James Taylor's Decision Management

Gary Cokins had a post recently titled "Are Performance Management and “Uncertainty Management” Synonymous?' in which he discussed performance management and uncertainty management. He makes the great point that the purpose of analytics should be to deal with uncertainty, not just analyze the past. He goes on to give a great example of a company analyzing their data to see what might help them predict which customers were a retention risk and how valuable this was.
What he does not really discuss is how to put this uncertainty to work in operational systems. Having figured out what predicted retention risk - having turned that uncertainty into a probability - the company had to put it to work. They did this by identifying a decision - whether or not to make a retention offer to a particular customer - by embedding that decision in their process(es) and by making that decision based on the analytic model they had developed. Without the decision point in the operational processes they would have got far less value. They would have gained understanding certainly but they would not have been able to take the thousands of small retention actions that made a difference.