Back in May, I was inspired by BJ Fogg’s book Mobile Persuasion which described the phone as being used for 3 distinct purposes: entertainment, utility and behavior change (FYI, I don’t actually recall if it was exactly entertainment and utility (or communication) but I want to focus on the third one). In any case, the book detailed numerous case studies where the phone had been used to influence and/or (attempt to) change a person’s behavior. Not surprisingly, many of these case studies used basic gaming mechanics and psychology before the current over-buzz on the gamification of consumer services. For example, I remember one case study where tests showed that from a motivational standpoint, a garden that adds or removes flowers when you succeed or fail is better than a garden that comes to life and wilts when you succeed and fail.
Anyways, thinking about how I could use a phone to drive behavior change, I focused on building and prototyping a few concepts as part of YumYum Labs in June/July and the results were quite interesting. We started with a high-level goal of seeing if a phone could be used to improve one’s health (and this was particularly personal to me since I had gotten out of shape over the past few years from my days of having done a triathlon!). In any case, we started with a simple model:
1. Create goals
2. Make achieving those goals fun
3. See the reward
Goals was used as a broader term; it could imply both completing a task-oriented goal and/or a learning/educational-oriented goal. For example, a sample task-oriented goal for the day could be parking a few blocks further away from the office to encourage more walking. Alternatively, an educational goal could be learning about how to identify a high caloric food by answering various questions throughout the day on your phone (using a combination of repetition, images and playfulness to drive retention).
The sub-goals were dynamic and would evolve over time and eventually they would be trained into your behavior. For example, drinking 8 glasses of water a day or taking the stairs was not something I needed to think-about anymore. That being said, we knew the goals themselves may not be enough of a motivator and so we integrated a social element where you were grouped with a set of anonymous folks with similar goals (your team). Initially, you would think that friends and family served as the best team but the reality is they often served as the worst. It was your friends and family that would say, “Raj, you look fine, let’s go for a beer!” From our data, the best team or motivators were other anonymous folks attempting to achieve the same goals, almost virtual identities where you can be yourself and not be ashamed of your situation. The bonus was if your team was in proximity since you could actually meet these anonymous folks with similar goals.
Pushing along, we tested the solution with a small group of folks and uncovered some very interesting learnings:
1. To our surprise, goal oriented applications only worked with a subset of folks (almost self-selecting). It was a very type-A way of thinking (eg improving oneself). The largest revelation was that the majority of people are dis-motivated and are happy to complain (maybe too strong) about life and/or their problems but are not willing to do anything about it. As depressing as it sounds, I live in Silicon Valley and so I see very driven people but the reality is that the mass audience is not Silicon Valley (and it’s still not clear to me that a phone or any piece of hardware is sufficient to motivate you out of a rut).
2. I was once told by a registered dietitian (RD), that the mere act of thinking about being healthy at lunch will often results in a 50-100% improvement in the type of food you eat; this of-course doesn’t account for ignorance (eg a cobb salad can be worse than a burger) but the intentions are sound and the ignorance can be solved through education. In any case, the application would wake-up at lunch and dinner and remind you to eat healthy – this was maybe successful but what we realized pretty quickly was that the killer feature was not anything we had predicted. Chat and messages reigned; being able to speak to others with similar problems is what drove usage (and interestingly is what drives the usage of most online wellness sites). The goals themselves brought like-minded folks together but it was marginally successful in changing anyone’s behavior. However, the application itself could be considered a success because of the daily engagement driven by chat.
3. Rewards can be too good; I know that sounds crazy but we had not anticipated users gaming the app for the rewards (eg I will take a photo of my lunch to show my in-app social circle that it was healthy and to earn a monetary reward but then I go to McDonald’s afterwards and get a Happy Meal!). It was an interesting dynamic watching users play the game but ultimately it was not “conclusively” accomplishing the underlying goal of making oneself healthier: changing one’s behavior.
In any case, it is an experiment that is continuing to evolve but one personal conclusion is that I’m not sure if you can change behavior using a mobile device on mass scale. Sure, there will be a self selecting audience that will succeed (eg see all the fans of LoseIt) but the best motivator is not a phone or hardware device, it’s someone you see on a daily basis who is kind of an a-hole and is willing to monitor you everyday and “get in your face.” Unfortunately, a phone or any piece of hardware can’t extend beyond being a dumb piece of hardware (the reality of things).