Getting Started: Digging into Drop-Off with User Behavior

Say you run a language learning app, and you notice that users have started to drop off after creating an account. That’s no good—you need your users to actually use your app. What happened? And more importantly, how do you fix it?

You may have some hunches, or maybe rolling back the latest updates could do the trick. But to be honest, you know you can’t build a great product on vibes alone. What you need are actual answers to your questions about user behavior—and you can get that with Amplitude.

In this post, I’ll show you some ways you can use Amplitude to quickly and deeply understand how customers use your product, and how to turn your behavioral insights into solutions even before drop-offs happen.

Read on for a practical guide to fixing drop-offs in onboarding, feature adoption, and more.

Tracking user journeys

You know users are creating accounts. At some point after that, you know they stop using your app. What are they doing in between? In order to understand why drop-off is happening, you have to know where it’s happening.

The simplest way to map your user journeys in Amplitude is with Journey charts. These help you see all the steps your users actually take as they work through your activation flow, whether they’re following the intended path or, to use a technical term, “wilding out.”

• To get started, give your Journey chart a start event and an end event. Amplitude takes care of identifying the all events triggered in between.
• Use a Pathfinder chart for a high level exploration. This is a great chart to start with if you don’t necessarily know what behaviors you’re looking for.
• Dive deeper with a Journey Map to compare customer paths, analyze them by frequency, and see path branches at a more detailed level.

In your language app, maybe you find the most common journey is account creation > onboarding tour > language placement test. Let’s also say that that’s where most of your drop-off is happening. Time to dig in more.

Digging into drop-off points

Once you have an idea of where drop-off occurs, you can start to unpack what’s causing the it at those points. Are there certain steps with more drop-off? Are there common characteristics about the users who drop off?

An easy way to do this in Amplitude is with Funnel charts. These let you examine specific events in order to see how many of the users that triggered the first event (say, creating an account) went on to the next event (onboarding tour), and the next (placement test), and so on.

• The funnel shows you how many users completed each event vs. didn’t, and you can start seeing trends with just two events. Maybe you find that 75% of your users make it through the onboarding tour, but only 20% finish the language placement test—that means the language test is your priority.
• When did the drop-off start? Change your date range to see how funnel behavior has changed over time. Maybe your drop-off coincides with a recent update?
• Who’s making it through the funnel and who isn’t? Click on the sections of each funnel step to find out more about the user cohorts. There could be common traits for users making it through.
• Keep a proactive eye on common funnels by adding them to dashboards or setting up alerts to notify you if there’s a change in conversion.

To start coming up with fix ideas, you can use the funnel chart to watch session replays of users who dropped off and see their frustration moments firsthand. You can also use integrations to push cohort information to marketing platforms like Braze for re-engagement campaigns.

Spotting power-user pathways

Amplitude gives you the tools to fix drop-off when it happens—but what’s even better is stopping drop-off before it starts. You can do that by using Amplitude to identify high-value user behaviors, then build ways to funnel more new users to those high-value behaviors.

• To make a cohort of high-value users, you can start with a retention chart for users who stick around the longest. Then dig in to see if there are common ways they user your app. Maybe they all take their first lesson in the first 24 hours—that’s a behavior to encourage for other new users.
• Next, create a guide to do that encouragement. Design the look and feel yourself, set up custom triggers so it only shows up for new users, and roll it out without any engineering resources. Just like that, you’re directing your users to their first lesson instead of hoping they’ll find it.

For more creative ways to activate power user behavior, check out the cohort tutorial that Anas Farah, Lifecycle Data Scientist at Square, recently ran for our Amplitude community.

Want to understand more?

With just a few Amplitude charts, you’re now well on your way to fixing your language app’s drop-off problem—and even finding ways to boost activation higher. When you actually understand how your customers engage with your app, you can build better products and experiences easily.

These tips are just the first couple of steps to get started on your user behavior journey. Visit our Getting Started hub for Understanding User Behavior to find more techniques to try. And if you haven’t already, sign up for Amplitude for free and test these out yourself.