With its upcoming Android M operating system, Google revealed a major UI breakthrough for smartphones. It’s called Now on Tap. An extension of Google Now, you simply tap and hold the Home button inside any Android app, and a card appears on the bottom of your screen suggesting a number of apps that can help carry out the next thing Google thinks you want to do on your phone.
So say you’re texting a friend about seeing The Avengers 2. Tap home, and a card with the movie will pop up, with links to reviews on IMDb, trailers on YouTube, and a way to purchase tickets like the Fandango app. Google has predicted a small handful of places you might want to go, and offered instantaneous bridges to get there.
Maybe this seems like a minor update, but in a broader sense, Now on Tap is attempting to solve the biggest problem on smartphones right now. Our million+ apps, once so useful, have trapped us inside a million+ boxes, each with very limited functions. So you’re messaging. Or you’re yelping. Or you’re googling. When you’re inside one app, how do you get to another app? If you're in Google maps and you've found the movie theater you're looking for, how do check what movies are playing and book a ticket and check out restaurants in the area? It's a nightmare of swiping back and forth between apps, while all the while trying to remember what you were trying to do in the first place. It's annoying.
So far, designers have already responded to this problem in two ways: First, the mega popular messaging apps like WhatsApp have begun to absorb more and more functions. They’ve become alpha apps, if you will, containing their own little cities of talking, searching, shopping, and mapping in a tightly integrated bundle. But it's hard to imagine this scaling very well, as the world of apps and services only continues to swell.
Meanwhile, Facebook, Google, and to a lesser extent, Apple, have to implemented a technology called "deep linking" or "app links," which allow you to tap on a link and go straight into the relevant section of an app. The philosophy behind app links is that you’ll surf your phone much like you surf the web. But app links come with a major limitation: A developer has to program an app with your destination already in mind. It has to foresee the future for all sorts of different people using its app, and play to the lowest common denominator. Sure, it makes sense that Google now links to Uber directly from Google Maps. But what about the dozens of other things you might want to do directly from Google Maps, such as make a date with a friend or make a reservation at an restaurant? You can't very well add a half-dozen links to any already crowded UI. This too is a system that can't scale with the ever-expanding universe on our phones.
Enter Google with Now on Tap. It combines the convenient interaction of app links with the remarkably good artificial intelligence of Google Now—that supermind that can already suggest a route to a nearby Wendy’s because it knows you’re hungry and knows you love Frosties, or pull up your flight reservation because it spied your Travelocity reservation in Gmail.
Now on Tap relies on an even larger universe of information, in the form of an even powerful version of Google Now. It used to be that Google Now could only mine information from other Gmail services, but Google has now opened up the Now API, so that Google Now can parse data from over 100 other apps, mining for actionable connections. Thus, much like Google can predict your search before you finish typing it, Now on Tap will predict your next multitask before you multitask it. And in doing so, it condenses anywhere you want to go on your phone next to the simplest of UI elements: A single button. One press could bring up any one of dozens of services that you might need in the moment. Any app that's been linked up to Google Now is fair game.
One day, perhaps it might help you discover services you aren't using but seem to need—thus solving the problem of app discovery, which has proven to be a severe bottleneck on the number of apps people use. Facebook already makes a ton of money advertising apps; imagine what Google could do, if its app advertisements were both highly contextual and immediately useful. The point is, whatever Google does do, it has quickly introduced what could become a powerful new paradigm in mobile computing—one that ties to its overriding goal of reducing user friction by answering questions before they've even asked.
But the big caveat is that Now on Tap will only become second-nature for users if it consistently spits out something useful. The service has to reliably steer people to the right place—otherwise, trust in the recommendations will wane and the service will become an afterthought. If Google fails in this endeavor, if it points you to a movie when you planned to see a play, or a fast food restaurant when you’re on a diet, then Now on Tap is no longer a trusty, personal assistant. It’s a dumb, nosy computer that’s getting in the way.
To be fair, Google Now has already been pretty successful at fortune-telling your needs. For many, it has become Android's killer app. But the complexity of mining information across just a few Google Apps is a far cry from trying to sort through dozens of competing apps, each of which might be the one you need. How often can Google be wrong before you decide that your own app-hopping is faster? I don’t know the answer to that. But soon enough, Google sure will.
Co-written by Cliff Kuang.