James Taylor's Decision Management

Over on the decision strategist there was an interesting post - Can Decision Systems Help Us Make Better Decisions? This was a thoughtful and interesting response but I am still going to have to disagree with a couple of points. Firstly the post assets that automation of decisions has not yet arrived. Yet many industries, especially those that must make risk-centric decisions, have long used automation to circumvent errors and bias. The use of rules and analytics to automate credit decisions, insurance underwriting and fraud detection is well established. The reality is that these systems make far fewer errors, are much faster and show less bias than the manual systems they replaced. It is true, however, that we have not reached the point where systems can do this for all decisions (though we get closer all the time). The post goes on to say:

The answer can be systematically determined based on several variables to find the maximum customer satisfaction for the lowest cost.The problem is that these systems remain inflexible.

These systems need not remain inflexible in terms of change over time (systems built with business rules are singificantly easier to change than people-based systems using policy manuals for isntance) and they need not exclude the skills of people. They might, for instance, expect the staff member using the system to enter some assessment of the customer's mood. The post makes this point, talking about decision systems losing "out on the subtle nuances that real people, with their years of experience interacting with other humans, can capitalize on". However decisions must also be made by websites and kiosks, devices with no person at all (thus requiring decision automation) and many decisions must also be made in circumstances where staff members are poorly paid, prone to bias or have high turnover. In those circumstances I think it is an open issue as to how much use the staff's input would be.
One of the post's last points - that "customers hate them" - is the most compelling. People do hate arbitrary, fixed systems and do prefer to think that the person to whom they talk has discretion and will listen to them. Yet they also want to self-serve more and expect to deal with someone who can get things done, not refer them to someone else. Both of these requiring systems that make decisions without people. Decision systems do work, will continue to work and are only going to expand.