Removing decisioning from the SDLC

Syndicated from ebizQ

My friends at IDIOM had a great tweet today – @Intelligentform said:

#Decisioning objective:nothing less than the removal of decision management from the SDLC – automated decisions should be managed as content

I retweeted it (I’m @jamet123) but I thought it warranted a longer blog post about why this is a good idea and how to do it. Why first.

Decisions are at the heart of many (most) processes and any but the most passive of systems. Any time you make a choice, select from alternatives or act you make a decision. Processes are full of these decisions – is this customer eligible, what price should I offer, what kind of product is suitable, is this transaction fraudulent. But this prevalence of decisions is not enough to justify removing them from the SDLC. They need to be removed from the traditional SDLC because they change all the time. Or, at least, you need them to change all the time if you want to ensure they remain compliant, effective and competitive. After all regulations change, competitors change their behavior, markets and consumer expectations shift, policies change and much, much more. And this high rate of change means that a typical software SDLC just won’t work – it takes too long to make a change to decision making. This is what drives many application development backlogs – changes to decision logic that are being forced through the SDLC.

As to how to remove decisioning from the SDLC that is a fairly straightforward process – Decision Management:

• Identify operational decisions
  Especially those that are complex, have complex interactions, change often or require business domain knowledge to understand.
• Develop decision services
  Reusable, stand-alone components that answer the question what decision should be made.
• Conduct ongoing decision analysis
  To ensure that the effectiveness of decisions is monitored, that decision making approaches are challenged with new alternatives and that the results of decision making feedback into improving how future decisions are made.

The construction and management of decision services is key, and is something about which I have written before.