Google announced yesterday something it’s been rumored to be working on for a long time: their first “streaming” apps, available as part of the Google search apps on Android devices. This is just one of a number of changes that might come to the way mobile apps work in the coming years. But one of the big questions is to what extent these changes will be driven by trying to better meet consumers’ needs and to what extent they will reflect companies’ attempts to pursue their own strategic objectives.
Google and app streaming
Google’s new model is one I’ve been interested in for quite a while. I did a consulting call with one of the largest tech companies in the world about a year ago and this was one of the main things we talked about. It feels like it is the single biggest disruption that could come to the way apps work today. Google is the most advanced in rolling it out, though Microsoft appears to be working on something similar. And there’s a reason why Google would want to pioneer this model: it fits well with Google’s overall strategic objective of reasserting the web over the app as the fundamental model for interacting with content on mobile devices. The reason? Simple: Google makes money primarily from web search and, for all its efforts to provide search results from within apps (100 billion in-app links so far), this still isn’t a perfect solution.
The question then becomes, does this model also serve consumers well? Or is this like so many efforts, from Google and other companies, primarily aimed at serving internal objectives and counter to what consumers want? What is clear is apps have become the dominant medium people use on their phones, with the browser and the open web used far less in most cases. However, the problem with apps as they work today is they’re fairly binary in nature: either you have them installed, in which case all their functionality is available (including in-app search), or you don’t, in which case they might as well not exist. The best most apps can do in a search scenario is pay to place app-install ads in relevant search results. In fact, this is exactly what Hotel Tonight – one of Google’s launch partners – does currently, but it makes for a pretty involved flow, which leaves you right back where you started, without any of the context of the original search:
The new app-streaming model solves two problems for consumers: it includes content from within apps the user hasn’t installed and which would otherwise be absent, and it allows the user to access the content within the app in a familiar app-like context, without having to permanently install the app. You go from a five step flow that leaves you having to do your search all over again to a single step flow that retains context and leaves your device in the same state as when you started. For at least some scenarios, that’s actually very consumer-friendly. I suspect that half the reason we all have so many unused apps on our phone home screens is we install apps for single uses that we have no intention of ever using again and just never bother to uninstall them. App streaming mitigates this significantly, while giving you the option of installing the app if you like it enough to want to give it a permanent home on your device.
The downsides of app streaming
However, there are some significant downsides to the streaming model too. Today, Google’s app streaming model only works on very strong WiFi, only with a handful of apps (which have to be participating in Google’s app indexing API), and only for apps which don’t have web equivalents, including Hotel Tonight, which only exists as a mobile app. So the model is inherently limited, not least by having to have good connectivity. One of the major benefits of many apps is they continue to function fine even without good (or any) connectivity, since much of the content is stored locally. Streaming apps can, by definition, only work when you’re connected. And when you are connected, they’ll typically use quite a bit more bandwidth each time you fire up the streaming version of the app than if you had it already installed on your phone, since all the UI and other assets have to be re-streamed every time you use it. In markets with poor and/or expensive connectivity, this would be a very sub-optimal model.
Another interesting wrinkle is how business models would evolve to deal with app streaming. A single price for a one-off download of an app is utterly straightforward, but how would you price the same app in a streaming model? It obviously works best for apps that monetize in ways other than app purchases (which, again, makes Hotel Tonight a good launch partner), but it’s not clear how it would work for others. Perhaps apps might offer short free trials through a streaming model, with the option to pay to download and unlock additional features.
Other innovations in apps
Of course, app streaming isn’t the only possible innovation in applications. I talked a few weeks ago about Apple’s virtually unbreakable model of having every app a user installs represented by an icon on the home screen and how that seems an increasingly poor approach. As such, I think what I call “headless” apps might well be another innovation we’ll see soon – apps can be installed but not take up space on the home screen. More broadly, notifications and especially actionable notifications will continue to grow as a way to interact with apps, perhaps in tandem with the headless model. There are so many apps which don’t really ever need to be opened, because all we use them for is notifications and quick actions in response to them. Interactions between applications will also continue to evolve – Apple introduced the Extensions concept relatively recently, but other platforms have had equivalent functionality for quite some time. Yet these interactions are often still fairly rudimentary and the depth and sophistication of the interactions could grow quite a bit.
Then, of course, there is the whole question of a layer of interactions that take place outside of apps, through things like search and virtual assistants, baked into the operating system itself. All the major platforms now include some form of built-in search which can tap into web and in-app content along with other data sources without launching apps explicitly. And Google is working on making its Google Now assistant context-sensitive with its On Tap features that take account of the app someone is currently using or the content they’re consuming.
The other thing that needs to continue to evolve – which I alluded to briefly above in the context of app streaming – is app business models. I’ve written previously about the challenges associated with in-app purchases in particular, but app stores need to evolve in their support for all kinds of features people are already familiar with when they buy software and services in other ways. Paid upgrades, trials, bulk discounts, and many other features need to be supported if developers are to be able to build sustainable businesses off the back of their apps.
Consumers have to be the drivers
I return to a point I made earlier – Google’s motivations behind app streaming are clearly driven, in large part, by its strategic imperative to feed the web rather than native apps. But that doesn’t mean it’s inherently consumer-unfriendly. But, as app models (and business models) continue to evolve, all the companies involved need to ensure they’re making changes designed first and foremost to benefit their customers. If companies instead feed only their internal needs, they’ll find themselves going down paths consumers won’t follow.