In an uncertain, changing world, most decisions are wrong, and success comes not from the inspired visions of exceptional leaders, or prescience achieved through sophisticated analysis, but through small-scale experimentation that rapidly imitates success and acknowledges failure.

So says John Kay in his latest column in the FT, The centralised road to mediocrity, and Amen to that.

It is not true that there are no planned successes (think of the iPod), but it is true that most successes are not planned. SMS is one example: only ever designed as a notification mechanism for voice mail, it became a run-away success with teenagers and had a major impact on the profitability of the mobile operators.

So what to do? Experiment and measure. Predict from past results, and validate the predictions. Analysis and modeling can help inform and inspire your choices and save you from costly mistakes.

The new manager is a person who both deeply understands the business issues and is very data literate. That is a very rare person.