Many of us have tried to change our health behavior during this pandemic. You might have downloaded a well-being app to help you change toward a healthier lifestyle or routine. However, in the end, the change you wanted just did not stick, and you stopped using the app. The question then arises, how can well-being apps increase user engagement over a longer period of time?

Making lasting changes is hard

In any effort to do something new, roadblocks come our way. Making changes in our lifestyle is a battle and most of us do not succeed at it in the long term. This outcome is also reflected in the low retention rates for many well-being apps (see Figure 1).

Figure 1. Retention rates for calorie, fitness, and sleep apps.  
Source: health tracker installs and retention data

Most of us think when creating new healthy habits, it is a matter of willpower. Mainstream well-being apps adopt this strategy as well and fail to understand that change is rarely a linear process. When we (inevitably) fail to change our behavior using this strategy, we lose a little bit of trust in ourselves and in the well-being apps. We might suspect we do not have what it takes, and we are just the kind of person who does not follow through on things. So we stop putting effort into the new behavior and disengage from the app.

So now what?

For well-being apps to be effective in helping us change our behavior, maximizing engagement is paramount. The app needs to get to know us, understand us and use these insights to our advantage.

In order to change someone’s behavior, we must first understand their context.

A well-being app succeeds when its content is tailored to:

• Your starting point;
• Your goals;
• Your specific habits;
• Your day-to-day life; 

With these things in mind, you are more likely to:

• Pay attention to the content;
• Come back to the app;
• Change your behavior. 

The goal is to address the personal need for support, whenever this need arises. 

Mobile technology can detect the habits and context (e.g. when a person comes home from work, enters a supermarket, etc.), and can provide interactive, tailored behavioral change interventions on a large scale (Figure 2).

Figure 2: Customer behavioral insights through mobile technology

Characteristics such as mobility patterns (e.g., home-work commute, pick up and drop off the kids from school, transport mode: car, bike, train, tram, etc.), physical activity levels (e.g. someone is a jogger, goes for a lot of walks, is often at the gym, is inactive, etc.), and characteristics of a person’s profile (e.g. workaholic, early bird, parent, etc.)  are used to create the optimal behavior change intervention.   

A notification to go for a walk is an example of a mobile behavioral change intervention to increase physical activity (see Figure 3). For a mobile intervention to be engaging and effective, it should adapt over time to a person’s changing circumstances, which is gathered through mobile sensing. Not only should the intervention be personalized to the contextual elements (e.g., time, place, social context), the content of the message should also be tailored to the individual. Different people require different approaches to reach similar goals. 

For example, we can observe characteristics like "workaholic" and "limited physical activity" for a person. A well-being app could suggest different 15 minutes walking routes on a map on multiple occasions. One day the app could suggest a walking route around her working environment during her lunch break. On another day a walking route can be proposed when she arrives home from work, and on a weekend a longer walking route could be suggested.

Figure 3. The difference between generic and personalized interventions

For this person, the suggested walking routes after work around her home are most effective. The well-being app learns this and minimizes the notifications during the lunch break, and maximizes the notifications when the person arrives home after work. The well-being app accommodates the real-life setting of the individual, eventually turning the walking behavior into a habit that fits that person’s schedule. 

Taking into account the person’s context and response to previous interventions (e.g. changing the timing, frequency or strategy), the well-being app maximizes engagement of interventions over time and promotes long-lasting behavior.

By making the target behavior easier to do, in this example 15-minute walking routes, the more likely she is to perform the target behavior and continue using the well-being app.

Conclusion

This blog only scrapes the surface of how digital interventions can bring about behavior change. 

The key takeaways are:

• User engagement is necessary for behavior change.

• Personalizing the experience is an excellent way to increase user engagement.

• Take into account the person’s existing habits and context to promote long-term behavior change.

By keeping the individual in mind when designing well-being apps we can increase engagement in well-being apps and effectively promote long-term behavior change and provide people with the support they need.

In the wise words of the philosopher Socrates, who said the first (philosophical) priority is to know yourself. We at Sentiance believe in a technology that tries to get to know us as an individual, understand our true potential, and coach us in a way we need to be coached in order to realize our full potential. The key for businesses to build trust with their customers is to make sure that the customers' data is used in an ethical way and improves their lives while ensuring data privacy.