Or maybe not.
There has been a search for the Next PC since the first PC. In the late '80s, John Doerr thought it was pen-based computing, and invested in Eo and Go, later merged into what a long-forgotten wit called Ego. Now Ego is long forgotten. In the early '90s, John Sculley thought it was PDAs, and he began his long slide from tech fame after his infamous proclamation that the Newton would win a major part of a "Trillion Dollar Market." In the mid-'90s, Jeff Hawkins said handwriting recognition was not possible on a small device, and then he went on to prove himself wrong a year later by founding Palm. He left Palm and innovated a second time by adding cellphone capability to a PDA , calling it the Treo; but by then was destined to lose out to the RIM and its addictive Blackberry, called the Crackberry among the Digerati. Why he lost is very instructive - more on this below.
In the 'Oughts we now believe that the Cellphone is the Next PC. And what a PC it is! The PC is a mature product with annual sales around 200M units, while the cellphone has rocketed past those volumes towards 1B units per year. Features in the cellphone either get adopted at massive scale, or dropped. Cellcams made the cut. Music is now being tried. Microdrives may be next. And keyboards. And so on. Is it the next PC? Let's take a closer look.
The conventional wisdom (CW) sees the cellphone as a black hole in the next decade in the same way the PC was in the '80s. The PC sucked into its orbit all platforms of specialty functionality. No more Wang dedicated word processors, no more Radio Shack reporter notebooks, no more Convergent Workslate spreadsheet machines, no more Four-Phase data entry computers, no more NCR POS computers, no more Commodore Video Toasters, no more small business database computers, and no more Osborne portables. All replaced by the higher-volume and general purpose ubiquitous PC.
Game consoles persisted. This is occasionally proffered as an exception, but game machines fill a different niche in the computer ecosystem. Check out this a chart of that ecosystem, borrowed from the VC pitch of the founders of Apollo Computer in the mid-'80s. They justified the workstation market with a brilliant synthesis of computing niches. They showed that computing categories vary by a 10x price/performance difference. In the prices of the '80s, PC's were in the $2k segment, while game consoles were in the $200 group. To continue their analysis: supercomputers at $20M, mainframes at $2M, minicomputers at $200k, and the then-empty corridor of workstations/servers at $20k. After the PC and the game console we have calculators at $20 and digital clocks at $2.
Interesting is that no major category has emerged in the gaps - the $750 device is a well-traversed death valley. Mobility tends to get a premium, so where a desktop is $2k a laptop could be $3k, and whereas consumer goods tend to sell faster at prices under $200, portable devices can sell in the $250-300s (iPods, high-end cellphones).
The CW on cellphones, however, may be wrong. What drove the PC to ubiquity was the need for software compatibility. Then scale economics drove out special purpose platforms. Today the lock-in is much less than it had been. Most users today need a small set of programs to be productive - Office, Adobe Reader, Windows Media Player, and of course a Browser. Much of the rest of the stuff on a PC is designed to protect it from threats and problems - not a productive use of computing resources. This change is due to the Internet. When the PC became connected, the game shifted from software compatibility to connectivity. Application logic can run in the network, and the point of control is no longer the OS in the device.
The two biggest successes of the cellphone era make this clear:
- Blackberry - a specialty device coupled to a service
- iPod - a specialty device coupled to a service
This is not for the lack of trying. There are a number of software companies competing with RIM and its Blackberry - but significantly, all with the PC business model of software separate from the hardware. There were mp3 players well before the iPod - remember the Rio? They were uncoupled from the music service, analogizing to the PC model of separating software from hardware.
This is also not for lack of industry reaction. The music industry would like to break the dominance of Apple and the iPod.
Thus, while the concept that either a feature makes it in the phone or not is instructive, we should not analogize any further to the business models of the PC era. Whereas the PC became a singular device - common underlying platform and common OS - the cellphone has remained a heterogeneous device with multiple platforms and OSes; and it is connected, meaning the application logic resides at least in part in the network, reducing the importance of the OS.
When might this change? One of the lessons learned in technology markets is what it means to cross the chasm from a cute technology to a mass market product. Geoffrey Moore worked with Regis McKenna and embodied a decade of learning into the must-read tome of that title. This lesson is so often violated that Yelnick wonders if anyone actually read the book. A technologist conceives of market penetration as working up the technical stack, from technology to platform to solution. Instead, what happens is a technology captures the attention of the early adopters, but to cross the chasm a company needs to offer a solution to a niche market. Once this beachhead market is conquered, the company earns the right to move into neighboring markets. Later success across many markets allows the solution to become a broad-based platform. Technology, solution, platform.
We see this vividly with the iPod. The early mp3 players were tried by the techno-savvy, but never became a phenomenon until Apple coupled them with iTunes, a legitimate music service. Now Apple dominates this niche with a solution. Is the iPod yet a platform? Not even close. Apple first tried to extend the iPod into a platform with an OEM deal with HP, but that deal has fizzled. Recently Apple tried to move into cellphones, but the initial iPhone - the Motorola ROKR - does not have the easy to use interface, or the memory capacity, of even the smallest iPods. It is not yet a solution for that segment, and so far is doing poorly. We should also expect Apple to move into video, but not until the full solution is packaged.
We also see this with the Blackberry. None of the competitors have made much a dent. The closest is Good Software, which has caught a hot product, the Treo. The Treo is well-designed and works well with Good. It is as close a model as we have seen of the old PC model - software/service separate from the device. Even so, it is not yet as smoothly integrated as the Blackberry, and has made only a modest dent in market share. Following Good is Visto, which has focused on an even broader mix of devices, but is not identified with any one like Good is with the Treo.
At some point the cellphone becomes mature, and the linkages between service and device standardized. Then we should see a rapid shift in the model. Until then, the business model of the cellphone era is to innovate in devices and integrate the service with the device. View the Treo with Good as the canary in the mineshaft for the transition to a more open business model, at least in the email-to-device niche. The next canary is the iPod - watch its market to see if the stand-alone mp3 player sustains its market dominance or gets sucked into the black hole of the cellphone (once the integration is done well enough on a cellphone to package a solution).
If we move beyond the canaries to the gorillas, the next big battle is the portable media player, which integrates video with pocketable devices. Video is already being tested on cellphones, to a mixed reception, but a potentially huge market. More on this in a subsequent post.