App Indexing, Predictive Services, And Unlocking Mobile Distribution

Editor’s Note: Semil Shah works on product for Swell, is a TechCrunch columnist, and an investor. He blogs at Haywire, and you can follow him on Twitter at @semil.

There is a “perfect storm” brewing in consumer mobile: Developers, companies, and investors see the explosive growth of smartphones (with no sign of slowing down), yet consumers only have so much bandwidth to interact with a small set of apps, let alone enough time in the day for another app. Consumer eyeballs are fixated on smartphones, triggering once-in-a-lifetime opportunities for application creators to reinvent products, interactions, and industries, but tragically, limited means of getting their creations discovered, or reengaged with, or paid through them. The result, for the time being, is driving app installs and engagement is all the rage, as companies frantically line Facebook’s pockets to help drive downloads and retention of their mobile apps while a bustling ecosystem of third-party app analytic providers wait to scoop up the remains. Something has to give, right?

In the past few weeks, we have begun to see the inklings of what the future of mobile search, navigation, and app discovery may hold. Forget about Siri for now, as it actually took a step backward in iOS7, if that was even possible. On Android, apps like Cover, which contextually places apps in your lockscreen based on when it knows you’re likely to use those apps, and Aviate, which intelligently surfaces information to your phone at the right time, recently launched at a time when iPhone fragmentation is starting to pick up and when Android handsets in the U.S. are getting better and better. Earlier this year, Google brought it’s mobile anticipatory compute engine to the iOS platform, giving iPhone users the chance to see how always-on integrated Google services can work at the application layer, though the battery costs from background processing impose hefty power costs.

All of this begs a high-order question: How will consumer interact with their phones in the future? Will it be through today’s “hunting and pecking” of apps in silos with a mix of a suboptimal mobile web interface? Or, will mobile operating systems learn our behaviors so well as to predict and anticipate what we will want to do or know next, either by the time of day, the way in which we hold our phone, or other signals? Or, will we continue to search for information on our phones as we search for information on the web with Google, by inputting keywords and having the ability to search across our apps (even the ones buried in the back pages or in folders)?

This last question became more interesting this week when Android announced in its latest KitKat 4.4 update that it would enable App Indexing across apps through deep-linking. The interwebs were abuzz with the possibilities this would open up for app discovery as well as re-engagement, helping extend Google’s core competency of indexing and ranking information to include applications, which to date remain in their own silos. Very soon, on Android (and not iOS), users will be able to search across their devices as well as have Google Now push information to them – it remains to be seen if Apple can or will want to move in this direction at an OS-level, or leave everything federated to the app layer. On iOS, there’s a small, early-stage startup based out of Palo Alto called Relcy which is trying to provide an app to let users search their phone, including other apps, on Apple’s mobile platform. And on both coasts, entrepreneurs have not forgotten about the mobile web, with startups like famo.us, Wildcard, and Instart Logic trying to reinvent what can be done with the content in the browser within a mobile context.

All of these advancements come down to how we search for information on our phones, how we can and will discover new applications, and how we can and will re-engage with those services through a mix of user-initiated search and machine-anticipated prediction. While voice command interfaces for mobile seem like a pipe dream (though, hey, it could happen eventually), bringing standard search back to our phones and becoming empowered to find information within disparate app silos could theoretically unlock a significant amount of utility and save time. For developers, of course, it could help reduce the pain surrounding two harsh realities – getting new people to discover your app and, once they’ve downloaded it, getting them to engage again (and again) with the software. Will these advancements in Android unlock distribution for mobile developers and be the push they need to leave iOS? Whatever does happen, the mobile platform that can help with app distribution – whether through user intent and search, or through predictive services, will attract developers in drove.