This question was asked of me last week by Dan, one of our Account Directors.
Dan is a guy who knows Oracle Policy Automation well, he’s seen how OPA allows you to take policy or legislation, turn it into readable business rules that are automatically parsed and accurately applied to a user’s circumstances. Specifically, OPA rules are typically directly attributable to the source legislation and policy.
OPA rules are typically directly attributable to the source legislation and policy
But what happens when part of making that decision, perhaps a single criterion, is open to interpretation, relying on the decision makers own intuition to apply it to an individual’s circumstances? In the absence of legal or policy guidance (which are easily added into Policy Automation rules), experienced decision makers will often have an innate feeling as to whether the person does or does not satisfy a requirement.
The problem is with each decision maker relying on their own perception and subconscious biases to make these judgement calls, you can get different results from different decision makers. Aside from the legal challenges this exposes you to, 41% of customers state different decision makers providing different answers is their biggest pain point when dealing with customer service centers.
But how, as Dan put it, “do you codify gut feel”?
The premise is simple: once decision makers have been in the job for a while, they start to see patterns in behaviour, and additional information, that helps them to better decide how a criterion should be applied. If we can somehow capture this experience in rules we can gain more consistent decision making.
Let’s take an example. Many years ago I worked with a government agency who provided social services payments to people in need. Many of these payments required an applicant to be “habitually resident” in the country. There was some general guidance on how to apply this to applicants who spent the majority of their time in the country or who were serving overseas on military service, but borderline cases were typically left up to the decision maker to look at the evidence provided and make a decision. Asking a room of 8 decision makers, they were all quite comfortable that this seemingly discretionary decision was being made consistently within their department. No cause for concern here. So… I asked… How do you decide if someone is “habitually resident”? The general consensus was that it’s usually pretty obvious – they got a clear feeling of whether or not this criterion is satisfied… but how? What factors do they consider to reach this ‘feeling’?
Decision maker A was the first to answer: “I just look at where they work and where their kids go to school”
Decision maker B added: “I tend to look at where they have purchased property and where their mail is sent”
Decision maker C added: “Yes and activity on their bank accounts is a good indicator”
Decision maker D added: “We received a memo last year recommending that we consider where they and their immediate family spend most of their time”.
Decision maker E added: “If they have joined clubs or other organizations that shows active community involvement, that’s a good sign”
At this point there began to be a few murmurs in the room: “Surely, you have to consider where their kids go to school?” “You mean you don’t look at where they work?”, “No”.
Each decision maker made the decision slightly differently
As we discovered, each decision maker made that decision slightly differently, and they were not comfortable with the inconsistency of factors considered. It became clear to all of us that the same applicant could have very different outcomes depending on whose desk their application form had the fortune of landing on.
A couple of hours of debate ensued. As they talked, we captured their thoughts and the level of weight to apply to each factor in OPA rules on a screen at the front of the room.
By the end of the day, we had a set of rules that they were comfortable with as a starting point. They chose to leave the final decision up to the decision maker, but now they were applying due diligence by using OPA to ensure that the minimum required evidence had been considered and that each aspect of evidence had been properly weighted by the decision maker.
OPA would also provide guidance on the recommended outcome where the evidence pointed overwhelmingly towards a particular result, and record where the decision maker deviated from that recommendation.
Which we all had a much better ‘feeling’ about.
Want to see the end result? This post explores how to model discretion with a similar example: Allowing discretion in OPA rules.