Making decisions is a big part of a CIO’s life, just as it is for any other line-of-business manager. And yet the process suffers from a misconception; the widespread, if rarely articulated, belief that decision making is about work and not part of the work itself. If CIOs can train themselves to structure the decision making process along the same lines as other parts of their job, the quality of their decisions can be greatly improved, says one prominent business author.
In a recent posting on “The Enterprisers Project,” an online publication and community devoted to examining the evolving role of the CIO, Nano Serwich draws attention to a Harvard Business Review article that distills some decision-making lessons for managers. The advice is drawn from a book titled Judgement Calls, written by Thomas Davenport and Brook Manville.
The book analyzes the decision-making process based on a study of twelve actual business cases in which organizations made the right decisions. In the Harvard Business Review article, Davenport distills the learnings into a short list of five precepts.
“How is it that managers facing high-stakes decisions, despite all the resources and knowledge available to them, often make them so poorly?” Davenport asks in the introduction to his article. “In large part, it’s because their whole perspective on decision-making is wrong. Managers think of major decisions as choices they must make in order for the work of the organization to proceed.”
The truth, Davenport says, is that decision-making itself should be viewed as work. If managers can train themselves to view decisions as tasks in their own right, they will then be more able to bring to that task the same discipline and direction they apply to any other part of their job.
“View decision-making as work,” Davenport says, “and you soon realize that many general principles about good work apply to it.”
The first thing is not to make decisions solo. Davenport debunks the “Great Man” theory of decision making, and invokes behavioral economics and decision bias research findings that all decision makers at all levels operate with decision processes that are flawed and non-rational.
No decision-maker is perfect: Davenport notes that “even the sainted Steve Jobs of Apple made some clunkers.” A variety of people should be consulted, with a systematic process for sampling their perspectives, Davenport says. He adds that social media can play a role, citing the example of EMC, which asked employees to offer their views on cost-cutting measures during the last recession.
Just as no one person has all the answers, no single decision-making tool is enough. In fact, relying on just one over and over again will produce suboptimal decisions in all areas. Instead, CIOs and other managers should combine analytical models, experimentation, experience, intuition and “‘wisdom of crowds’ approaches.”
Good decisions are invariably based on solid and exhaustive measurement, and although here again no one approach is enough, Davenport says data and analytics would be the single option he’d choose – if he was forced to choose just one. “Rigorous decisions depend on gathering some data and performing some analysis. Although no one should abandon human responsibility for good decisions to computerized analytics, they can provide a massive amount of help.”
Post-decision reviews are also invaluable, Davenport says. “Decisions will get better if you establish the habit of reviewing them after the fact. This requires ‘a culture of honesty and self-examination,’ as a manager at Chevron told us when we visited there.” If you can’t figure out why good decisions worked and bad ones didn’t, you’re not likely to improve the quality of your decision-making. Davenport cites a small home builder which has a process to examine decisions that don’t go to plan. “Managers there systematically assess not only ‘why didn’t this house sell?’ but ‘what made us decide to build it the way we did?’”
Finally, decision-making needs a structured process, just like any other kind of work; it’s not an ad-hoc thing, though many managers see it that way. The process will dictate who gets involved at what point, what kind of data and analysis should be applied, and how quickly the decision needs to be made. “If you’re worried about ‘analysis paralysis,’ or excessive focus on analytics (rare in my experience), a process will address the issue,” Davenport says.
Davenport wraps up by saying that anyone who feels he’s overstating the case is probably a very gifted manager, one to whom principles like these are second nature. Or else they’re clairvoyant and can make the right decision intuitively and correctly every time without other perspectives, logical processes, or analytical tools. But, he concludes, “if it’s that easy for you, perhaps you’re actually not doing it very well.”