Psychology, AI & Games: Reshaping Healthcare Design

Maria Borysova at UXDX Community: Designing for Real-World Impact. Video: https://youtu.be/GPMzs47V6YQ

Readable transcript: edited from the recording's captions for readability (fillers and false starts removed, punctuation and section headings added). Wording is the speaker's own. Timestamps are positions in the video. Names marked [?] could not be verified against the audio.

From knowing to doing: the motivation gap in health

[00:00:08] Thank you so much for the intro. A few words about me. I've been working with healthcare companies, helping them design products and services, for the past seven years. I had a chance to work with smaller companies, helping them bring a product from scratch to launch and then through the growth stage to the unicorn stage, and I also had a chance to work with corporations like CVS Pharmacy and others. Today we'll be diving into gamification in healthcare.

[00:00:40] To be honest, games have a powerful influence on our behavior. When I was a kid, my mom found a computer game called something like Clean Teeth, so I could learn to keep my teeth clean by playing fun simulations. I was running around the teeth, cleaning them to prevent cavities. These simple games stick with us because they tap into a very human behavior: we love challenges, but we hate being forced into them. That's why, many years later, I still remember this game.

[00:01:16] Our health choices are significantly shaped by our upbringing. We observe what our parents do, how they behave, whether they work out or not. Later on, when we are exposed to other stimuli, we change what we learned earlier, and maybe we can improve our health. In recent years we can definitely see that more and more people, especially millennials and Gen Z, are more interested in investing in health and wellness purchases. We have a lot of information about how to be healthy, how to eat healthily, how to ensure that our lifestyles are as optimized for longevity as possible. But at the same time, we don't really know very well how to maintain these healthy behaviors consistently.

[00:02:12] So the problem is probably not knowledge, and not even a lack of interest. As we can see, Gen Z and millennials definitely want to invest in this. The problem is motivation and engagement. We will be exploring how gamification can bridge this gap between health knowledge and actual health action.

[00:02:34] Gamification is essentially a combination of different effects. There's a challenge. If we win the challenge, if we meet the conditions that were set in it, we get a reward. Then our position shifts somewhere on the leaderboard. We may receive rewards, for example badges or trophies. Then we want to share it with our social network. Then our status gets updated. And everything is centered around the point system. This is exactly how we understand whether we are good enough to receive a reward or a badge or not.

Evolutionary origins: status in the tribe

[00:03:19] If we go back to the tribes, and ask how it happens that we are so obsessed with games, achievements and recognition, we can trace interesting evolutionary origins. For most of human history, survival was impossible without group membership. Early humans faced predators. They did not have all the resources they wanted, unlike now, when we can have any food. For us right now it's not about calories being consumed, it's about calorie restriction. There were also many other challenges that required collective action, and being rejected from the tribe often meant death, quite literally death. Not like right now, when it's mostly emotional damage and that's it.

[00:04:20] This created pressure to develop behaviors that secured our position within the group, within the tribe. High-status individuals received the best treatment: the best food, the best sleeping positions, safety and great influence. And this shaped the way our brains work even now. Our ancestors had to be very sensitive to social approval and disapproval, because if they were not approved by the group, by the community, it's quite possible that they would die alone.

[00:05:01] That's why our brain's reward pathways respond strongly to social approval. It is ingrained in human nature. We release dopamine and other chemicals when we receive positive social feedback. For example, we have achieved something, the community has approved us or our achievement as valuable enough, and we release dopamine. On the other hand, social rejection activates the same brain regions as physical pain. So it's already part of our nature.

[00:05:46] How it manifests right now is that we pick up our phone, and the difference between, for example, social media addiction and addiction to our health apps is not very big. We pick up our phone, check our status, and based on what we see, either we've received approval from the community, from our fitness app or nutrition app or our coach or anything else health related, or not. For example, we have achieved our health goal, dopamine increases, we feel happy, and then after some time dopamine decreases.

[00:06:28] There is no huge difference in how this works, whether we're speaking about health-related products, social media, or other maybe unhealthy behaviors. We are always triggered by something that helps us release dopamine and feel good, and after that we keep repeating the same behavior.

Leaderboards

[00:06:53] Going back to what I was talking about in the first place: we must ensure that our position in the community, our position within some tribe, is high enough. That's how we survive. So we will obviously release dopamine when we see, in this case, "You are in the top 1%," from Strava, as a reward, as recognition, as understanding that we are so high on this leaderboard. It's virtual, really. We don't even know these people in this tribe, but we already understand that we will not die, per se. We are in a very good position.

[00:07:41] Going back to leaderboards, almost all health-related and fitness-related apps use them. This one is a screenshot from the Whoop app. The same can be found on Peloton and other apps. They use leaderboards and badges to stimulate users to progress within the community, and I can compare myself to others. I can check the details and actually see where I stand in this social hierarchy. The same in this leaderboard from Strava.

[00:08:18] One important distinction I need to make here is that these leaderboards look like static screens, but they're never static. They were never static in traditional societies, and they're not static now. Positions can be gained and positions can be lost. If you performed well yesterday, for example you killed a huge animal, or you helped to cure a kid, or something else that positioned you well in the community, this works only until the next person demonstrates a value bigger than yours. So you're always in a perpetual anxiety cycle about maintaining your position. It's not new. It was always the problem in the tribe. You always had to keep an eye on the leaderboards, same as we do now.

Rewards, trophies and delayed health benefits

[00:09:22] Interestingly, if you think about rewards and trophies, they are the natural next steps from these leaderboards. As parts of tribes in our ancestral environments, we received natural feedback when we did something. For example, if a human successfully hunted, they received immediate feedback: they received food, meat. After that, they most probably received social recognition in their tribe, because they brought so much meat and everybody was happy about it. As a result of that, they received elevated status, and this again triggered dopamine release. So the feeling that I've done something right reinforced the behavior, and obviously this worked. That's why we have this similarity.

[00:10:20] This immediate feedback loop is very important for learning and adaptation. Finding food, shelter, avoiding predators: we used exactly the same reinforcement mechanism for all of this. Dopamine release, learning, repeated behavior. Right now it is much harder to get this dopamine release when we're speaking about healthy behaviors like regular exercise, proper nutrition or even preventive healthcare. The results are delayed by weeks, months, even years. So why would I engage in some healthy behavior right now if I'm going to see results in ten years, or maybe when I'm 70? There is no real feedback right now, and these rewards and trophies at least try to bridge this gap.

[00:11:19] These are some examples of rewards and trophies that we receive in sport apps and health-related apps, in this case for different challenges that we can accomplish. Again, we receive these trophies and rewards, we get the dopamine release, and that's how we keep repeating these behaviors.

[00:11:47] Interestingly, if we go back to tribal recognition of our rewards, and the case when I came home with a lot of meat because I was such a great hunter, we can trace this in fitness and health-related apps as well. Many weight-related apps, so dieting and nutrition-focused apps, fitness apps, even sickness-related apps, will quite often include communities where we share stories about why something has worked. Usually we obviously want to establish our status, so we want to share our achievements, so again we're positioned better within the community. We can post about our marathon completion, and in this way we signal that we are providing some value, similarly to how we were providing value many, many years ago in the tribes. This way we highlight our achievements and make sure that we're earning recognition within our groups.

Feedback loops and the gap between action and reward

[00:13:02] All of this is closely tied to the point system we've already started talking about. We know there is a reward feedback loop. I perform some action based on a trigger. First there's a trigger, then some behavior, I receive a reward, and based on this reward I decide that I want to keep doing the same thing, keep repeating my behavior, if the reward is something I feel is meaningful and I enjoy it. For our ancestors, this made sense in their daily behavior. For example, again, I hunted an animal, I got a reward from the community, and then I keep going and doing the same.

[00:14:00] This is a very simple example, but for more complex behaviors it's exactly the same idea. The feedback loop is very short, so we get feedback immediately. I found great shelter, it is safe, we can avoid predators. I receive feedback that this is great and I should keep doing this, keep making these kinds of decisions that help the tribe to survive. I'm the best mom keepers area [?] in the tribe.

[00:14:32] The benefits of healthy behaviors, for example regular exercise and nutrition, are delayed. So we're breaking this habit loop, because there's no way I'm going to remember what behavior triggered a reward if the behavior was done ten years ago, or sometimes even months ago, which is long enough. There is a significant gap between action and reward, and that creates a huge motivational challenge, because in this case a human cannot really learn the way we were designed to learn.

Streaks, progress bars and point systems

[00:15:13] This brings us to an interesting question: how can we bridge this gap? We need to signal to the human, to the user, to the patient, that something is going right. Some great examples of this are streaks, progress bars and other ways to signal that the action the human has performed was great, was positive, was healthy. Progress bars work in the same way: they update in real time with a signal about the progress I've just made. Streaks give us immediate feedback about something that was done a second ago. For example, I've logged a healthy activity, I've finished my workout, or I've had a very healthy meal.

[00:16:09] All of these examples, streaks and progress bars, make sense and they help us, but they're quite limited. Health behaviors are very abstract. Reduce the risk of heart disease: how do we measure that with a streak? A streak is only about counting the days that I engaged in some behavior. That's not enough for us. We need more depth. A progress bar is also not enough. We need to understand what we're working towards, and gamification helps us with this. We transform abstract benefits into something measurable and concrete.

[00:16:52] For example, this is the Weight Watchers app. They have a unique points system, which defines the number of daily points you can use to eat certain types of food, and based on what you log, you use these points. In exactly the same way, ZOE Health, which also helps with dieting and healthy nutrition, gives us feedback about how healthy the food we've just logged is. There's a specific number tied to a specific food or workout or any behavior we're tracking within the app. We collect these points, play the challenges if there are any, achieve these progress markers, and immediately our brain experiences satisfaction.

[00:17:52] To be honest, I can even share my personal example. I was recovering from a knee injury, and it was a very long healing process. It was almost unbearable, because I could not see the progress. But when I started tracking progress markers on a daily and weekly basis, I could actually see how I was improving. Even those tiny increments, tiny changes every single day, compounded into a strong upward trend. These numbers provide a structure that we lack in very abstract health concepts. The process does not feel endless. It's not like I'm logging this food and I don't know how great it is or how long I should keep going.

[00:18:45] Another great example is point systems used to reinforce healthy behavior. Again, in the health app, after I log a meal that might not be very healthy, they give me immediate feedback on what should be improved about it. That's a very nice gamification feature, and it might be enabled by AI as well. We understand what was included in this dish, and we suggest to the user, for example, instead of using very sweet fruits, use something less sweet, or add nuts. This way the user learns immediately. There is no gap.

[00:19:30] Another great example of a point system is Wings [?]. They track multiple health behaviors, and there's also a compound health improvement score. Practically, I earn points for engaging in the healthy behaviors I've decided to track within the app, like eating healthily, and they're all aggregated into one health improvement score that I look at every single day. It provides me with a short-term reward cycle and also a long-term one, because in the short term I may fail today, but I will not fail in a month if I keep doing something that is good for my health.

[00:20:14] Whoop is a good example as well. It uses daily scoring for sleep performance and workouts. Again, the same idea: quantifying my health achievements, the ability to influence them, and immediately seeing the feedback in the health app. A very nice example is Pacing [?], because this one is related to sickness and actually tracks exertion level. Based on heart rate variability, they track how tired we get performing certain actions, for example going for a walk or tidying up. If a person is sick, it's important for them not to get too tired. So they give immediate feedback, and we learn: if this behavior is bad for us, we won't engage in it, and if it's great for us, quite possibly we'll continue incorporating it into our lifestyle. The same goes for sleep, quantifying the quality of sleep.

Anticipation of reward

[00:21:26] Interestingly, I just wanted to quickly mention this one. It seems like right after the action is performed, we receive some reward; this is the first chart here. But right after the cue is recognized, when we know that we will perform some behavior and then get a reward, we already start getting the dopamine spike. Obviously, if the reward is provided too late, we see the decrease here, and different changes when we receive a reward a little bit early or later. But I wanted to highlight the anticipatory reward surge [?], which means it's not just about receiving the reward after I've done something right. Our brains have anticipation of rewards.

[00:22:22] How did this work back with our tribes? The response was there to motivate very complex, persistent [?] behaviors. For example, I needed to continue to track prey for days when I was hungry. In this example we're again using ZOE Health. Here we have weekly or monthly reports with analysis of what a person has done right and wrong and what should be improved, and this creates a state of motivated anticipation. Fitness apps do the same. We know that if I'm doing everything right today and I keep eating healthily or keep working out, I will see that my score for the week is excellent, and I anticipate it already.

Can gamification harm us?

[00:23:14] Which brings us to an interesting, controversial question: can gamification harm us? Obviously there are pros and cons to every method. On the one hand, immediate feedback loops motivate consistency. That's how we evolved. Looking at my personal examples and many examples from other people, even on a basic, shallow level, we keep moving every day just to complete a specific number of steps, so we don't lose our streak, or because we don't want to see a gap in the activity rings or anything else. So it does help us.

[00:24:00] However, we also need to keep in mind that health metrics, especially when we display them publicly through apps and social platforms, Strava again, become not just personal feedback. They become a social performance, and this creates pressure, and a focus on external validation rather than our internal, intrinsic motivation. One of the ways I think this can be balanced is integrating not only quantitative feedback but also qualitative feedback. For example, I can assess how I feel right now from one to ten, but I also need to say why. The same for a workout: how did it feel? So we create space for a narrative, for the story of why this happened, alongside the number.

[00:25:00] While numbers are quite valuable tools, we also need to design for humans holistically, and keep in mind that we are responsible for our products and our designs. Feel free to connect with me on LinkedIn or Medium, and subscribe to my Substack about design for healthcare, and not only healthcare, but mostly behavior design. Thank you.

Q&A

[00:25:30] Host: That was great, thank you so much. I have too many questions to actually get through, so I might get one or two in. The first one: Duolingo is famous for streaks, which you talked about. What company do you think is doing really well with gamification, in general, I'm assuming?

[00:25:56] Maria: Well, obviously, as they mentioned, Duolingo is doing great. I would say all the examples I've mentioned here, which are dozens, especially the ones that have been on the market for years, definitely do it well. Let's say Strava does it well, [?] does it well, Peloton does it well. Mostly all the top apps. You'll be able to learn something from their experience.

[00:26:36] Host: And specifically, there was a follow-up on healthcare. Which one would you recommend for healthcare?

[00:26:45] Maria: Well, healthcare is a very broad and big segment, so it depends on exactly what kind of problems you have as a user. You should definitely keep that in mind. My favorite would probably be Strava again, just so as not to introduce any new names.

[00:27:06] Host: Great to know. And the other piece was around complexity and challenges. You've talked about the benefits of gamification. What are the risks?

[00:27:22] Maria: That's a great question, because obviously the risks I mentioned, like having this social performance instead of being intrinsically motivated to perform a behavior, will harm us. We probably all know some people who are very anxious when they are unable to track something, or tracking doesn't go well, or the feedback they're receiving from the app is not what they wanted, and they become super anxious about it. They're optimizing for the metric rather than the value they actually want to get from it. So we need to keep in mind that we probably need to go much deeper than giving the user a number, which is a very shallow way to assess something, especially when we're speaking about such a complex and deep category as health.

[00:28:18] Host: Yeah, it's really valid, really interesting. Thank you so much, Maria. I really appreciate it.