I have eaten lean meats and fish. I have swapped unhealthy snacks like cake and biscuits for healthy ones like popcorn and seeds. I have eaten many portions of vegetables such as broccoli, leeks and kale.”
— A participant in the “Lessons” experiment
“Learn By Doing” Concept
The results of our “Learn By Doing” experiment — testing a service designed to convince people to try a healthy meal by giving them free ingredients and instructions for how to cook it— indicated that giving free ingredients was effective at helping people overcome the most common barriers to adopting a healthy diet. All participants either did cook their recipe box meals again or showed strong intent to buy ingredients in their next shop in order to cook them again.
We also learned that the service both helped people learn and feel more confident — two of the three influencers of behaviour change (the other is importance – see the Mojo behavioural design framework above). Participants were excited to share photos with us and some even decided to share their photos on Instagram or with family and friends. These participants commented on the positive support they received as a result and the contribution this had to their confidence level.
Interestingly, people also commented that the supplied ingredients and images of meals included in the instructions helped them to learn about portion control.
And many mentioned feeling inspired to exercise and walk more as a result of having enjoyed cooking and eating healthy meals.
What next?
Out of the two concepts, “Lessons” is a more financially viable solution for us — we have the capability to build a service like this, and an existing platform through which to distribute it.
We want to iterate on it by exploring more interactive lesson formats — like those used by language learning and brain training apps — as we believe it would further improve knowledge retention and enjoyment.
We also need to figure out how this type of service might fit into the existing experience we offer with our platform. I don’t mean how it fits into the IA, but how our existing features need to evolve to meet the user needs defined by the third How might we… statement and the other quadrants of the Mojo framework.
Follow me if you’re interested to see how this project develops, or email me at olly@tictrac.com if you want to know more about our approach or the concepts.
