Consider the camera phone. In the space of less than five years, a product that most market analysts scoffed at when it was introduced now outsells digital cameras by a 5-to-1 margin. At this rate, camera phones will soon outsell film and digital cameras put together. Not only have sales exceeded all expectations, but the product also has had a huge impact on the semiconductor industry.

That's just one recent example of the unpredictable nature of the consumer electronics market. The success of the camera phone traces its lineage way back to the Sony Walkman, another well-documented surprise market success that was equally panned by the pundits when it was introduced.

But for every such success, there are several much-heralded products that failed to live up to expectations.

In the long history of the consumer electronics industry, no one has been able to anticipate consumer behavior with reasonable reliability. Yet developers press on.

Indeed, photo messaging is so seductive a concept, for so many players, that a whole industry has sprung up around the quest for a "tipping point." Whether consumers, in mass numbers, can be convinced of the utility of zapping pictures from handset to handset is a matter of critical concern to many industries: Phone makers are looking for new applications and features that will help them ward off the specter of market saturation; cell service companies are counting on new revenues from multimedia; photographic film suppliers, obviously, have reason to be wary; and, perhaps not so obviously, even the printer business may be affected.

This is how the exercise goes: If you had a camera with you at all times, would you be more likely to take pictures? Probably. Is there something that would guarantee you would take more pictures? Perhaps, if it allowed you to do something with the pictures immediately. Might the ability to beam the pictures to a local printing kiosk for pickup on the way home be the ticket? (If so, printer manufacturers could strike it rich.) Or, would embedding a voice message be the tipping point that would have grandmas everywhere clamoring to see-and hear-baby photos zapped from the proud parents' phones?

Nobody knows quite what influences human behavior; predicting the consumer electronics market is not for the faint of heart. Still, as you might expect, history affords certain broad conclusions.

The first is that "logical extensions" in terms of performance, size and cost tend to do well. What that means is that "smaller, faster, cheaper" are fairly good bets. The evolution of the personal computer is an obvious example. So, too, are the DVD and the LCD TV.

The DVD, in fact, wasn't just a logical extension; it was a logical discontinuity compared with the videotape cassette. The DVD provided all the features and convenience of videotape, but with far better picture and sound quality; greater durability; the ability to skip instantaneously back and forth; and the ability to access extra material, such as directors' cuts and even simple videogames, on disk. So far, it hasn't been much of a recording device, but that is about to change. And in any case, despite this drawback, the DVD's ascendancy over the videocassette reaffirmed that prerecorded playback-rather than recording-was the far larger market.

Just as the advent of the CD, as a replacement for vinyl, expanded the market for recorded music, movies on DVD are proving to be a boon to Hollywood. What is particularly exciting for the movie industry is that people are often choosing to buy DVDs rather than rent them.

It's also a logical extension to think that TV sets can never be too wide or too thin. Indeed, increasingly they are both wide and thin, thanks to the growing popularity of flat panels based on plasma and liquid-crystal displays. And, as their cost comes down and size goes up, LCDs increasingly are the display of choice over plasma; they are longer-lived and can double as computer monitors.

Still, some very logical extensions either have proved to be duds or remain suspect. Microsoft's much-heralded Tablet PC was conceived as a more powerful, larger version of the surprisingly successful Palm Pilot-but it went nowhere. Microsoft is now applying its considerable software muscle to making the platform smarter and more useful.

The logical extension to text messaging-yet another one of those delicious surprises that initially was pooh-poohed but then took on a life of its own-would be to provide small or clever keyboards as part of the phone, to reduce the gymnastics involved in typing words on a phone keypad. But would the idea fly? The first such devices are just hitting the market. Time will tell.

It's also clear that design-the ineffable synergy that results from a perfect marriage of form and function-is an important predictor of success in the consumer electronics market. No other company gets it right more often than Apple Computer, whose gorgeously distinctive products earn praise from reviewers and almost cultlike devotion from customers. Similarly, Nokia, the world's leading handset maker, acknowledges that phones aren't technology, so much, as they are extensions of the end user's personality. Automobile manufacturers caught on to the possibilities of that consumer mind-set long ago.

Consider Apple's iPod, which took the portable-music-device market by storm. Its runaway success is usually attributed to the fact that it does one thing (play back downloaded music) and does it very well (storing thousands of tunes). That alone makes the iPod a worthy successor to the venerable Sony Walkman.

Two recent developments epitomize the conundrum of consumer electronics: TiVo and satellite radio.

When it debuted in 1999, TiVo, whose name now has become synonymous with the digital video recorder (DVR), promised to revolutionize the TV business by allowing users to sidestep advertising and escape the tyranny of TV schedules. But, despite gaining some ardent supporters-FCC commissioner Michael Powell famously called it "God's machine"-TiVo has flopped.

Or has it? Actually, what has flopped is the original business model: Only a small fraction of the projected subscriber universe signed up for the service. Now, while it will still sell its subscription service in the U.S. market, TiVo appears to be concentrating on licensing its technology to hardware manufacturers and pay-TV operators.

TiVo supplies its DVR technology to DirecTV, the biggest satellite-TV operator, and recently signed a deal with Netflix. Sony also licenses TiVo technology for its DVRs. And set-top box makers such as Scientific-Atlanta Inc. and Motorola Inc. are independently entering the market. TiVo, the brand, may not make it, but the revolution it promised is imminent.

The benefits TiVo promised to TV viewers were, and still are, obvious. It may turn out that those benefits will be delivered through others.

Satellite radio, on the other hand, appears to be promising similar benefits to radio listeners (commercial-free radio, for example), that, somehow, don't appear as comprehensively compelling. Commercial-free music is fine, but it's still not as good as being able to choose exactly the music you want, when you want it. (The iPod is clearly better on that score.) And being able to drive coast to coast while listening to the same station doesn't sound all that exciting. The conventional-and largely convincing-wisdom is that it's radio, after all, and radio is a niche business that plateaued a long time ago.

Wrong again! The handful of satellite radio firms that have been in business for a couple of years have already racked up more subscribers than the DVR business.

But the outcome of the biggest battle for the hearts and pockets of consumers will hinge on the answer to this question: How smart should a home be?

The home as digital nirvana has a number of variants. The PC-centric vision-endorsed, not surprisingly, by Intel-has the PC as the media center that controls films, songs, games, all transmitted wirelessly to screens and speakers throughout the home. Sony and others from the more traditional consumer electronics camp want game consoles as the uber-controllers of the digital home. Even the DVR and set-top box crowd wants in on the action. Then there are the fringe futurists who foresee refrigerators that re-order milk, door openers that sense the identity of people who approach and other, more esoteric automata, all connected over the Web.

Given the track record of prognosticators in the business, it would not be wise to bet for or against any of those camps.

Carver Mead, professor emeritus at Caltech and prolific contributor to electronics technology, said something recently to Forbes that may apply perfectly. "Any specific prediction we may make about upcoming invention is bound to be wrong-even if we are talking about our own work. Our predictions are really just our aspirations. As we learn more and follow our heart and our instincts, our aspirations change.

"Eventually our work leads us to a place we never imagined. But it's usually completely different from anything we ever predicted. And, much better, always much better."