Software Everywhere — David Beers on Mobile Computing

About once a year I cave in and indulge the slightly narcissistic urge to draw up a design for a mobile device that I think someone should be selling. I say narcissistic because such exercises almost invariably end up being about what you would buy, not what would be attractive in a broad enough market for the product to be viable. Mobile computers and smartphones differ from any other computer that's ever existed in that they are so personal in every sense of that word. Everyone needs something a little different. Even so, it's fun to think about what kind of product you would want to release if you were the product manager for a company like Palm, Nokia, RIM or HTC.

One thing I'd be thinking a lot about is how to move mobile devices— especially the high-end, high-margin smartphones—in an even more personal direction. I believe we're a long way from knowing what market segments exist—or could be created—for smartphones. That's one reason that the market has not taken off quite as quickly as many analysts predicted. It seems to me that everyone keeps repeating the mantra that "personalization is the killer app" and then seems slightly stumped that most of the personalization that really interests people has been at the low end of mobile phone functionality: ringtones, callback tones, wallpapers and the like. Why aren't the much greater personalization opportunities afforded by third party applications driving consumer interest in smartphones the way they did for PDAs five years ago?

There are a lot of reasons, but one that I haven't heard anyone mention is that smartphone hardware is less customizable today than PDA hardware was five years ago. Back in 2000 when I got my first PDA the hot device was the Handspring Visor, which had a wonderful expansion port called the Springboard slot. Dozens of third party hardware companies produced Springboard modules that enabled you to turn your Visor into a digital camera, a serviceable smartphone, an MP3 player, a WiFi-enabled war-driving machine, a GPS navigator, a barcode scanner, a mass storage device, or what have you. The Springboard phenomenon was so successful that hardware companies invested in products that today seem almost ridiculously niche: $400 modules designed to improve your golf game, personal massagers, and medical modules to deliver timed dosages of insulin to users that are diabetic. At the peak of the PDA phenomenon there must have been 30-40 modules available and they made the Visor addictively customizable.

Handspring was among the first mobile computing companies to foresee that stand-alone PDAs would quickly be outmoded by mobile "communicators" with cellular radios and they can hardly be blamed for turning their focus away from the Visor line and toward the new Treo. But one of the unfortunate casualties of that transition was that they weren't able to carry over the vision of user-customizable hardware.

If I were Palm's product manager, restoring that vision would be one of my top priorities. As nice as the Treo is for people that want to stay on top of all their communication channels while they are mobile, it suffers from the same problem that all smartphones do today: the design decisions of what to optimize and what to compromise, no matter how well-conceived, are baked permanently into the device. This deprives customers the chance to make many of their own decisions about what mobile computing uses are most important to them. Or even to morph their device to suit the occasion: business trip, vacation, morning trail run, moving around the WiFi-enabled office or college campus, surfing on the couch.

Modularize so users can "build their own" smartphone

I should be able to decide myself that having 4 gigs of storage or a WiFi adapter is more important to me right now than a camera, for instance. The camera, the memory and the WiFi adapter should all be modules that pop in and out of my phone as easily as changing the battery. Instead of releasing one new Treo at a time, Palm should release four or six of them—all the same base unit but with different modules orienting the device toward different market segments. Standardize the form factor and interface (SDIO would be a good choice) and watch the module ecosystem grow your loyal customer base. Nokia had to build a complete handset to provide functionality for runners and athletes—the kind of niche product that only a company the size of Nokia could risk. With a modular system like this a smaller vendor like Palm would only have to release a module with an accelerometer and GPS that would pop into its base smartphone. Or let a third party assume the risk of building the module and bundle it after you know there's a good segment for it.

Don't compromise on the physical interface

There's another baked-in compromise found in most smartphones today that I'd want to eliminate, at least for the high-end models: stop forcing people to use the same physical interface to make and receive voice calls as they do to send an email or browse the web.

Ever since the BlackBerry took off, there's been a distinct trend toward replacing the 12-key dialing pad with a tiny QWERTY keyboard. That's great for entering text, but lousy for dialing a phone number—or even speed-dialing while your eyes are on the road. Voice is still the primary function of a smartphone, so if you want to expand the smartphone market you shouldn't be sacrificing the voice calling experience to add other functionality. Nokia came up with the right answer with their 9000 series Communicators: a clamshell design with a standard phone keypad on the outside and a full keyboard (and larger screen) on the inside. This form factor has come a long way from the day when it was referred to as "The Brick" but the current models are still a bit too big and boxy to appeal to anyone but a business user or gadget freak. And they lack touchscreens, which simplify the use of complex applications enormously. What I have in mind for my ideal smartphone is something more like this, which is kind of a cross between a Nokia 9300 and a Samsung SPH-i500:



Viewed from the side and back you can see the modular influence of the Handspring Visor:



If this handset looks a little chunky for a "dream machine," consider that it's still quite a bit smaller than the year-old Nokia 9300, which similarly packs in two keying systems and two screens.


Given that we are adding a module system and a touchscreen digitizer to the interior screen, I may actually be pushing beyond the envelope of what can feasibly (or economically) be done in the way of miniaturization. Clearly, small size is a big factor in smartphone adoption, especially among new users. But how long can it be before a "convertible" design like this can be realized with the slim profile that so many consumers expect today? I believe the company that achieves it will be the one that sees mass market consumers starting to drop the extra C-note for a smartphone and the carriers chipping in whatever difference is needed to make the deal go down. And my guess is they will subsidize a handset like this heavily: the roomier keyboard will deliver a text entry experience that even the Baby Boomers will enjoy and the near-VGA screen will make the web and streaming multimedia a pleasure—both use cases that drive demand for premium wireless data plans.

If this form factor and modular design catches on the way I think it would I'd expect to see platform and software developers experiment with interesting ways of using the wide aspect ratio inner screen. Imagine your list of selectable RSS feeds in a column on the left and a browser window on the right. Or a split view of a document that lets you see a whole page at a time. Or the ability to have two chat windows—or even two separate applications—share the screen. It's admittedly a bit of a geek dream, but flash a device like this on the screen during a couple of episodes of 24 or the next Mission Impossible movie and I suspect you'd bring out the inner geek in a lot of folks shopping for a new phone.

Or is that just me projecting my own particular brand of gadget lust on the masses? Probably. I don't care anymore. Someone just build this thing so I can buy it, ok?

Update: After posting this I was contacted by Matt Hamrick, who is part of a group that really is building their own smartphones from parts ordered online. Their hardware designs are being released under the Creative Commons License to encourage others to join in their "open source" hardware project and according to Matt a couple of the designs are a lot like my drawings above. Read about it in my followup post here.