Worstall on Weds If the theory of disruptive innovation is wrong then why do companies act as if it's true?

Proof is rather in the pudding after all.....

The latest topical shouting match in my corner of the economics-meets-tech world is an essay by Jill Lepore in the New Yorker. All this talk of “disruptive innovation” is just that – talk – and the theory itself doesn't actually hold out in the real world. So, Yah! Boo! And all that really.

The real point of this essay by a Harvard professor is the usual one of such essays: to show that all those who are not Harvard professors are stoopid, so there.

Leave aside the “angels on pins” aspect of her argument and the problem with the dismissal of the disruptive innovation argument is that it does in fact explain a great deal of our world. There's a truthiness to it which makes it still relevant.

Newbies always kill off established businesses

For example, we know very well that large companies, well established in their industries, are a great source of innovation over time. Using William Baumol's definitions for a moment, innovation means the continual advancement, by incremental steps, of a product or technology. However, they're not a great source of invention: those great leaps to another way of doing things entirely. Those are (almost) always the province of new entrants to the markets: either new firms or those moving from one marketplace to another.

It's thus a standard assumption, an assumption well backed by masses of empirical evidence, that large companies do increase productivity in those small steps. Yet the major advancements come not from the development of extant businesses but from the entry of new ones into the market and the decline of the old. Or, as we might put it, the upstarts clubbing the oldies to death with their new inventions.

Given this, there's obviously an interesting theory to be constructed about disruptive innovation. The one we've actually got, the one that Lepore is arguing against, is that these new products are initially worse than those they are replacing in many manners. But consumers see something there that extant technology doesn't provide and thus it replaces the old.

Practical examples? Well, if you must...

There is explanatory power there: in 1895 the car was clearly inferior to the train, the dominant transport technology, for actually getting somewhere. But the car offered the opportunity to get anywhere, which is what was so disruptive.

I suspect that a truly useful fact about this disruption is that, almost by definition, it won't come from the incumbents in the market. Not because of any economic woe, just because of the incentives faced by the managers and shareholders of said incumbents. One of the three great management books of all time, Up The Organisation (the other two are Parkinson's Law and The Peter Principle, those three contain more good sense about management than an entire Business School faculty) had it as “pissing in the soup”. Why would you pursue a product or technology that is simply going to compete against your own highly profitable current products?

You might, of course, if you see that others are coming to steal your lunch using that new technology but by then it's often too late, as Kodak found out with digital photography.

So this version of the theory tells us that it won't be the record player manufacturers who create iPods, won't be (until late in the game) Microsoft that releases a free operating system and so on. There are examples that prove (as in the exact sense, to test) this. It shouldn't have been IBM that released the PC. They did, but almost as a guerilla project and without thinking through the effect it was going to have on their core business.

Google, Big Auto, digital maps and driverless cars

All of which brings us to the next technological wave about to engulf us, the driverless car. There's much muttering about how Google isn't getting much joy out of trying to cooperate with the behemoths of Detroit.

Disruptive: Driverless cars will DESTROY DETROIT. No, really

The latter would like to see small and incremental additions of useful technology to current cars (aid with parking, for example) and the former wants to go all out for entirely autonomous vehicles.

One can see this from both sides and in doing so it's possible to divine why there's unlikely to be much agreement between them.

Detroit obviously wants to continue with the current model of each adult (more or less) having their own car which gets used for perhaps 90 minutes a day and then rusts quietly the rest of the time. That's what supports their unit sales, after all. The important thing here is maintaining the idea of individual ownership of a vehicle.

Google, on the other hand, is clearly interested in the idea of near-public fleets that are in use almost all of the time. This would, of course, kill off the volume of sales. The two business cases (or “use models” if you prefer) are clearly in conflict. So it's not at all surprising that Google and Big Auto are not getting very far in cooperating. One is trying to offer a disruptive technology, something that will entirely upend the economics of the extant industry. The other is the extant industry which would really prefer not to be upended, thank you very much.

Thus, if Google's model of driverless cars does come to pass, it's going to be through a new entrant to the marketplace, not via one of the incumbents volunteering to be taken out behind the woodshed.

True disruption: finding a guaranteed revenue stream

Another potential sticking point is the maps which were developed by Google and are essential for its robo-cars to operate, says Sven Strohband, a robotics expert who worked at Volkswagen until 2006 and who was not involved in the discussions.

That data, compiled by Google, can be extraordinarily detailed, down to the height of kerbs or location of signs. “The question is who owns the data,” he said. "You need to have frequent map updates and your car can only go where you have really accurate map data."

This answers the question of what Google might be able to get out of their work on these cars. If the technology depends upon those accurate maps, maps which need continual updating, then Google can (perhaps should, or could) charge an ongoing fee to everyone using said technology. For road layouts do change gradually over time and so you'll need to have today's maps in order to be able to get to places.

$20 a month would produce a very good revenue stream indeed if every car, even the reduced number of shared vehicles, simply had to have such a subscription simply to remain in operation. ®