If you’re building and maintaining more than a couple of mobile apps, you know the new app lifecycle presents a kind of hellish arithmetic. It goes something like this:
(# mobile OSes supported) x (total # apps) x (releases per year) = experiences that had better amaze end users
In other words, mobile is driving expectations for superlative app experiences, even as it increases both the number of device+OS combinations IT must build for, and the rate of release.1
Now, plainly cross-OS development and testing solutions such as those included in our Platform help quite a bit. But mobile also requires a shift in how we think about the lifecycle. Agile taught us that short, frequent release cycles are better than long, artifact-heavy ones. With the rise of mobile, organizations have to go a step further, focusing on the principle of minimum viable product (MVP) release versions.
The essential idea of MVP is to put something in the hands of users, quickly, and then – just as quickly – improve it based on their actual use. As Eric Reis puts it, MVP allows teams to gather the maximum amount of validated learning about customers with the least effort.
Ah, but there’s the rub: MVP depends on knowing, quickly, how users are actually using an app.
In our experience, the vast majority of companies have little means of evaluating this. Most are content simply to wait for user feedback. There are a couple of problems with this approach. First, it tends to skew to polarities, capturing the sentiments only of the most thrilled or disenchanted users. Second, it’s a lagging indicator, leaving companies little time to correct problems before users move on.
Earlier this month I took on this problem as part of my track session at ALM Forum. The key is having real-time analytics that show you, in an instant, the how/what/when/where of app adoption and usage. Remember too that with mobile comes a wealth of information unlike anything we had in the days of web. This includes variables like user location, their device type, app version, OS version and device orientation, just to name a few. The trick is seeing it, making quick sense of it, and using it to inform your next release.
As you’ll see in my deck, we’ve seen customers embrace mobile analytics and experience a hockey stick effect in their app adoption.
So what are the mobile metrics that are so critical to fast, effective app lifecycles? Nolan (our CTO and co-founder) previously shared the five mobile metrics that no company should be without. As a refresher, here they are again:
What you measure: App Installs + Opt-ins
What this answers: How many new users did I acquire? How many devices are receiving push notifications?
Why it matters: Acquisition is the beginning of the journey with your users. Knowing how people download and install your app after a given call to action can help you improve your marketing over time and reach more people. Push notifications are also a key way to engage with users in real time, so getting permission to send them is an important part of the app install process. It’s here that you must make clear your value proposition for the user, explaining why it is worth agreeing to the interruption of a push notification in exchange for the timely and valuable information or insight you can provide.
What you measure: Number of Sessions + Session Length
What this answers: How many times did users open my app? How long do users spend in my app?
Why it matters: Engagement is everything. It doesn’t matter how many downloads you have if no one is actually opening and engaging with the product. By measuring the average session length, you can determine if users are successfully using the app for its intended purpose. If the majority of users are dropping out before they have time to complete a full activity, then there are clearly pain points that need to be addressed by the development team.
What you measure: Number of Active Users/Total Installs
What this answers: How many unique users per day or week are using my app? How many of last week’s users came back this week?
Why it matters: Retention helps expand on the story that acquisition is telling. Total downloads only tells one piece of your app’s story. Retention allows you to measure changes week over week and month over month. A high percentage of your total users staying active is a sign that the product is engaging and performing well.
What you measure: User exit points
What this answers: How many users who begin a process actually get through to completion? When and where do they drop out?
Why it matters: If users aren’t completing your intended flow, then your application isn’t serving it’s intended purpose. Using funnel analysis to understand where and when users are leaving is arguably the most important metric you can track. Insights collected on conversion need to be acted on immediately, otherwise you are missing opportunities to derive value from your users.
What you measure: Ratio of app crashes to app sessions
What this answers: How common are app crashes? Where do users encounter crashes? How many crashes do I have to fix?
Why it matters: The quick-trigger attitude of mobile users means unstable apps die a quick death. Apps that crash over and over again are going to be deleted and forgotten about. You need diagnostics to understand where, when and why crashes happen. As a bar, we see world class mobile organizations achieve crash:session ratios of 1:100.
As Jeffrey Hammond at Forrester Research notes, “without an effective measurement capability, it really doesn’t matter how fast you can release – you’re still shooting into the darkness.” 2 But don’t be daunted feeling you’ve got to have every single one of these immediately. The important thing is to start – now.
And remember, we can help.
1 In honor of the recently passed U.S. Tax Day, we refer to this scaling problem as the “mobility tax.”
2 Hammond, Jeffrey. “Measuring Mobile Apps.” Forrester. December 2013.