Putting back users to the forefront: sustainable engagement tips from behavioral science
Apply to Irrational Labs’ Behavioral Economics Bootcamp.
Posted by Evelyn Gosnell
Watch out. If you’re an app developer, this is not a pretty graph. You might even stop reading, in what behavioral scientists call ‘The Ostrich Effect’, the desire to bury our heads in the sand in an attempt to avoid information we don’t want to confront.
What’s going on here? Why do people download a meditation app and then immediately proceed to … not meditate? Why do they download a fitness app and then … not exercise?
This is called the ‘Intention-Action Gap’, the all-too-human tendency to not follow through on things that we want or mean to do. It’s understandable. We download an app because we have good intentions, but then we get sidetracked. We have emails to respond to, a presentation to prepare for, groceries that need to be bought, children to pick up from soccer practice, an Instagram story to post. We’ll follow through tomorrow. There are all sorts of wonderful things we are planning to do…tomorrow.
Luckily, behavioral science can help close the intention-action gap, offering a toolkit to help change behavior for the better. Here are three ways we can apply lessons from behavioral science to drive sustainable engagement:
1. Set people up for success in the beginning
Part of setting up users for success involves creating the mental model they have for your product. Let’s say you’re in the language-learning space. The average person, especially one who has not learned another language before, may not have a clear idea of how to approach learning a language. How often should they practice? For how long?
There are two choices here. 1) You could let them figure it out on their own. 2) You could help them by creating a mental model on how to use your app.
See the following example from Duolingo below:
These labels of casual, regular, serious, and intense are entirely subjective, of course. But what this approach does provide is a frame, a starting point for the user to understand the basic idea that it should take between 5–20 minutes and they should practice daily.
Dating apps do a similar thing with the design of their interface. See the example from Bumble below:
This clearly sets up a mental model that the user should likely be talking to more than one person at a time. Whether six is the right number is a separate topic; what we can probably agree on is that limiting conversations to one at a time is not the right approach, and this design creates that mental model.
Another way to help set up users for success is by using implementation intentions. This is the idea that fleshing out the details of an action increases the probability of taking that action.
My favorite example of this comes from an experiment on voter turnout run by David Nickerson and Todd Rogers. In their experiment, some potential voters were randomly assigned to be called with a fairly standard script, such as information about the election’s time and poll location. A second group of potential voters heard not only the standard script but also were prompted with a few additional questions, asking them when they would vote, where they would be coming from, and how they would be getting to the polls.
What happened when election day rolled around? The second group had 4.1% higher voter turnout. Why? Thinking through the details of what it would take to actually show up to vote made them more likely to realize in advance, “Oh yeah, I’ve got that marketing meeting that always runs late on Tuesdays. I should probably go before work.” If they hadn’t thought this through, Tuesday might have rolled around, they might have gotten stuck in that marketing meeting, and never made it to the polls. The act of thinking about the details in advance can serve as a planning function.
Apps can use implementation intentions to help users follow through with the actions they need to take to stay on course. Take Shapa, a health and wellness app that helps people lose weight. Since a recent research from the University of Georgia has shown that stepping on a scale daily is correlated with weight loss, Shapa encourages its users to step on their scale. This is not an existing habit most users have, so during the onboarding, Shapa asks users to think about the next day. When would they like to step on their scale? The user can project themselves into the next day, think about the components of their morning routine, and figure out the moment that would suit them best. Shapa then sends its weigh-in reminders accordingly.
2. Measure the right things
Peter Drucker famously said, “What gets measured gets managed.” In the app world, this means that the metrics we focus on dictate how we design apps. And we’re often focused on the wrong metrics. Daily Active Use? Sure, we should probably know what our daily active use is. But should we allow it to be the primary metric that we’re obsessed with? Not necessarily. For some, it’s simply the wrong focus. For instance, it tells us nothing about whether users are actually progressing toward what they set out to achieve. And isn’t that why we’re here?
What we may want to measure is whether we’re helping users achieve the outcomes they want. Let’s go back to Shapa, which aims to help people lose weight. That is the desired outcome: weight loss. That is a key metric for Shapa to look at. After you identify the outcome, you back into how to achieve it: what are the behaviors that will be most likely to lead to the desired outcome? (There is often more than one). Then, you design the app around encouraging those behaviors.
For Shapa, one of those behaviors is stepping on the scale, as we saw earlier. The weigh-in screen takes a prominent place on the user’s dashboard. Users are also shown their historical performance on this behavior; they can see their weigh-in history in a section called ‘Progress Review’.
Another nice example was shared by Badoo at last year’s Playtime event. Badoo owns a group of dating apps and decided to test a feature where users could send each other hearts. What happened? It seemed like it worked: 20% of users sent a heart. Winning! Engagement metrics up!
But what happened when the team looked more deeply at the data? The response rate after receiving hearts dropped, -6% for men and a terrible -35% for women. The retention rate for women also dropped by 1%. This is another example that highlights why it’s so important to measure the right thing. Badoo has shared that the behavior it’s trying to drive is meaningful conversation. That is a very different thing than just “engagement,” or “daily active use,” and results in different design choices.
Another way to think about this is that if you’re trying to design for a behavior, you’re also trying to design against the opposite of this behavior. Given that Badoo is designing for meaningful conversation, they created a Bad Openers Blocker feature, showing small pop-ups with messaging like “you can do better” when the message would not likely lead to meaningful chats.
A third example comes from the fitness space. Imagine you’re designing an app that aggregates classes at various gyms and studios. Users have access to all these classes and can sign up through the app. What I’ve seen frequently in this space is designs that make it really easy to book the same class again, often looking like this:
The question is: is repeating a class really the behavior we want to be designing for? If we think about the desired outcome from the user’s perspective, it likely has to do with a) starting or maintaining a fitness habit, i.e. going to classes, and b) perhaps getting more fit. Trying a broad array of classes could increase the probability that the user finds classes he or she enjoys and will want to stick with, and novelty can help stave off boredom. We also know that the human body also benefits from continuously challenging muscle groups by varying exercise to avoid plateaus.
Understanding the desired outcome allows us to design for the right behavior — changing up classes vs. repeating classes — and the final design would reflect that. Knowing that the user previously took a class on Tuesday at 6 am and that this time slot is likely available, the app could propose an alternate class at that time. It could pull from its user data and find that people who like X class (original class user took) also like Y class, and then propose a Y class. In other words, imagine that user data indicated people who take barre also like pilates; in this case we could recommend a pilates class to a barre-class-taker.
All of these examples illustrate the same point: to drive long-term engagement, it’s important to be looking at the right (long-term) metrics.
3. Build trust
The third element of building long-term engagement is trust. Trust in apps is at an all-time low, a recent report from Common Sense media showing that 72% of teenagers think that apps are manipulating them. That trust is an important element of long-term engagement goes without saying. The question is how to think about trust, and how to build it.
Trust is partly about framing a relationship as a long (vs. short) term one and about acting in the other party’s best interests. I have a personal example from working for my boss, Dan Ariely, who incidentally spoke about trust in a recent TEDx talk. This story dates back to when I had first started working for Dan. I was paid a monthly salary — a standard arrangement. A couple months in, Dan said, “How about I just pay you now for the year? It seems like that would be easier than dealing with this monthly.”
Think about what that conveyed to me. With those words, Dan signaled that he saw our working relationship to be a long-term relationship based on trust, not a short-term, transactional one. It would be a risk for him to pay me upfront for work that I hadn’t completed yet, which is why it was a strong signal of his trust in me. As you can imagine, it only furthered my trust in him and my motivation to do a good job.
The opportunity here for apps is to build trust with their users. This involves sincerely caring about user outcomes and aligning with their goals. It also means acting in their best interest, which means sometimes acting against your own best interest. Imagine a world where ride-sharing apps suggest to a driver to stop driving. That communicates they value the driver’s interests above their own, which is a very powerful message for a driver to hear. It helps convert a transactional relationship to a longer-term one.
Conclusion
In the hyper-competitive app landscape, it’s no surprise that a flurry of business concepts dominate the discourse. Engagement. DAU/MAU ratio. Monetization strategy.
It’s time to put users back to the forefront. What was their original goal in downloading a particular app, and how can we help them follow through with achieving that? That is what sustainable engagement is about, and that is what should ultimately drive your business outcomes.
Most app developers have positive intentions. They want to create products that help people. They want to create products that help people meditate, exercise more, sleep better, save more, and be more productive. There’s an opportunity to connect more deeply to this goal of helping users close their Action-Intention Gap. And the good news is: behavioral science can help.
What do you think?
Do you have thoughts on sustainable engagement? Let us know in the comments below or tweet using #AskPlayDev and we’ll reply from @GooglePlayDev, where we regularly share news and tips on how to be successful on Google Play.
Also don’t forget to check the previous articles in this series: