RETURN ON TECHNOLOGY: Why do technology costs outpace productivity gains?

I thought it was a joke. You know, like, “How is a sneeze like a drum solo?” And the answer would be something like, “You know it’s going to happen, but you’re powerless to stop it.”

So here it is, straight: How is business technology like the invention of electricity? But it’s not a joke. It’s an article in Slate magazine online (www.slate.com/id/2167909/). That article opens with a bit of surprising history. Electricity was widely available starting in the late 1800s, but it didn’t make any mark on American productivity until the 1920s and beyond.

Why so long? That’s an interesting question. It’s interesting because the same question haunts modern companies about their own new technology, since it, too, doesn’t seem to pack much of a productivity punch. The cost of new technology rises faster than productivity. That’s a mystery and a frustration.

It turns out that when steam power was king, factories were built tightly bunched around the central shafts the steam engine drove. My grandfather was a millwright, and he worked on machinery that picked its power off the whirling shafts in a plant. When electricity came along, it let the factory owner replace the big honking steam engine with a big honking electric motor, and things went along just about as before. My grandfather doubtless never missed a beat.

Then came World War I, scarce labor, and in the aftermath of the conflict, a booming economy (at least by early-20th-century standards). All sorts of devices started getting their own little electric motors. Freed from their long shafts, plants began to spread out horizontally instead of vertically. Industry began to design itself around electricity, instead of shoehorning electricity into the old factory. Work flow became based on natural assembly requirements, instead of shaft placement.

This might sound like a trite point, but it has a huge implication for today. Modern companies, with all good intentions, now buy software packages that cost millions of dollars to implement. Yet they expect the new technology to mold itself into the shape it finds in its new home. Companies routinely spend millions of extra dollars to write custom software in addition to the available code in these mega-packages, just to make the software do what the old paper systems did, only a little faster and using more fossil fuels.

New wine isn’t happy in old bottles, though. Left to find its own channel, new technology doesn’t just do old things a little better. It creates new things. It remakes. It eliminates steps. It redefines. But if its owner doesn’t want to redefine, all that expense has resulted in grafting computers onto the status quo, so productivity rises tepidly, if at all. At some point, one company or other will reinvent itself around the technology, instead of vice versa, and it will leap ahead of everybody else.

For example, when data from all over the organization is restricted to only a few managers who read paper reports, decision-making has to be made topdown. But if that data is instantly available company-wide, decision-making can be pushed both downward and sideways. The organization becomes flatter and wider, more decentralized. Decision-making can be faster, more responsive to local conditions.

Wal-Mart uses this model, allowing its far-flung stores to order and carry locally popular items, in addition to the general ones. Amazon used its technology to completely change the idea of what “retail” means. Even Wal-Mart has to limit its inventory size. Not Amazon. There are companies making fortunes from programmers who never meet one another. It’s true that first we invent our technology, then it invents us.

But the companies I mentioned didn’t have to reinvent themselves. Technologies permitted them to start and grow. There was nothing to dismantle first. Existing companies find it much harder to morph into new shapes. For an example of how hard it is, look at Disney’s wrenching shift from hand-drawn animation to computer animation, and how much the Pixar tail is now wagging the Disney dog.

The reluctance to let technology drive change appears to account for much of the failure of new technology. By many estimates, more than half of all technology projects fail utterly in their productivity goals, and a still higher percentage fail enough to be annoying rather than beneficial. New technology can boost productivity only if the organization radically reorganizes to use the new capability.

That’s tough to do because, while the rewards are great, so are the risks, especially if you’re the first guinea pig making the change. But by the time the sluggards fall into line, the first adopters are eagerly piling up profits. They do it by using what economist Joseph Schumpeter called “creative destruction.” If there’s a scarier phrase to scrawl on the boardroom whiteboard, I don’t know of one.