Two weeks ago I applauded Reed Hastings, Netflix’s CEO, for listening to his customers and reversing his decision to split Netflix’s streaming and mail order businesses. This past week we saw another high profile course change as H-P’s leader, Meg Whitman, stepped back from the tech giant’s plan to spin-off its PC business.
“The decision was actually very straightforward,” Mrs. Whitman said.
At first blush that statement seems surprising. Just seven weeks ago, in making the decision to sell the PC business, H-P had estimated related costs at $300 to $400 million. H-P’s most recent analysis revised that figure upward to $1.5 billion – four to five times the initial projection!
How does that happen?
It happens, as the title of this post suggests, because too often we accept critical assumptions without challenge; because we base our analysis on spotty, high-level data; and because we think in terms of a single event instead of the probability of a range of outcomes.
Perhaps the better question is not How this happens, but How often this happens. The answer, I believe, is too often.
List your assumptions clearly, and then test them – especially the most obvious ones. Seek out an additional level of detail in your data, and ask yourself what’s missing, and why. Play What If?, estimating the likelihood of difference scenarios, and how your answer changes as you vary key input variables.
These actions are at the core of great analysis….and exceptional decision making.