Lisa Kart, Roy Schulte and Teresa Jones of Gartner recently published Find the Best Approach to Decision Management. It’s a great piece and if you are a Gartner subscriber you should definitely read it and share it around your organization. Even if you are not you should seriously consider getting it from Gartner as it makes a solid case for Decision Management. Obviously I am not going to copy and paste it here but here’s the outline they provide with some commentary on my part about resources here on the blog or at Decision Management Solutions that might help you (Gartner outline in italics, my comments not)
The piece starts off with a nice description of Decision Management. If you want more background on this you could try my books (Smart(Enough) Systems with Neil Raden or the more recent Decision Management Systems) and sign up for our forthcoming intro webinar.
Decision Management consists of three stages:
- Decision Discovery (and modeling)
- Decision Services (construction using technologies like business rules and predictive analytics)
- Decision Analysis (ongoing, continuous improvement)
- Use Systems With Predefined Decision Models to Make Nontrivial, Repeatable Decisions
Gartner’s focus here on non-trivial yet repeatable decisions is key. These largely operational decisions are what drive success in your organization and can be managed, automated and improved. I really liked the strong recommendation to develop a decision model for these kinds of decisions too – check out DecisionsFirst Modeler (our cloud-based, collaborative decision modeling tool) as well as our upcoming decision modeling training and this white paper on Decision Modeling with the new DMN standard.
- Use Decision Support Approaches for Decisions That Require Both Analytics and Human Judgment
I have written before about decision support v decision management and about the role of the OODA loop – check out my most recent book. I would also say that we have had some real success building decision models for non-automated decisions (using the model to position predictive analytics in the decision for instance as well as for designing better dashboards/reports). Many systems are a blend of decision support and decision management.
- Choose Algorithms and Tools According to the Nature of the Decisions to Be Made
- Complex-Event Processing
I really like the emphasis here on a mix of technologies and on bringing them together appropriately. There’s a lot more on this in our freely available Decision Management Systems Platform Technologies report. Once again decision modeling is your friend here as it let’s you see exactly how you can/will mix and match these technologies.
Besides the platform technologies in our report there are lots of pre-packaged decision management applications. One of the richest areas for this is in Marketing – check out this blog series on Marketing Decision Management systems.
Three bits of advice I would add to the recommendations the folks at Gartner include in their paper:
- Use decision models (and the new DMN standard) to specify decision support and decision management systems for any repeatable decisions
- Make decision management and decision management systems part of how you build your application portfolio
- Add critical decision management technologies like business rules and predictive analytics to your platform