Life in the Beta Lane

Another week in the US, another new release of a revolutionary product from Apple. The iPhone 4 has already sold upwards of 1.5 million units and it would have sold many more had the manufacturing lines been able to produce them quickly enough.

This extraordinary market performance aside, the iPhone launch was marred by a huge controversy over cell signal degradation. It appears that if you are left handed and you hold the phone in a particular way, you lose cell signal and your phone drops calls. That’s arguably close to a fatal flaw for something that ultimately is, despite all the other applications that run on it, first and foremost a ‘phone’. Has something gone terribly wrong at Apple, and by implication -- since Apple currently is at or close to the pinnacle of it -- in America’s technology innovation system?

That’s the wrong question. The right question is, what can we learn from this incident about technology innovation in the real world and how consumer expectations misunderstand and potentially distort it?

Imagine yourself for a moment the architect of a small family house. In a fascinating book titled How Buildings Learn Stewart Brand wrote “All buildings are predictions, and all predictions are wrong.”

Good architects know this implicitly. When the family tells the architect how they believe they will use the house, what their patterns of behavior are going to be, and thus whether they want the kitchen open to the living room or behind a wall, they are making de facto predictions about the shape and feeling of their lives over a long future.

Those predictions are almost sure to be wrong - not because the family is stupid or insincere, but because the family will change, the world in which they live will change, fashions will change, and the unexpected will certainly happen. Maybe ten years from now they will have more children than they could have imagined or they won’t cook their own meals and won’t want a kitchen at all. It’s the multiple intersecting uncertainties of human behavior that guarantee a vast on-going market for building renovations.

Now imagine yourself the engineering team that must design a communications and computing device that millions of people around the world will carry in their hand. The team is operating at once very close to several technological horizons: Physical materials, battery systems, radios, software, and more. They are faced with the challenge of integrating these and other elements into a consumer device that will dominate an extraordinarily competitive market of alternative devices. And they have to do it all beautifully, secretly, and at a price point that is plausible during a worldwide recession.

Software engineers know this kind of problem well. Another fascinating book that has stood the test of time is Frederick Brooks’ The Mythical Man Month, reflections by an IBM engineer on the intersection between human capacity to design and communicate about complex technological systems, and the infinite ways in which those systems will be used when they are released into the wild.

Brooks explains eloquently why software has bugs: the number of possible pathways through even a moderately large body of software code is infinite and they can’t all be tested. People will use software in ways that engineers can not anticipate. The essence of real field testing is lack of rigor -- you’ve got to get the software out into the world and let it live in the randomness of machine diversity and human use to find out where the unexpected and unforeseeable problems will emerge.

That’s life in the ‘beta’ lane, and it’s the way real technological innovation happens. Even in simple software packages there are too many variables to test and control. No one can ‘get it absolutely right’ the first time (or the second time or the third time). The real test of good design is how quickly and transparently the system manifests its bugs, and how easily they can be fixed.

Add the layered complexity of a multi-tiered system where you are responsible not only for the software but also the hardware, the physical design, and the integration among all these elements -- and where you don’t control the people who will use the phone, the network on which the phone operates, and other important variables -- and the challenge should be evident.

If you shoot for perfection on the first day of release, you’re guaranteed to fail. Even worse, you’ll probably pull back from the technological frontier in order to exert more control over the variables that you can control -- and thus reduce the innovative potential of your team. And, it still won’t yield perfection. It will just yield a less innovative product, that will fail in some other unexpected way.

This is where consumer expectations have gotten out of line and in doing so pose a threat to the innovation ecology. Consumers, at least in the US, have come to expect perfection in use at first encounter. They don’t want to live in the beta lane. If companies try to meet these expectations, they will undermine their innovation systems. And to no good effect, because the bugs will still appear. They’ll just be different bugs, on a less interesting and revolutionary platform.

What’s needed here is a bit of honest and straightforward consumer education. The test of good design at the technological frontier is how quickly bugs manifest and how quickly they can be fixed. There are trade-offs in innovation and you can’t have everything you want on day one. This doesn’t mean that consumers should be willing to put up with faulty products (they shouldn’t). It means that intelligent consumers and audacious innovators need to meet each other in a kind of rapid dialectic of bug and response: like a 4.0 / 4.01 / 4.1 cycle in software revision at a very fast rate. This is life in the ‘very beta’ lane but at a speed and responsiveness level that lowers the marginal disutility of (initial) "flaws" in products.

For better or worse, Steve Jobs doesn’t quite have the personality to explain this gently but persuasively to American consumers. But someone needs to. The stakes here go beyond the success of the iPhone and Apple Inc -- they go to the innovative potential of modern economies. Innovation at the horizon is a large numbers game of trial and error. The only way to reduce error to zero is to reduce trial to zero.

Email the writer: