Apple Can Build the iPhone. Why Can't It Get Internet Services Right?

In the end, does Apple want to be a hardware company or an internet services company?

Editor's note: Once a market changes, how the current leader finds its future is a permanent question. Apple is the company facing it now: it can no longer sell more iPhones every year.

We can watch Apple probing for that future in several directions at once — internet services, smartwatches, cars.

Apple is a great company. It has changed the world twice, with the personal computer and the smartphone. But the shift it has to make now has few successful precedents.

As the mobile internet finishes spreading, other leaders will run into the same problem. Apple's search will be instructive. We are following its attempts in a series of three articles. This is the first.

At every product launch, Apple's executives work a few well-rehearsed jokes into their presentations to warm up the room.

Yesterday's WWDC 2016 was no exception. Demonstrating a new iMessage feature, Craig Federighi, Apple's vice president of software engineering, told the more than 5,000 people in the hall that "it's not polite to order food just for yourself," and used the chat window to order 5,000 Indian curry puffs.

When the price, $43,177.43, appeared on the giant screen behind him, the audience laughed.

Federighi was replaying one of Steve Jobs's classic jokes. At the first iPhone launch in 2007, Jobs used the phone's built-in Maps to find a Starbucks, then tapped the screen to call and order 4,000 lattes, to go.

In nine years, Apple went from a small company that couldn't sell 2 million phones a year to the most valuable company in the world, selling 150 million iPhones in 2015.

But iMessage, announced at WWDC 2011 — the last launch event Jobs hosted himself — never joined the ranks of the world's major internet services.

The American venture capital firm KPCB published its well-known annual internet report at the end of last month. The main social and messaging apps it compared were WhatsApp, Facebook Messenger, WeChat, Instagram, Twitter and LinkedIn. iMessage was not mentioned at all.

When iMessage first appeared, more than 100 million people around the world were already using the iPhone, and all of those rivals were still small: WeChat and Facebook Messenger had just been born, and WhatsApp was one year old.

Five years later, each of those three services has more than 600 million users, more than all iPhone users put together. iMessage is still just a replacement for text messages on the iPhone.

That iMessage is only an accessory to the iPhone would not matter, if the iPhone kept selling in greater numbers and kept bringing Apple more revenue from hardware. But starting this year, Apple's future is tied to falling iPhone sales.

Internet Services Have Never Mattered This Much to Apple

More than 60% of Apple's revenue now comes from the iPhone. And iPhone sales in the first three months of this year came to only $32.86 billion, $7.4 billion less than the same period a year earlier. By Apple's own forecast, the next quarter will be worse.

Selling fewer iPhones does not mean users are switching to Android. In mature markets such as the United States, iOS's share is in fact higher than it was a year ago:

Apple's problem is that its existing users have no reason to buy a new phone.

According to Apple's own financial reports, by the end of this March only 40% of existing users had upgraded to a 6 or a 6s. Most people willing to pay iPhone prices have already bought one, and most iPhone owners wait nearly three years before replacing the phone in their hand.

In other words, the number of people using an iPhone hasn't shrunk. It has simply become very hard to find new ones, and most of those who already own one rarely switch. Until something breaks through in smartphone hardware or software, none of that is likely to change.

By outside estimates, there are about 550 million iPhone users today. If Apple can't sell more new hardware, persuading them to buy more internet services is the simplest place to look for growth.

Apple thinks so too. On the past two earnings calls, chief executive Tim Cook has pointed out that Apple's internet services revenue is still rising, and called it "the power of Apple's ecosystem."

But in the first quarter of this year, when iPhone revenue shrank by $7.4 billion, services revenue grew by only about $1 billion — nowhere near enough to cover the hole the iPhone left.

Yesterday's Keynote Shows Apple Is Taking Services More Seriously

In the 120-minute keynote yesterday, Cook and nine other Apple executives took turns on stage to announce this year's upgrades to four operating systems: the watch, the TV, macOS and iOS.

The schedule was so tight that several App Store changes had to be announced a week ahead of the keynote — the first time in more than a decade that Apple has released news before a launch event.

Even in an event that packed, internet services still got a sixth of the time. Fifteen minutes went to iMessage, with three executives on stage walking through its features, and another five minutes to the heavily redesigned Apple Music.

The upgraded iMessage can do a lot:

• Connect to third-party apps — ordering food, as Federighi demonstrated, or sending money to a friend;
• Richer animation: GIFs, stickers on photos, even full-screen effects flashier than WeChat's;
• Resize a message, such as shrinking the text to sound like a whisper when you apologize;
• Suggest, automatically, that you swap words for emoji.

WeChat has already shown the world that a messaging app connecting you to the friends and family around you equals a sizeable business. Apple is now rebuilding iMessage into a product that competes head-on with WeChat, WhatsApp and Facebook Messenger.

Earlier reports said Apple would bring iMessage to Android and make it a complete cross-platform internet service. That wasn't announced at the keynote, but the Android version of Apple Music also arrived some time after the iPhone version last year. An Android iMessage app will probably follow later this year.

In those 15 minutes yesterday, iMessage announced many times more new features than it had in the previous five years combined.

And shipping everything in one big release is a major reason Apple's internet services aren't good.

Apple Runs Its Internet Services Like a Hardware Company

Phil Schiller, the Apple senior vice president who has run the company's marketing since 1997, once compared Apple's product launches to the release of a Hollywood blockbuster.

Like the publicity blitz before a blockbuster opens, Apple saves its heaviest promotion for the short stretch just before a product goes on sale.

Yesterday's event looked much like the premiere of a big franchise film: devoted fans pitched tents and waited a day early in the cold San Francisco wind. Hours before it began, thousands of attendees filled the plaza in front of the Bill Graham Civic Auditorium, looking very much like a fan rally before the opening of Star Wars: The Force Awakens.

Apple's product development and its marketing both revolve around a few launches a year. For hardware, that is no bad thing; Apple's success over many years proves it.

But almost no successful internet company works that way. Google, Amazon and Facebook, and Tencent and Alibaba in China, none of them push product updates on a yearly cycle.

The Facebook and Uber apps for the iPhone ship a new version roughly every two weeks.

When Google, Amazon and Facebook test a new version of an online product, they first open the feature to a set share of users, then decide from how those users behave whether to give it to everyone.

At these companies, data is the basis of every decision. Douglas Bowman, Google's former visual design director, once wrote that he had no authority even to decide whether a line on a web page should be three pixels or four; it had to be A/B tested.

Apple doesn't do this. It doesn't open a feature to 10% of users and then adjust it according to the data. Apple decides internally what a service should be, finishes it, and delivers it to all users through an operating system update. (iOS does have betas, but they are aimed at people who choose to register, download and install a test build — not the general public in any real sense.)

Apple's two most important internet services today, iMessage and Apple Music, both update along with the operating system. Since iMessage launched, iOS has been upgraded on average once every six weeks, which means six weeks is the shortest interval iMessage can possibly manage — three times as long as the Facebook app's.

In practice, because many iOS upgrades are emergency patches for system problems, iMessage is updated far less often than once every six weeks.

In hardware this was not a problem. Competitors had no chance to iterate quickly either; neither BlackBerry, nor Microsoft, nor Nokia could make better calls than Apple.

But the best internet services today all iterate fast and adjust as they learn. Apple clearly knows how internet companies work. It has chosen to believe it can make the best choice on its own.

Apple's first choice isn't always right. The redesigned Apple Music is one example.

The Apple Music shown at last year's WWDC had a fairly complicated interface: dense text, an app whose type and background colors changed with the album playing, and a tab pointing to Connect — in plain terms, a place for musicians to send messages to listeners.

This year's version simplifies the interface and hides the Connect page nobody liked. Trent Reznor, who runs Apple Music, explained in an interview after the keynote that Connect was removed because the team realized "it ultimately wasn't a place people wanted to go, so we spread what used to be posted in Connect throughout the rest of the app — artist pages, the browse section, and so on."

That process took the Apple Music team a year.

This May Be a Difference in Culture

Apple's people are careful when they discuss how the company runs inside, and the best reporting on the subject amounts to fragments.

What is certain is that Jobs long ago rebuilt Apple into a company that runs around a few products a year. Inside, it is organized by function, not by product. And the top executives — Cook, Schiller, Federighi and the rest — deal directly with the details of the major products.

That structure lets Apple build the best hardware and sell it. Even with a product that is plainly not mature, like the Apple Watch, Apple can sell 10 million a year, far ahead of any comparable rival.

No other technology giant today is built like Apple. Steve Ballmer tried to reshape Microsoft along similar lines in 2013. The reorganization had run only a few months when he announced early retirement without a successor in place — which amounted to being pushed out.

With a spare, efficient structure, Apple concentrates on three or four important launches a year and releases a product only once it has been polished to perfection inside. In that arrangement, its internet services are more like dependents of a few pieces of hardware.

You can read Apple's view of the world in the way yesterday's keynote was arranged. The 120 minutes divided roughly into four blocks — watch, TV, macOS, iOS — split by the final hardware product.

That is how Apple sees the world. However important Apple Music and iMessage are, they arrive as part of the iOS 10 release.

And Apple's position on personal data security narrows what its internet services can do.

At this year's event Apple also stressed the importance of artificial intelligence, opening up Siri and adding natural-language recognition to iMessage. But as soon as he finished introducing the AI features, Federighi added that all the computing happens on the device and nothing is uploaded to Apple's servers.

Nearly every company today, Google included, is working hard to collect user data and process it on servers to make its artificial intelligence more accurate.

If Apple's most important goal is selling hardware, that choice is the right one: once users trust Apple's hardware completely, they are unlikely to move to another platform. Just as most iPhone users today won't switch to an Android phone, and Mac users rarely go back to a Windows PC.

But if Apple really wants internet services revenue to cover the hole the iPhone leaves, the privacy limits shouldn't be there. Nor should the pace of product updates depend on operating system releases.

Amazon, a company with limited experience in artificial intelligence, released the Amazon Echo and within a year it was considered better than Siri, which has been around for five.

The Music app built into 550 million iPhones keeps encouraging you to pay for a subscription. But over the past six months Apple Music has clearly slowed in attracting paying users. Its total grew by only 4 million — behind Spotify, a small startup.

Under its present constraints, Apple's internet services plainly cannot compete with the best internet products in the world.

In 2007 Yahoo's co-founder Jerry Yang invited Jobs to speak to the company's executives. Yang had just returned as chief executive and wanted to rescue a declining company.

Jobs told them: "In all seriousness, Yahoo could do anything. You have outstanding people and more money than you can spend. What I can't work out is whether you want to be a content company or a technology company. Pick one. I know which one I'd pick."

Today the same choice sits in front of Apple.

Cover image from Tony Avelar / AP