Let's take a look at how decisions should be managed, why framing decisions is so important, and how to frame them, in this post.

Businesses make thousands of decisions every day, and a very interesting fact is in many organizations, they mostly (32%) use the rule of thumb or they use an ad-hoc manner (25%) to make the decision.

decision-making study

 

This is relatively an old study done back in 2008, but from the conversation we have with many organizations, we think the stats for today are not that different.

What this means is in an organization, the decisions are not formally managed, they are not identified and structured and therefore, there is no owner of a particular decision. This is quite odd, as opposed to how organizations treat processes. There is a process owner, they formally document and structure processes across the value chains and they recognize processes as an important element of improving and running a business.

Framing Decisions

There is no objection to the importance of decisions. We all “know” the risk associated with bad decisions, however, when it comes to ensuring risks mitigation the approaches are particularly “process” driven. We don’t reserve any objection to the importance of processes however the numbers we shared here and the conversation we have with many business leaders raise the question of why we don’t treat decisions as not only a critical part of the business but also why not structurally and systematically manage them?

Considering the reach of a decision in business, the positive and negative impact of good and bad decisions, and the volumes of decisions being made in organizations every day, it makes sense to manage them as a first-class citizens in organization.

Let’s talk about an example and understand how we can frame a decision in a systematic, structured manner. In financial organizations, “Credit Decision” is how they understand the creditworthiness of a client. Then this decision is used across many processes and different reasons such as

  • determining if a client is eligible for a particular product,
  • increasing the number of credit cards of a client,
  • identifying the right product to a client,
  • …and many other scenarios (e.g. processes)

Let's find out how we can frame a decision such as “Credit Decision” that will consistently create a quality outcome. A decision, as explained before, requires some input which might be in some shapes and forms of information stored inside applications, databases, files, etc.

Decisions are complex, and humans deal with complexities by breaking them down into smaller, understandable, and manageable units. So, let’s apply the same principle here, we are going to break down a complex decision by framing this decision.

If we represent the “Credit Decision” with a rectangle, and it requires two data inputs for “Income” and “Expenses” we can model it as below:

For example, now let’s drill down to more details of the “Income Decision” which requires “Income Consistency” and “Minimum Income Check”. So, we can break it down like this:

And we keep breaking down a complex decision into more individual, simpler decisions until we can build this whole picture on what is the “credit decision” for example in this context:

framing decisions

 

This is called modeling or framing a decision. Meaning we show the components of a decision:

  • Input data
  • Other dependent decisions
  • And relations between them

As you can see here, we clearly defined how a “Credit Decision” is made, however, we have not defined where the inputs are collected. Or we have not specified how each of those smaller units of decision making (e.g. Stability, Frequency, etc.) create the outcomes.

This is a very simple method that enables you to model a complex decision in a very effective way.

Conclusion

Modeling or framing decisions using decomposition techniques allows you to break down a complex decisioning scenario into more understandable and reusable units (e.g. sub-decision, business knowledge, etc.). Also, specifies the relation between those smaller units, meaning, how the outcome of one decision is related to other decisions as an input which you can follow by the direction of the arrows between nodes in the model.

A decision model is not just a graphical representation of a business decision, it is an executable model! In fact it's an artifact representing live decision specification which you can deploy, test, debug, embed and simulate.

Remember framing decisions using these techniques makes them independent of where the data is stored and how you access it, which gives you more flexibility on testing, simulation, deployment, etc. Later you can use the decision model as part of orchestration to connect to databases, files, REST API, and other forms of the data source.

Published January 27th, 2022 at 08:12 am