The downside of automation


Nathan Zeldes, former Intel engineer and author of the seminal paper on Infomania, argues that IT tools can reduce productivity. He doesn't suggest that computers and information technology, writ large, is a bad thing (he's an Intel guy, after all), but rather that any specific IT tool might not be good for the organization. He describes a typical situation:

I’ve seen many an MD cursing under their breath while struggling to enter my examination data and conclusions into a new computerized system. Instead of scribbling a few illegible lines on paper and chucking it into a manila file, to be processed later by an assistant, they had to use an unfamiliar and possibly ill-designed piece of technology, and it took them much longer. And because of this they had less time to apply their real value added, their precious ability to cure the sick.

Zeldes isn't advocating a return to the 50s, complete with pink collars, steno pads, and 3-martini lunches. (Although, who knows - he might be a fan of Mad Men.) He realizes that the benefits of IT are enormous. But I think he raises an interesting issue: the downside of IT systems and automation.

Zeldes says that usually technology

gets deployed with little attention to the wider implications. Thus, if a tool enables the manager or engineer to do the admin’s work, the temptation to remove the admin and become “lean” and “efficient” is great. But the fact is, an admin is paid much less than a highly skilled engineer or manager (or surgeon); and the latter only has so many hours in a day, which may be better used for doing higher level tasks. This is not to say that we can’t streamline some of the work by having it done by the manager; the question is which part, and to what extent. As is often the case, it’s pretty much about identifying the correct balance.

Toyota is famous for being very slow to introduce new, expensive, technology: they never want to automate a broken process. That slowness to add technology also enables the company to understand how it will affect the value stream, and whether that's wise.

When I see companies leaping at technological solutions for time and attention management, I have a feeling that they're in for a big disappointment. Buying a piece of software isn't a cure for poor work flow any more than buying a bigger pair of pants is a cure for your weight problem. Understanding the root cause(s), developing multiple countermeasures, and going through several PDCA cycles is a more reliable route to success.

2 Comments