As more and more companies adopt Business Process Management Systems (BPMS) to automate key business processes, they are also seeing the value of enhancing this technology with Business Rules Management Systems (BRMS). When processes rely on complex business decisions, it makes more and more sense to use business rules to automate these decisions. Not only are business rules better for delivering clarity and agility in decision-making, the use of a BRMS allows rules to be shared across processes, across BPMSs and between a BPMS and other implementation platforms.
As more and more companies adopt Business Process Management Systems (BPMS) to automate key business processes, they are also seeing the value of enhancing this technology with Business Rules Management Systems (BRMS). When processes rely on complex business decisions, it makes more and more sense to use business rules to automate these decisions. Not only are business rules better for delivering clarity and agility in decision-making, the use of a BRMS allows rules to be shared across processes, across BPMSs and between a BPMS and other implementation platforms. Examples of decisions include what offer to make, whether to accept or decline, whether to underwrite etc. If a business takes this approach it will have a centralized location to update decisions and the actions to be taken as a result – a rules repository, if you will.
Many companies are also talking about analytics in business processes. Gartner, for example, has released some thought-provoking stuff on how the combination of business intelligence and a business process platform might be used to create an environment in which processes are self-configuring and driven by customers/transactions. To Gartner “Analytical Process Controlling (APC) enables an organization to investigate process execution realities, evaluate their business impact, and ultimately improve and optimize its business processes.” So if business rules are used to automate decisions in a process, how can analytics be integrated and to how close to this vision can we get?
Before considering this perhaps we should outline the various kinds of analytics. There is not much agreement on terms but essentially you get three main kinds:
- Reporting or business analytics – looking at all past information and reporting what happened, making it easy to understand the data
- Data mining analytics – beginning to investigate what the trends in the data and business are. Adding an element of human discovery and making it possible to predict the future of the business
- Predictive analytics – determining what the decisions and actions should be in the future for a specific transaction or customer based on analysis of past results
Let’s assume you have automated critical decisions within a process using business rules. One common example that can illustrate this point is loan originations. Clearly there is a process but equally clearly there is a critical business decision embedded within it – should we offer credit to this person on these terms? To completely automate this decision we need to automate this decision, and that requires a different approach from those used in the past. Enterprise Decision Management (EDM) is an approach to combining business rules and analytics to better automate operational decisions. Using EDM to bring both business rules and analytics to bear on a decision can improve the precision, consistency and agility of these decisions. Let’s consider how analytics can complement business rules.
The first category of analytics, reporting, can help you understand your process and the impacts of the decisions you make but it is not really “bringing analytics into process”. Equally providing dashboards and other reporting to those conducting manual reviews may be helpful but it is still not really bringing analytics into the process. You could use business rule logging to record the critical decisions within a process instance and then use reporting analytics to investigate your process execution – like what loans were made. Still, you are only using analytics to understand the process, not to drive it. This is, sadly, where many people stop when thinking about analytics in a process.
The first way you can really impact the process analytically is to use data mining to find rules. Data mining for rules means using statistical analysis techniques to analyze historical data and derive thresholds and segmentation rules that are statistically significant – in our example you might use these to ensure that the various routes (high risk, low risk, standard say) through the origination process are used for meaningful subsets of those applying for loans. Now my routing and process control is starting to be analytically driven or fact-based not purely judgment-based.
The use of predictive analytics offers more opportunity for analytically driving the process. Embedding predictions of likely future behavior can help make business decisions where risk or opportunity is uncertain. In the loan example, I might predict the risk that an application is fraudulent and refer risky ones for manual investigation or for a cross-check that is too slow or expensive to use in every case. To do this I apply new rules that take advantage of this prediction and drive the process in a new direction. Now my data, and my predictions of the future based on it, are really starting to drive my decisions and thus my process.
This combination of rules and analytics is often referred to as Enterprise Decision Management and, as an approach, is highly complementary to Business Process Management. Adopting EDM to manage your decisions and then embedding these automated, analytically-enhanced decisions into your processes, can help you use your data to improve your processes in new and very profitable ways. Essentially this approach allows you to build processes that are driven by the transactions themselves – the data in the transaction, or metadata attached to it, would determine what scores the models would generate and the combination of scores and data would determine which rules fired and that would decide which steps to take to complete the transaction. This let’s you “invert” the process and have it flow from the customer to the organization. An example of this would be an origination process where the data entered by the customer is used to drive models and rules that determine which products are available and what additional data and steps are required.
Analytics can be used to do more than just report on a process, you just need to have thought about decisions separate from processes.
