Platforms All the Way Down

More people know how to use Facebook, which was started in 2004, than natively speak any single language; the particular syntax of Google’s email service, which soft-opened the same year, is familiar to about three times as many people as live in the United States; Pinterest won’t say how many people use its app, but those people have hit the “Pin” button at least fifty billion times. These platforms, which are conduits for human attention, have become staggeringly large.

A single person likely participates in sustaining many of these companies at once, which can make it difficult to understand how they relate to one another. You receive Gmail messages as you check Facebook; you get buzzes from Instagram as you scroll through Twitter. One moment you’re deep within an app platform reading news, two seconds later you’re sending private messages through another app that also hosts news. Media from one service appears in feeds elsewhere; the tranches of your identity represented by each swell and shrink constantly. This takes place, increasingly if not mostly, on your phone, where notifications pull you from place to place transparently and quickly enough that the process doesn’t feel like an intentional one. The phone, like the computer or the browser before it, seems neutral.

But of course it isn’t: Every path a user takes through the software on an iPhone, or an Android phone, is intentionally designed. iOS is an expression of Apple’s ambitions just as the News Feed is an expression of Facebook’s. Today’s iPhone is the product of years of refinement; its software, iOS, has been significantly altered over the last eight years, but has been refined largely in the service of one main function: running apps. Apps in a grid. Apps that open and close. Apps that were created first to access the large number of people using smartphones; apps that helped make smartphones more vital, turning that large number into an enormous one, and altering their platforms’ characters in the process. The iPhone, and Android phones, are, for different people, ways to check Facebook, or to use Snapchat, or to play games, or to take pictures, or to text on one of a dozen functionally similar services. But the hierarchy never changed. The platforms that host our conversations and our media and our social performances still answer to platforms of their own.

In recent years, Facebook’s News Feed created an enormous opportunity for the publishing business. Facebook, the place where people went to see what their friends were up to, became a central destination. News Feed collected and spread links and comments and videos, creating a virtuous cycle of attention — to organizations posting from the outside, its growth and ability to drive attention suggested endless demand. This culminated in something like industry dominance; news organizations and entertainment sites and video producers, some more intentionally than others, found that, even on the internet, the best way to reach the most people was still through a platform that isn’t their own. For Facebook, publishers filled a void in the service, supplying near-endless matter for conversation, reference and consumption within the feed; for publishers, Facebook provided growth.

Facebook’s influence on what publishers produce is obvious because publishers produce media: They wear their sensibilities on their sleeves, and express them in familiar language, so it’s easy to watch them change. Just as Twitter’s brand of competitive conversation breeds and favors certain behaviors, Facebook’s metrics-sorted News Feed breeds and favors its own. Media must account for its container, and a platform is an opportunity for publishers only insofar as they are willing to produce media in the platform’s native languages and forms. These languages and forms change as the platform does; the platform changes as its owner sees fit. Facebook as a video host, or as a host for articles, strains the definition of partnership. How can you partner with the company that provides the entire context for your existence? This arrangement likewise strains the definition of a publication, which — having lost some degree of self-determination — is reduced to the sum of its content.

What’s easy to miss from within Facebook’s platform, where its relative size implies total self-determination — is that it is similarly dependent on two others: Android and iOS. Facebook reported 3.54 billion dollars in revenue last quarter, seventy-three percent of which came from smartphones. At the same time last year, twenty-four percent of its monthly visited on mobile; now, that number exceeds eighty-five percent. This is staggering but contingent growth: It is billions of dollars gathered from advertising placed in apps used on platforms designed and ultimately controlled by Apple and Google. Social platforms’ businesses are contingent on the mobile platforms alongside which they have taken over the world.

This situation is not lost on Facebook, which is large enough to characterize as a competitor to Apple and Google. The company attempted to make its own phone, which failed. It has been more successful in breaking its app apart into pieces, isolating messaging from the News Feed while seeing growth in both. It has acquired other, simpler services — like Instagram and WhatsApp — and it has taken steps to survive in a world in which the app grid is no longer the first thing people see when they check their devices, which, it turns out, may not even be phones. (Facebook is attempting to buy its way out in front of that, too.)

Recent announcements regarding the parent platforms have given this task a sense of urgency. Google described a feature in the next version of Android that reaches within apps to provide contextual information to others:

If you and a friend are discussing a movie, just long-press the home button to pull up a card with information about the movie, like showtimes and ratings. If you’re watching a YouTube video of a famous actor, just call up Google Now to get more information about that person.

Foursquare, another service the company included in its presentation, would seem to function not as an app with its own interface and friend connections and advertisements, but as a piece of infrastructure in a Google service. This feature, an extension of Google Now, a fuller attempt to reconstruct Google’s smartphone software around “cards,” which glean recommendations, predictions and updates from, among other sources, your email, searches, and maps. It recommends articles not in a feed housed within an app but on the home screen, based on sites you’ve read and what else their readers read. I’m not sure I’d call the recommendations Now has given me good, exactly, but they are extraordinarily specific and often unexpected — they project, above all, the presence of lots and lots of data.

Apple, yesterday, showed off something similar — a change to iOS that treats apps less as silos than as transparent, accessible entities:

Proactive assistance presents the most relevant information without compromising users’ privacy and suggests actions at a particular moment — even before you start typing — automatically suggesting apps to launch or people to contact based on usage patterns, and notifying you when you need to leave for appointments, taking into account traffic conditions. iOS 9 can even learn what you typically listen to in a certain location or at a particular time of day, so when you plug in headphones at the gym or hop in the car before work, it can automatically display playback controls for your preferred app. Typed search queries deliver more relevant results from more categories, including sports scores and schedules, videos and simple math calculations.

Both moves make sense: They give these companies, who are not shy about reminding the public that they have presided over the creation of vast new industries, more control over what has grown, they feel, on their turf. Such changes might also fundamentally change what an app is, and how it should function: An operating system in which an app’s utility is extracted from its interface and integrated into the main software is an operating system that regards apps in a meaningfully different way than the iOS or Android of yesterday. Just as a publisher’s proposition to an advertiser is undermined by Facebook’s ability to advertise further up the referral chain, an app — even an app like Facebook! — might find business models reliant on platforms within platforms less appealing. (You could argue that in-app interfaces have already been demoted by notifications, which provide direct access not just to apps but to specific parts and actions of apps. A simple messaging app exists substantially within notifications; many in the first wave of Apple Watch apps are, essentially, notification filters.)

This is the proper context to understand Apple’s music services and News app, which, not unlike Facebook’s Instant articles, will host publications’ content natively. (It will also allow publications to sell their own advertising, or supply ads for a cut of revenue.) Whether or not Apple’s News app will succeed is as much a question about how Apple’s platform will change as it is about the specifics of the first version of the app: Do people want a Flipboard-style panel of nicely formatted news stories with no social context? And if they do, won’t publishers face the same identity obliteration they found on News Feed, where publications’ names and reputations have been relegated to captions between panels of official platform content? Or is finding a foothold on Apple’s platform about making sure you’re at least present when media is proactively sent to users according to the judgment of software? Apple’s past attempts at bigfooting their developers haven’t always worked, but they have also been relatively constrained — a News app that’s just one more app in the grid doesn’t seem to change much on its own. But, again, Apple provides and exerts influence over that entire context.

In 2011, Apple put news apps behind a single icon; this week, it brought them out. For years, its App Store recommended games to users based on a simple algorithmic determination of “What’s Hot.” Now, in a change that might be small and obvious for Apple, but which will significantly alter the way in which developers make and market games, its recommendations are made by people. Who knows where News ends up in a couple software updates! Or where, in some new and obvious software scheme, a standalone Facebook app exists. Or how deeply Apple’s music service becomes embedded into its phones, tablets and watches, and what remains available to Spotify. What is an App Store, exactly, when your phone suggests new apps based on things you do or say, or, as they already do, based on locations you visit?

It could be the case that Apple and Google’s attempts to re-centralize the various functions of their platforms under tighter control will fail. Maybe, in making their own platforms more assertive, they create a new class of arbitrary interfaces that competitors — on their platform or on new ones — can improve upon. Maybe the basic functions of a smartphone operating system become easy enough to replicate that an increasingly territorial Apple or Google could lose its claim on the category, just as a social platform that annexes and anoints too many of its partners might risk drowning out the simple but compelling human network that drew its audience in the first place.

The melding experiences of living and working on a queasily centralizing internet will be defined by these relationships. Platform conflict, private or public, sensible or capricious, explicit or implied, will determine the manners in which we read and watch and communicate and produce online. A type of co-dependency that would be familiar in any advanced industry becomes incomprehensible at such scale and seemingly boundless acceleration — a billion users as unsure about what they’re seeing as where they’re seeing it from.

Anyway! This is the internet we have, at least for the time being: the internet of platforms, all the way down.