What do IKEA, tea kettles, drugs, and your app all have in common? A first time user experience. Whether or not you’ve designed one, all products have a first time user experience, and a carefully crafted on-boarding process significantly improves user engagement. In this post, we’ll take a look at how we were able to give 20% more free internet to new users by adding a tutorial to our FTUE (First Time User Experience). As a quick side note, we cover the tutorial here, but the FTUE subject covers a wide range of app features. Things like app load time, registration funnel, and apk size all affect your user’s first time experience.
The idea of adding a tutorial surfaced because we saw that a large % of users that registered never tried an app on our platform. We’re an app that pays for mobile phone plans in developing countries in return for users trying apps, so that metric was obviously concerning to us. First time users were not understanding the fundamental benefit of mCent.
Taking a cue from some of Google’s apps, we initially built a showcase view tutorial, which highlighted different sections of the app and described what each section did. Our goal was to have a brief explanation of all the different parts of our app. Once we released it into the wild, we saw an immediate impact on our users. The tutorial increased the number of app trials per new user by 10%, and it was simple to see how the tutorial helped increase engagement and revenue. However, over time it became obvious this tutorial was unscalable, because the length of it was increasing with every added feature.
It started first when Jameson, one of our data scientists, suggested adding a slide for referring your friends. You can earn tons of free internet with referrals, so why naturally we should be showing this in the tutorial? We added one slide then. Shortly after, we launched two new big features, the activity feed and our messaging platform, which meant adding another two slides to the tutorial. Our tutorial was now 7 slides of scrolling, before you could start using the application. It was a really poor way of introducing users to mCent considering the hoops they jumped through to find, install, and register. With that in mind we went back to the drawing board.
We built the second iteration of our tutorial based on 3 learnings and assumptions:
- The magic moment of our app, regardless of any new features, is when users complete their first app trial
- As our app increases in complexity, it will become impossible to tell users about every new feature that we launch
- In our core markets, reading through verbose slides of information, especially when english isn’t their first language, massively increases the friction of completing a tutorial.
We decided to focus only on the core experience of completing offers. We dropped the showcase view, and instead focused on getting users to try their first app. We simplified the text so that there was less reading to do. We made it so that you could go through the tutorial and understand the app, even if you didn’t read it. The improvements we made by streamlining the tutorial and focusing it around our core loop were clear. The number of completes per new user increased by another 10%, meaning that between both iterations, we had increased the number of completes by 20%. While this number may seem small, when you multiply 20% by every single new user that comes in through the funnel, it meant significant overall impacts to our system.
In the end, that meant that more users were able to enjoy free internet access, bringing us closer to providing free internet access for a billion people in developing countries. Pretty sweet stuff. If you’d like to learn more, come work with us!
- Having a good FTUE can make up for subpar quality at the top of the funnel.
- Create a tutorial that is action driven, rather than descriptive.
- Simplify rather than trying to fit the kitchen sink into the FTUE.
- Constantly challenge your assumptions. FTUE was previously dismissed as something unneeded
- The tutorial should impart understanding even if users don’t read any of the text. Minimize walls of text.
- Don’t expect users to make it to the end of the tutorial. Many will drop out, and the experience should be strong even for those who did.