The Decision Model: Business Rules Grow Up

The world of business rules and business rule management has grown up. Welcome to the world of business decisions—a much more compelling technique with which companies can manage their operational business policies.

Until now business decision analysis and maintenance—the means by which business logic is discovered within current business practices, mined from opaque legacy systems, represented in transparent format and managed as a business asset directly understood by business subject matter experts—has been rather vendor-specific and lacking in rigour and scalability. Many methods are available (e.g., EDM, ABRD and BRS), but all are either lacking in formal structure, subjective and/or poorly supported by tools. Developments of the past few years have changed this.

With the advent of The Decision Model (TDM) and tools (like BiZZDesign, Sapiens DECISION and OpenRules) that support it, an era of new rigour and effectiveness has dawned within enterprise decision management.  TDM, a precise method and framework for expressing business decisions, has addressed many of the flaws of business rules and prompted a slow evolution from business rules to business decisions. We examine these flaws, what TDM can do for you and the promise and power of this approach.

Introduction

So what are business decisions and how can you benefit from them? In this article we cover the potential benefits of the evolution from business rule management to business decisions management using an approach like TDM. If you are new to business rules, you want to understand their many benefits or you want to know the difference between business rules and business decisions, please follow these links. If you interested in understanding what the evolution from rules to decisions can offer, read on…

Drawbacks of Business Rules

For those companies far-sighted enough to see the value of cleanly separating their living business logic from: the heads of their staff who orchestrate manual procedures, the source code of their systems, the definition of their business processes and their (likely out-of-date) requirements documents (and there are many reasons why you should), business rules and business rules management have quickly became a fact of life. For all their considerable strengths however, business rules can have some drawbacks. This, coupled with some poorly designed tools, has given some companies a bad initial experience of business rule management.

Much of the evolution of business rule management systems (BRMS) in the last three years has been driven by the difficulties experienced by rule adopters. Typically these include:

• Growing Pains: there is real difficulty in safely growing rule repositories in excess of 2,000 rules without descending into chaos. Although most BRMSs physically support large rule repositories, they lack any higher level structure to coordinate the relationships between rules and to consolidate these large repositories. Rule-flows and folders just don’t cut it in practice.
• Lack of Coordinated Validation: although many systems can validate individual rules, they frequently cannot validate the mutual consistency of rules in the case that they are used collectively to reach a coordinated conclusion. For this, extensive testing facilities must be manually developed by adopters of business rules.
• IT Focus: many BRMSs, despite their best intentions, invariably tend to move the focus from the business to the technical user. For example: they provide a technical rule language which users need to learn, they support rules with a strong technical flavour, they require knowledge of a programming language to support complex rules or they ensure that only development tools can be used to maintain artifacts like the vocabulary of business terms, rule-flow or fact types. This creates an early (and unnecessary) dependency on IT teams.
• Silo Development: most BRMS focus exclusively on rules and ignore the context of their application. They divorce rules from their process context and make no attempt to connect them to a business process. They also do not associate rules with their business rationale (e.g., business motivations, policy documents), making it harder to judge and impossible to monitor their on-going effectiveness.
• Lack of Agility: most BRMS don’t empower business SMEs and analysts to own the business vocabulary (the names of the terms, the business words, with which rules are built) and form of their rules. This control rests with IT and this means they have to be involved in every significant change. This is an important barrier to agility and undesirable for many other reasons.
• Lack of Business Testing Facilities: a vital part of developing business rules is testing them, an activity in which business SMEs and analysts should have centre stage. I’m not referring to systems integration testing which should be controlled by IT, rather the initial unit and regression testing that occurs as decisions are initially defined and refined. Business SMEs should be actively involved in the testing of their decisions. Testing in many BRMS is very hard to configure and very often this has to be done by IT, which distances business experts from the process.

Business decision approaches, like the Decision Model and DMN and the tools that support them—business decision management systems (BDMSs)—overcome many of these problems.

Overcoming These Issues with Business Decisions

The Decision Model (and the tools that support it) offers three essential ingredients that ‘change the game’:

• A set of structural and semantic principles which establish the  rigorous integrity of a business decision and its inferential dependencies on a hierarchy of rules. This allows it to retain its coherence regardless of size and be validated as a coherent whole;
• A focus on business ownership of decisions, rather than IT;
• A defined and distinct ‘context’ for decisions within the set of business architecture. A well defined integration of decisions with, but separation from: business processes, organisations, rationale, events, data and decision services.
• A standards based approach which means that all decision are documented using the same notation (in this case TDM). This standardization opens the door to multiple tool support.

These work together to provide the following advantages not offered by rules alone:

1. A rigorous structure that binds sets of business rules into a cohesive whole. Just as databases provide referential integrity to enforce consistency, decision modelling supplies inferential integrity to keep decisions totally consistent. Many tools that support the approach offer extensive validation functionality ensuring high levels of decision quality and cohesiveness. This facility also enhances ease of understanding of business decisions, essential for their effective review.
   

   Example Decision Model – Coherent Structure

    

2. Enterprise Scalability. This cohesiveness radically decreases the entropy associated with large rule sets and facilitates much bigger repositories. The explicit dependency relationships between decisions and rules helps to quick assess the impact of any proposed change.
3. A means of establishing traceability between decisions and key business KPIs. The approach recognizes that decisions are not defined in a vacuum, there is a clear and documented association between decisions and: their business motivation, business organisational structures, business process, business events and business data. By explicitly recording which decisions support which business policies (in which process, for which organisation, triggered by which events), we can: record the rationale of each decision, quickly assess the decision impact when policies have to change and build a foundation for measuring the on-going effectiveness of decisions.
4. A mechanism to support the business ownership of decision mining, definition, vocabulary, testing and evolution. Many TDM tools support vocabulary management directly by the business. TDM has enabled mainstream tools that support business vocabulary definition, maintenance and decision execution. Many TDM tools also support business analyst creation of test data, some even automate it. Together these facilities offer unprecedented independence on the part of owners of decision logic to mine, develop and test their decisions without the need for IT involvement (until deployment).
5. A strong business focus, so that decisions can be understood without any technical knowledge.  The best decision tools can be used to promulgate business decisions across an enterprise because of their use of English business terms and decision table format, rather than technical decision or programming languages. This removes the need for specialised technical training and allows more staff to benefit from an explicit statement of company policy. Above all, TDM supports communication of business requirements.

Conclusion

These benefits allow BDMSs to provide a more enhanced safe agility than traditional BRMSs. This allows much a faster time to market for new decisions and less change latency for existing ones. This is strengthened further by allowing businesses to quickly determine which decisions are impacted by an incipient change of business policy—so teams know the impact of any proposed changes more quickly. Tools allow business SME teams to achieve much more—from initial requirements mining and building ‘play’ systems to test ideas, through to development of a complex network of decisions—without the constant need for IT support. The process of business and IT cooperation is greatly streamlined as a result. Lastly they serve as a vehicle for communicating business decisions throughout the enterprise.

This is an exciting time to be in decision management. With the additional facilities offered by the toolset and the increased rigour of the underlying approach, there is much more we can offer customers who need to take control of their business decisions. The benefits are more accessible than ever before.

Already tried using business decisions? Why not judge your business decision maturity with our informal quiz?