I wanted to explore the influence of aesthetics on how persuasive an app can be for my master's thesis. To do this, I reviewed literature on persuasive mobile apps. Then I decided to tackle sedentary behaviour. Why? First, people were sitting for longer times, and second, it would be easier to observe how they react to prompts in a short-term study.
I was familiar with behaviour change theory and persuasive design principles from prior projects, and visual appeal is considered to make a system more persuasive. There's also literature on aesthetics and its effect on usability; people find better looking systems to be more usable; just like how people associate higher quality traits with 'better looking' individuals.
There was one more thing though. Most persuasive frameworks were around using positive reinforcements, but there were studies who suggested use of aversive stimuli would make it even more persuasive, making positive reinforcement to pop more. And in a previous study, we did confirm that. So the decision and motive to design it with both positive and negative messages were there.
Can a mobile app following persuasive design principles, while also utilizing aversive stimuli, succeed at persuading its users?
Does aesthetics influence the persuasiveness of an app?
So how was I supposed to design an app that was persuasive and that looked good? Knowing how aesthetics also affects usability, I had to be careful. If it was 'too ugly' it could hinder the usability, and if you can't use it, well, you know, you wouldn't even get to being persuaded.
Thankfully, the literature had something to say. So I used Gestalt principles (set of principles that is based on how we group and understand visual information), colour theory, and typography to change aesthetic levels. And to make it persuasive, I used Persuasive design principles.
This is how it all began, from the Paper app on iPad to finally on Affinity Designer.
I created the prototypes using Origami app on macOS, and tested each variant with 12 participants, total of 48 participants.
It was a modified take on usability testing to incorporate additional tasks (so they were busy doing things when notifications appeared to observe whether they interrupted their tasks). I asked them to think aloud as they completed the tasks, then at the end completed a questionnaire where they rate usability and aesthetics of the system.
Potato is a favourite!
Persuasive group participants were more motivated than the control group participants. They were emotionally connected to it, felt sad and responsible; they wanted to stand up to help it! Control group participants would dismiss notifications, saying "it can't tell me what to do!', but when there was a potato, a sad potato in danger, things were different. So it worked!
"Graphic designer's nightmare"
said one of the low aesthetics participants. They commented how bad it was for about half an hour. Yet, it didn't seem to have a significant enough effect on persuasion. The questionnaire results showed they were rated worse than the higher aesthetics variants for how it looked, but not by a huge margin. So maybe, despite everything I did, they weren't "ugly" enough?
Usable all across
The results also showed all variants had similar usability ratings, which was awesome, mission accomplished! Comments and deeper look into results show similar stories though; less attractive variants were less good, and people wanted to change how the low aesthetics variants looked like.
Yes aesthetics didn't significantly influence reported persuasion, but, based on the comments, maybe it's not about that directly. If it looked ugly, people wouldn't even get the app (I was forcing them to use it at the end), so maybe that's the key!
So if your persuasive app is ugly, then people will just skip it, even before using it and getting exposed to all the persuasive awesomeness.
There's no perfect study (I think), and it is very crucial to admit what you could have done better. In this study, there are two things.
First, to ensure greater difference in aesthetics levels between variants, I could create multiple designs, test it with people and use the highest and lowest rated ones. I had limited time so that I couldn't.
Next, it was too short. People only used the app during the session, and although they reported being persuaded, a longer term study would show the effectiveness of the app. Maybe people would get bored over time, maybe they would find notifications at some point, who knows?
The study got published in the Graphics Interface Conference in 2018.