This content is part of the Essential Guide: Get up to speed with BPM news, apps, tools and strategies

Best practices for implementing BRM

BRM is being leveraged by enterprises in order to enhance agility and communication.

Business rule management (BRM) technologies, often characterized as decision management (DM), can play a valuable role in bringing business agility to the enterprise. Leading BRM tools and DM products include IBM Operational Decision Manager, Fair Isaacs Blaze Advisor, Red Hat JBoss Drools, OpenRules, Progress Software Corticon and InRule.

"Business rules management is particularly well-suited for dynamic business environments where agility and consistency are key non-functional requirements," said Edson Tirelli, principal software engineer at Red Hat as well as the Drools project lead. BRM is also a great solution for cases where expressing models using a higher-level language or abstraction is important in order to empower business analysts to author, audit and/or maintain business rules (and processes) themselves.

BRM is widely used in financial and insurance markets, communications, government, healthcare and logistics. Tirelli said that healthcare is one of the most fascinating fields for BRM applications. BRM tools are growing for the corporate systems for healthcare management, embedded devices to monitor patient health and front office patient management.

Find the right balance between BPM and BRM

There is almost always a need for business rules everywhere you see analytics around data mining, predictive analytics and machine learning.
James TaylorCEO, Decision Management Solutions

BRM tools are often used to complement business process management (BPM) tools. BPM tools and methodologies are purposed to improve the efficiency and quality of workflow through a sequence of tasks/activities, said Mark Allen, CTO of decision management at Progress Software. In contrast, BRM tools and methodologies are purposed to improve the efficiency and quality of tasks that involve decision-making. The capabilities complement one another.

Most BPM tools include basic BRM functionality. But it is usually limited to routing or work assignment tasks and not well-suited to automate operational decision-making, said Allen. This limits BPM utility for managing rules around how to adjudicate a claim, or determine the qualifying discount rate.

Make analytics actionable

BRM capabilities show tremendous promise in making analytics and business intelligence more actionable, said James Taylor, CEO of Decision Management Solutions, a business rules consultancy. "There is almost always a need for business rules everywhere you see analytics around data mining, predictive analytics and machine learning."

Leading analytics vendors like SAS, IBM and SAP are starting to wrap business rules around their analytics capabilities that can be embedded into enterprise applications. This makes it much easier to leverage business rules in the front office to improve customer treatment.

Leveraging the cloud and beyond

Taylor sees a tremendous movement towards employing BRM results to the cloud. The shift of using rules for customer treatment is driving enterprise architects to consider new ways to deploy results to the cloud quickly since call center reps and agents are often working from cloud applications.

Many enterprise architects are even looking to deploy the rules themselves to the cloud. Enterprise architects need to weigh the pros and cons of the GRC requirements of keeping the rules behind the firewall and sending the results to the cloud compared with staging the rules themselves in the cloud. Deploying a rules repository in the cloud can provide some benefits in terms of integration with other services and mobile applications. But Taylor does not expect to see most established enterprises moving the rules to the cloud directly because they are sensitive information.

Some enterprises are even starting to push business rules to mobile devices as well. Taylor expects to see BRM pushed out to devices for the Internet of Things to help distribute decision making regarding the collection of data. In these cases, the BRM engine will help to ensure that an enterprise can collect actionable data without overwhelming networks and applications with unnecessary data. This will allow more decision making to be federated through the network.

Find the right balance between IT and business

"It is more important to start with a top-down approach with the decisions to solve rather than rules," said Jacob Feldman, founder and CTO at OpenRules, an open source BRM provider. The best results come when business people are put in charge of writing the specification and business rules themselves and work in concert with technical people to implement the information system. The IT team should in turn focus on integrating the rule system into the back end IT infrastructure. This process requires developing an enterprise architecture that enables the direct execution of business rules.

Many organizations tend to have the business people write rule books that are passed on to programmers to implement. But Feldman cautions that this is a bad approach. When developers are in charge of implementing rules, they are likely to use programming tricks to implement the rules. As a result, the rules can be difficult for business analysts and subject matter experts to understand.

The subject matter experts should also write the test cases for the business rules and decisions, said Feldman. This enables a test driven approach that leads to more maintainable and extendable systems.

It is also important to have a clean contract between IT and business as to who owns the rules. "Defining roles and responsibilities cross-functionally is one of the biggest differentiators between those that see a lot of success and those that don't," said Jim Wray, VP of product management and sales operations at InRule.

This can be a daunting task for many IT departments that fear business people will break something, said Wray. In some cases, it is better to have business people sign off on rules, and in other cases, to make the changes themselves.

Automate for volatility, complexity and volume

One good starting point for implementing BRM lies in identifying decisions with high volatility, plus complexity or volume, said Progress Software's Allen. Volatility refers to the frequency of change of the underlying rules. If the rules never change then hard coding them is not a problem.

Complexity of rules involves the number of factors that go into a given decision and the total number of rules. The more factors and rules, the more difficult it becomes to understand the interrelationships of the logic, and the time to program increases exponentially.

Volume refers to the number of times a given decision is made. Very complex decisions (such as how to route an oil tanker, or whether to approve a commercial loan) may be lower volume, but still high value. Other decisions that are higher volume but lower complexity (such as eBay's seller hold policy) are also very high value.

Start small

It's also a good practice to start with moderate scope, explained Allen. This means decisions with enough complexity to see the value, but small enough to complete within weeks. Some of the different forms of value to measure include cost efficiencies (eliminating manual tasks), reduced risk (improved compliance with policy and regulation), increased revenue (optimized decision-making) and faster change cycles (opportunity cost). IT benefits include reduced development and maintenance cycles.

Introducing BRM technology into an organization is similar to introducing any new technology. "Success can be achieved by having key people with experience involved in the planning and execution of the project," said Red Hat's Tirelli. "Starting small and organically growing to encompass the whole solution also helps to make sure the solution is built on solid foundations."

Next Steps

Learn more about intelligent BPM

Business process management: How it's being shaped by analytics, cloud services

Dig Deeper on BPA and BPM