By John Gruber
Atoms are the world’s first shoes to come in quarter sizes. Experience them today at Atoms.com.
And so what to make of the HP/Apple iPod deal?
It’s a big deal for HP, and it’s a very big deal for Apple. I thought there were a bunch of big announcements at Macworld Expo — GarageBand, the iPod mini, the updated iLife suite, the new branch of mathematics that says 40 percent of Mac users on OS X means “the transition is over” — but the HP deal is much bigger. I’m guessing that Apple had in fact hoped to announce it at Macworld, but couldn’t because negotiations between the companies weren’t finalized until late Wednesday night, according to the New York Times.
Why is it big? Well, for one thing, because HP is big. HP is a huge company; in the fourth quarter of 2003, HP reported nearly $20 billion in revenue, and over $1 billion in profits. Apple, by comparison, reported $1.7 billion in revenue and $44 million in profits in their most recent quarter.
In terms of PC market share, HP is second only to Dell — nearly one out of every six PCs sold is from HP. In fact, it’s hard to think of any company which would have been a bigger deal for Apple to license the iPod to. Dell has already launched their own digital music initiative (and how are those DJs selling?), so they’re out of the question. IBM is a bigger company overall than HP, but not in terms of PC sales or consumer electronics. Only Sony would be a bigger deal — and it’s hard to imagine Sony licensing a portable music player from another company.
Computer industry platforms exist at very different levels. For example, “Linux” is a very low-level platform — it’s just a Unix-like operating system kernel that can boot computers and provide access to hardware peripherals. KDE and Gnome are desktop GUI platforms that are built on top of the Linux platform. Tivo is another GUI platform, entirely different than a “desktop”, but also built on top of Linux.
Mac OS X and Windows encompass both these layers — low-level operating system kernels that boot computers and control hardware, and high-level graphical human interfaces for users (and APIs for developers to build their own apps).
I think that what Apple has built, with their three-pronged iPod/iTunes/iTMS, is an even higher-level platform, specifically for digital music (and perhaps in the future, digital media).
Platforms exist for different reasons. Low level platforms are perfect for certain tasks. If you run a high-traffic commercial web site, your servers might run nothing but low-level software. Just Linux or BSD, along with software built for Unix-like platforms: Apache, Perl, PHP, MySQL. No GUIs. No desktops. Just servers running low-level software, very quickly, and very reliably.
Desktop platforms have made Microsoft tens of billions of dollars and made Apple a household name. A desktop platform is what normal people expect to see when they turn on their new computers. There’s a very small hobbyist market for PCs that ship without desktop software pre-installed, but major players like HP and Dell need a desktop platform; and in the mainstream market, that platform is Microsoft Windows. The alternative for PC manufacturers is to replace Windows with an open source desktop — but even then, they’re including someone else’s platform. Apple is the only personal computer manufacturer that builds its own desktop platform — and it costs them a lot of money to do so.
With digital music, here’s what people want to do:
HP needs to provide their customers with solutions to all three. Apple leads the industry in all three. Hence the appeal of Apple’s iTunes/iPod platform to HP.
Why take a risk developing their own portable player? By licensing the iPod from Apple, HP starts with a market-leading product. Even from a company of HP’s resources, it’s far from certain they could create something of comparable quality from scratch (I’d say downright unlikely, judging by their handheld Windows CE devices). And if you’re going to license some other company’s player, why not license the best?
And in terms of the music store and desktop music software, it makes even less sense for HP to develop their own. Out of the three prongs — player, software, store — the only one where there’s serious money to be made is the player. You can’t make serious money selling iTunes-like software, because iTunes itself is free. It’s like trying to sell web browsers or email clients — the market is so small that it’s only appealing to small software companies.
The same goes for online music stores. Let’s be generous and say that Apple can make $0.25 profit per $0.99 song. (A more realistic estimate is probably $0.10, but let’s be generous for the sake of argument.) Let’s also speculate that they make $50 in profit per iPod. (They’re definitely making more than that on the higher-end models.) Even with these low-ball estimates, Apple needs to sell 200 songs at the iTMS to make the same profit they get from selling just one iPod. And remember, in reality, the difference is probably even more striking.
For a company like HP, which generates $20 billion in revenue in three months, it’s hard to get excited about making money in $0.25 increments. They could sell 100 million songs and it would barely make an impression in their bottom line.
The iPod is where the money is, and that’s what Apple offered to HP. It’s almost a sure thing that HP is going to make money selling their players, and their customers get great software (iTunes) and the best online music store (iTMS). This gives them an enormous advantage over their arch-rival Dell, which offers an inferior handheld player, an inferior music store, and inferior desktop software.
It’s clearly a huge win for HP. But what’s the upside for Apple? The surprise isn’t that HP said yes, the surprise is that Apple made the offer in the first place. Apple is not exactly known for licensing its hardware to other companies.
And at first glance, it’s hard to see where this benefits Apple. Surely Apple will make less profit per HP iPod than they make on their own — HP is obviously going to get a slice of the pie. The goal is obviously to sell more iPods, period, but Apple is already doing a good job selling to the Windows market.
Second, there’s the simple fact that HP, as a major PC manufacturer, is a competitor. Sure, most HP customers aren’t even considering a Mac, but some of them are. It muddies Apple’s message that to get the best digital hub experience, you should buy a Mac.
But the upside for Apple is long-term, not short-term. It’s hard to create a successful platform, and even harder to create a market-leading platform. You need the right product, the right price, and you need to ship it at the right time. But once established, leading platforms having staying power. (Cf. MS-DOS.)
There’s a widely-believed canard that Apple’s mistake in the 1980s was in not licensing the Mac OS to any and all PC makers. The idea being that if they’d done so when their usability advantage was at its peak, Apple would today occupy the position of OS monopolist, and Microsoft would not.
This is sheer fantasy. The entire point of the Macintosh, from the onset, was to focus on quality. Anyone who claims Apple could have achieved greater than 50 percent OS market share without sacrificing what makes the Mac the Mac has no comprehension of how Microsoft built the Windows empire.
Widespread licensing of the Mac OS was never an option. What was on the table, however, was a limited licensing strategy. Apple could have chosen a handful of A-list companies to license the Mac; companies like Kodak, Xerox, and AT&T (which was a major player in the computer industry in the 1980s). Even in the best-case scenario, the Macintosh never would have achieved a majority market share. But it is conceivable that Apple could have carved out a significantly larger slice of the industry than what they ended up with.
I am not arguing here that Apple should have done this, or that it would have been a sure-fire success. All I’m saying is that they could have done it, and it might have worked — securing a larger long-term market share without diluting the Macintosh brand to lowest-common denominator status.
Windows-style widespread licensing is for commodity items. Limited licensing is for premium items. The iPod is never going to achieve Windows-like monopoly market share. But that’s not the point — like the Mac, the iPod is about quality, not ubiquity. But unlike the Mac, Apple has the opportunity to secure a market-leading position for the iPod.
The market leader doesn’t have to hold a monopoly. (Cf. the auto industry.)
The HP-Apple licensing deal elevates both companies. The day HP ships their iPod, it will become the premier company in the Wintel digital music game. And for Apple, the HP deal changes the iPod from product to platform.
Bold storyline #1 in this saga is that Apple has for the first time licensed market-leading technology to another company. (The mid-90s Mac OS cloning experiment doesn’t count, because the desktop market share war was already over.)
Bold storyline #2, however, is the fact that this deal has nothing at all to do with Microsoft. This is important, because Microsoft very much wants to control the future of digital media platforms.
Their plan, not surprisingly, is to capitalize on the monopoly status of Windows. It worked with office software. It worked with web browsers. And they want it to work again with digital media.
Tim O’Reilly had a great line in his 2002 WWDC keynote address:
Microsoft gets a lot of heat for not leaving enough on the table for others. My mother, who’s English, and quite a character, once said of Bill Gates, “He sounds like someone who would come to your house for dinner and say, ‘Thank you. I think I’ll have all the mashed potatoes.’”
Microsoft has taken nearly all the mashed potatoes on the PC desktop platform. HP is a company with a proud history and thousands of very smart employees. But it’s nearly impossible for them to stand out in the Windows PC market; the commodity nature of Windows has made a no-name Brand X PC nearly equivalent to a PC from HP.
The HP-Apple deal is not exclusive — Apple reserves the right to license the iPod to others. But HP can rest assured that Apple is not going to license the iPod willy-nilly. The iPod will remain a premium brand; ergo HP’s iPod will be clearly superior to no-name music players.
Even Microsoft executives seem to have fallen for the line of thinking that Microsoft’s technologies are de facto industry standards. In response to the HP-Apple deal, a Microsoft spokesperson declared — I kid you not — that it would be bad for “choice”. As The Apple Turns offers a fine translation:
Check it out, this was the best a company spokesperson could muster to undercut the announcement: “Windows is all about choice… we believe you should have the same choice when it comes to music services.” Translation: “Use any service you want as long as it sells Windows Media, buy any player out there as long as it plays Windows Media — but for heaven’s sake, don’t buy one of those wretched iPod thingies or we’ll be completely boned with our whole plan to monopolize digital media commerce and then we might actually have to start innovating for our paychecks for a change.”
The previous times that Microsoft has parlayed its operating systems monopoly into other monopolies — e.g., office apps and web browsers — they were still within the realm of the PC industry.
The difference with the digital hub (to use Apple’s parlance) is that it extends beyond the PC industry. The market for digital cameras, for example, is not a subset of the PC industry — it’s a subset of the photography industry. Likewise, the market for digital music players is simply people who like music. Just because the early adopters of digital cameras and MP3 players were indeed computer nerds doesn’t make them PC peripherals — they’re just cameras and music players.
The Microsoft company line seems to be Why wouldn’t a PC manufacturer base its digital music player on Windows Media?
Which, it seems to me, is all wrong — and perhaps a sign that Microsoft’s institutional paranoia has finally been wrung dry. In light of the HP-Apple deal, the better question would be, Why would a PC manufacturer go with Windows Media for its digital music products?
If they took all your potatoes the last time you invited them to dinner, why invite them again?
Apple’s music platform runs on Windows, but it is independent of Windows. It is therefore independent of Microsoft. In addition to the fact that Apple was able to offer HP a superior music platform than Microsoft’s, it’s also the case that Apple is a much more appealing partner. Given the relative sizes of the companies — even after accounting for Steve Jobs’s ego — it’s obvious that Apple is in no position to push HP around. The amount of revenue Apple generates per quarter, HP generates in a week.
HP’s blockbuster deal with Apple will have one exciting side effect. The company will be working with Apple to add support for Microsoft’s superior Windows Media Audio (WMA) format to the iPod by mid-year. You heard it here first.
Of course, the reason you heard it there first is that it wasn’t true. Unlike Thurrott, Wired News reporter Leander Kahney actually spoke to someone at HP, who told him:
“We’re not going to be supporting WMA for now,” said Muffi Ghadiali, product marketing manager for HP’s digital entertainment products group.
Predictably, however, Thurrott’s speculation was linked all over the web, although commentary seemed more focused on Thurrott’s claim of format superiority (cf. this article at Slashdot) rather than on whether the idea even made sense.
And, frankly, it doesn’t make sense. You might make a case that the iPod could support unprotected WMA files, but this follow-up column from Thurrott shows that he’s clearly talking about WMA files protected with Microsoft’s proprietary DRM. How could it possibly benefit either HP or Apple to cut Microsoft in on this? They might as well pull down their pants and get out the Vaseline while they’re at it.
Anyway, speculation on WMA support isn’t nearly as intriguing as speculation on who might license the iPod next. Those are the rumors I’d like to hear.