A little while ago, we hosted a webinar about “Why Decision Management?“. However, including the business case presentation alone seemed somewhat incomplete. Indeed, we also needed to focus on ‘how’. As a result, we created a presentation that covered both aspects.
In this blog post, I will focus on the key takeaways of the presentation. However, I encourage you to watch the video, especially if you are a manager, new to decision management technology.
To begin with, let’s define what we mean by Decision Management.
Decision Management is a Discipline
supported by Technology
for Capturing, Automating and Improving
Large-Volume Transaction Processing
for the purpose of making Decisions
Capture Decision Logic – Why and How
The typical audience for Decision Management tools includes mainly Business Analysts. It happens that people ranging from IT to Business could use these tools. Nevertheless, Business Analysts (BA) remain the main users. Their role is to translate business concepts into business requirement documents. While convenience leads to known entities such as Word and Excel, there is value in leveraging DMN Tools. Not only do they guide BAs through the discovery process, but they also give you a leg up when implementation comes.
When projects do not allow for a discovery phase, or when expertise is solely in the heads of experts, other capability in the decision Management world can lead to an effective elicitation. RedPen mimics the work of the expert making manual decision, and capture business rules in the process. Of course, you will need and want to author business rules in other ways too. For example, one ruleset might be better represented as a Decision Tree. That being said, the compliance team may want to review it as a Decision Graph or a Decision Table. In the end, flexibility in representation and approach is the key to a successful project.
Let’s keep in mind that the purpose of capturing the Decision Logic is to automate processing, and to eventually improve the decisions.
Automate Decisions – Why and How
Each organization has a different internal architecture, different mandates. Decision Management technologies make it easy to deploy your decision logic in your architecture of choice. For one company, it might be flexible REST services. For another, it might be the speed of native components such as Java classes or .NET assemblies. For other, it might be the elasticity of a grid framework. In the end, what matters is that the deployment is easy to integrate in that architecture you have chosen.
Architecture is far from being the only consideration though. You must also consider:
- Runtime Performance in all forms
- Release Management
- Deployment Management
With all that in place, you will benefit from automated decision that reduce manual processing, without the nightmare of managing code.
Improve Decision Performance – Why and How
Once decisions are automated, you can fully appreciate the power of Decision Analytics. Not that we recommend you only consider your Key Performance Indicators (KPIs) at the end of the cycle. Of course not. A significant part of their value appears at authoring time. By defining them early, and checking progress as you add more and more logic, you will be guaranteed better quality business rules.
Furthermore, actual historical or pooled data samples can reveal expected business performance through large scale simulations. Finally, when the final outcome might be delayed, Champion / Challenger experiments allow you to test competing strategies. For example, you might make a marketing offer as a result of your business rules. Yet, the distribution of offers is not enough to anticipate future business performance. The key question is whether or not the customer will accept the offer.
In extreme scenario, like fraud detection or marketing offers, you might want to explore new frontier. We have seem incredible business performance by giving machine learning tools to Business Analyst or fraud Experts. In doing so, they complement the work of Data Scientists, and can produce predictive models very quickly, in those incertain circumstances.
In conclusion, Decision Management technologies offer amazing return on investment for your decisions that happen in large volumes.
Learn more about Decision Management and Sparkling Logic’s SMARTS™ Data-Powered Decision Manager