A few weeks ago, TechCrunch ran an excellent post called “The Future of Apps Should Be Better Apps.” In it, Anshu Sharma responds to the Wall Street Journal’s pitch that chatbots are going to replace apps with a resounding “No, not that easily.” Sharma makes a solid point. It could be that we end up chatting with our phones more than tapping them. It could be that we transition to the mobile web, and we might even end up somewhere in between web and native apps. But it won’t be the most hyped up or flashiest future that wins the day. It’ll be the solution that does the best job of solving the complex issues that apps have. That’s not going to happen overnight. It’s going to happen as the result of small discoveries and a gradual movement towards more user-friendly computing. We may be agnostic on “what comes next.” What we do know is that analytics is going to pave the way. [Tweet “#Analytics is going to pave the way to better #apps”]
99 problems
In his TechCrunch article, Sharma mentions a few problems with apps that he thinks developers will have to solve in upcoming years:
• Downloading and installing apps is annoying
What: You go to www.acme.com on your phone only to get a big, aggravating pop-up telling you to install the Acme iOS app.
****• Signing up for apps is cumbersome
What: You have to create an account with a new password just for the Acme app, adding to your cognitive load and increasing the chances that you’ll forget these credentials later.
****• Hard to find apps again
**What: **Later, when you need the Acme app again, you can’t find it in your phone or remember your credentials.
These are issues that are more about usability design than they are about platform or form factor. It’s not difficult to imagine a chatbot being equally as annoying to set up and manage, especially if they start to become mini app ecosystems unto themselves. And the mobile web, while it no doubt holds promise, was abandoned for apps precisely because it had similar problems—it was cumbersome and disconnected. These are some of the most intuitive problems that apps have today. There are others that are trickier to see for consumer and developer alike.
The thorny ones
Here are some common app problems that we think a lot about at Amplitude:
******• **Cross-platform cohesion
**What: **Apps don’t always make it convenient to switch from mobile to desktop to tablet while maintaining a consistent experience across all platforms. This is made especially difficult by the fact that completely separate apps have to be written for Android and iOS.
******• **Features that don’t add value
What: Too many apps are loaded down with features that users don’t care about, hiding what could be decent functionalities behind them. “Just because you can doesn’t mean you should” is a lesson that often gets lost because people want to give users everything they could ever want.
**_****• **Excessive _optimization of metrics at the expense of UX
What: It’s blatantly obvious when companies that say they’re optimizing for “engagement” are really just optimizing for short-term returns. They put users through unnecessary hurdles, blast them with intrusive ads and ping them constantly, while justifying it all with a small boost in conversion percentage or website traffic. In the long-term, this kind of behavior just drives users away.
These are tough ones because they’re really hard to see. Unless you have a deep understanding of user behavior, you’re not going to be able to identify the features that are actually driving users away from your app. Even worse, you might not identify your “secret sauce” feature that could be the foundation of incredible growth. [Tweet “Without analytics, you won’t find the feature that could be the foundation of #growth”] We thought a lot about these kinds of thorny problems in developing Amplitude, and we write about them now because we want to help all developers make better apps. We think analytics is the best way to do that.
How mobile analytics changes the game
When Instagram was first getting its start, it was just a Foursquare clone with a bad case of feature creep and a cute name—Burbn. A lot of people thought Burbn was the hot new social location app. They had $500,000 in seed funding to work with and the benefit of being in a great space. But co-founders Kevin Systrom and Mike Krieger knew that something was wrong with their big idea. Most of their features were going totally unused. After carefully studying their analytics and looking at how users were behaving inside their app, they cut away all the features until there was only one left. It was the camera. Much to Krieger and Systrom’s surprise, people really enjoyed taking photos, putting filters on them, and sharing them with their friends. The rest is history. Paul Graham, in Hackers & Painters, put it best: “Software should do what users think it will. But you can’t have any idea what users will be thinking, believe me, until you watch them.” We tend to assume billion-dollar ideas come out of nowhere. But it wasn’t magic that led to Instagram’s creation. It was a meaningful and thoughtful analysis of user behavior that led Krieger and Systrom to their good idea. Without user data, they would have kept building Burbn. With it, they started building Instagram.
Keep it simple, stupid
When we spend too much time thinking about the next big thing, we succumb to magical thinking. We imagine that the next big thing is just going to pop spontaneously out of some 22-year old Stanford grad’s brain, and that after that happens we’ll ditch apps forever for (chatbots/voice AI/mobile web/whatever). [Tweet “When we spend too much time thinking about the next big thing, we succumb to magical thinking. “] Those kinds of stories can energize and excite us, and of course there’s some counter-intuitive element of surprise in every great app. No one would bother with apps unless there was. But intuition alone isn’t what builds great products. It’s collecting data. It’s talking to your users. It’s using that data to make decisions. It’s hustle—pure, unadulterated, quantitatively-driven hustle.
