James Taylor's Decision Management

Beth Gold-Bernstein wrote a piece on Business Activity Monitoring called Speeding Toward Business Visibility. She identified four distinct kinds of BAM solutions:

• Provide real-time monitoring of business events related to Key Performance Indicators (KPIs);
• Provide analytical capabilities and management dashboards;
• Provide complex event correlation;
• Provide predictive capabilities

Now the first two strike me as extensions of Business Intelligence to a more real-time, dashboard-monitoring kind of approach. The others, however, start to move into decisioning. Beth went on to say that "Companies are looking for ways to become more responsive" and this comment made me wonder about the first wave of BAM products. I think they have a problem, what I call the "so what" problem. There is a tendency to assume that monitoring, especially if I can make it more real-time and easier to consume with a dashboard, is somehow an inherently good thing. To this I sometimes say "So what"? If I cannot change my behavior in response then so what? I must make decisions based on this new, real-time, easy to read information for there to be any value in it.

Interestingly Beth goes on to say:

"overwhelmingly the most important was the ability to make the company more responsive and move from reactive to predictive mode - to be able to anticipate problems before they become problems.
This capability requires real-time and predictive capabilities such as neural networks and inference engines."

So there we go - you must be able to become more responsive and more predictive. This is the "what" in "so what". As Beth notes you need to use predictive analytics to help identify trends and to do so not in a report that a person can see but in an executable format that a system can use. Likewise you need an inference engine or rules engine to turn the information you have assembled, and the predictions you have added, into actions - "If this is true and this is true and this is likely to be true in the future then do this and this and this".

Beth concludes with some comments on what's important:

"the most important technical capabilities ... complex event monitoring and reporting was the most important capability (31%), and management dashboards was second (27%). Predictive capabilities and analytics were tied for third with 11% each."

So this ability to process events, potentially complex ones, and (presumably) do something is key. This is why effective BAM needs decision management - you must be able to automate and manage the decisions that are impacted by the data that comes from monitoring these activities.

There's more on this on my other blog here and here.