Improving productivity by automating management

The book The Numerati by Stephen Baker has generated a little buzz. It's about companies' efforts to model human behavior from real data generated by web surfing, electronic communications, etc. The thing that grabbed me about this review in BusinessWeek was the line about IBM's effort to "improve productivity and automate management."

I guess we've come a long way from the days when managers got together to decide which of the company's functions were going to be outsourced, and which were going to be automated and people replaced entirely. The notion of automating the management function has a sort of visceral appeal to anyone arbitrarily replaced by a downsizing or outsourcing effort. If you're running a company just by the numbers, and attach little importance to leadership, team building, and other kinds of "soft" skills that many managers put little stock in, maybe you could "automate" management. But I doubt it.

I skimmed the Workers chapter of the book that's excerpted in this review. It's not nearly as dire as the article in BusinessWeek describes. But it's a fact that some managers latch onto tools as a means for solving some of their problems without making a committment to learning the processes and assumptions implicit in the design of the tool. Buy, install, use. And if the results aren't what you expected its a problem with the software.

So we're safe from "automated management" right now. But then I thought personality testing as a means of making hiring decisions would never go anywhere.