A Panel of Experts: How to Introduce The Decision Model to Your Organization

(Ground Up and Top Down)

With the rapid adoption of The Decision Model, the most frequently asked question is: “How do I convince my organization to try it and eventually adopt it as a standard?”

Two related questions from two different perspectives are:

• Do I have to find a way to introduce The Decision Model from the top down?

• Can I introduce The Decision Model from the ground up?

The top down approach is often the one taken by Enterprise Architect organizations or Enterprise Centers of Excellence. The ground up approach is the one taken by business analysts or project managers on a particular project.

Good News!

The good news is that success stories prove that both approaches work and can even occur in parallel. In fact, in December 2011, a group of four experts participated in a panel specifically to address these questions. Each panelist is a successful evangelist for The Decision Model in their organizational sphere of influence. By evangelist, we mean someone who spreads knowledge within an organization about The Decision Model with a goal of organizational conversion or adoption.

The experts had used The Decision Model for complex business logic as well as other uses, such as Data Quality Logic. As panelists, they shared the challenges and opportunities served by their decision model projects along with practical lessons learned.

The Panel

The audience included business analysts, directors, business people, project managers, and enterprise architects. Apparently, 40-50% of the audience was familiar with The Decision Model while the rest were just learning about it. A full recording of the panel is here.

One panelist introduced The Decision Model from the ground up. Another introduced it from the top down. The commonality is that all have experience creating decision models, but the target technology for automating them differed from legacy systems, to a home-grown engine, to multiple target commercial BRMS environments. The variety of target technologies proves that The Decision Model is not only technology-independent, but it deploys easily and performs well in various environments. Of particular interest was a surprise related to commercial BRMS products and The Decision Model (more on this surprise below)!

This month’s column summarizes the most useful parts of the panel. It begins with an introduction to the panelists. From here, it presentstheir answers to five important questions.

The Panelists

There were four panelists, chosen to represent various industries as well as differing experiences with The Decision Model on real-world projects. All four panelists are often active in the LinkedIN Group called The Decision Model.

Nick Broom

Nick has over ten years of experience in business analysis for major financial and legal institutions and in healthcare regulation. He specializes in business process design using BPMN, development of business rules and logical data modeling.Below he talks about introducing and actively promoting the usage and benefits of The Decision Model within a multi-national insurer.

David Pedersen

David has over thirty years of business and IT experience in a variety of industries and is possibly the most experienced Business Decision Architect in the world. He is a contributing author of the book The Decision Model: A Business Logic Framework Linking Business and Technology. He is the author of many articles on Decision Modeling, including “Better, Faster, Cheaper Part II – The Decision Model Meets Data Quality Head On” at Modern Analyst.

Gil Segal

Gil has over twenty-three years of experience working with Business Rules and Business Decisions technologies. Gil is the Director of Enterprise Solutions at Sapiens Americas and the subject matter expert on Sapiens’ business decision management system, Sapiens DECISION.

Panelist #4

Panelist #4 has over seventeen years of experience in defining IT and Business Architecture, Business Management, Business Operation Management and Decisions and Rules Management. He was instrumental in designing architecture and infrastructure for business rules. He was also responsible for managing and directing a business rules operations group. He assisted in leading efforts to build a Business Decision Management System and manages several other critical decision modeling Projects.

Question 1: What was your or your organization’s motivation for considering The Decision Model

Panelist #4 discovered that The Decision Model is easy for both business and IT to understand. It is rigorous with corresponding rules for correctness. The idea that decision models can be implemented in any technology was attractive. It also offered a comprehensive solution to a large set of rules and logic, such as 50k rules.

Gil was familiar with The Decision Model from its early days. However, he became an enthusiastic advocate when he sat through a validation session of 40 senior business people and IT. Not only did everyoneunderstandeven the complex models, all played an enthusiastic role in validating them for correctness and completeness. As a business rule vendor, he had heard vendors talk about business user empowerment but no one really achieved it and not through to deployment.

Dave had many years of experience with the “insanity” of business rules and logic getting lost in systems(regardless of target technology) and the inability to reasonably test them. He saw The Decision Model as a comprehensive way for business rule capture, integrity validation, and testing, and for never losing the logic in technology again.

Nick witnessed the “light bulb moment” when he put decision models in front of both business and IT people and they asked a lot of questions about them. Of particular interest was that their questions were notabout the notation such as “What is this?” but were immediately business-focused such as “Why is thisvalid condition for this conclusion?” The clarity of requirements and immediate positive response across audiences of business and IT was rewarding. An advantage was that the goal is not just business signoff but clarity for deployment.

Question 2: What kind of experience have you had with The Decision Model

Nick became aware of The Decision Model when reading a business process article from Bruce Silver.  Nick introduced The Decision Model as part of a requirements approach on a project, thereby evangelizing The Decision Model from the ground up.

Gil was familiar with The Decision Model before its publication. Subsequently he now plays a key role in the development of the Sapiens DECISION product to support decision modeling and governance. As a result, he is familiar with decision models from various clients in various industries.

Dave has created decision models for a wide range of uses (e.g., business logic and data quality logic) and across many industries such as the mortgage industry, technology, and healthcare insurance.

Panelist #4 participated as his organization introduced The Decision Model from the topdown, establishing a center of excellencefor formal governance over decision models and deployment. He has been a lead on many decision model projects and is a certified decision modeler .[1]

Question 3: Can you provide insight into the sizes and complexities of your decision models?

Nick’s first project produced 14 Rule Families, 40 Fact Types.

Gil’s smallest decision model consisted of about 3 Rule Families with a few rows and a few columns. The largest was about 50 Rule Families, 120 fact types, averaging 5-7 rows each.

Dave’s simplest decision models were single Rule Families. Medium sized decision models were 30-40 Rule Families, some with 8-10 Fact Types and 40-50 rows each. The super large decision models were more than 450 Rule Families, but the Rule Families themselves were small, often only 1 to 4 rows. In his experience, the size and complexity of decision models varies characteristically by industry.

Panelist#4 worked on one project delivering 50 decision models which turned into 400-500 views of decision models. There were 2000-3000 Rule Families, with 20,000 rows in total.

Question 4: In what technologies has your organization implemented decision models?

Nick worked with legacy systems, taking an approach from requirements with decision models to COBOL.

Gil worked with decision models created in Sapiens DECISION and deployed to an engine internal to DECISION and to two other external commercial BRMS products. This is how he discovered a surprise related to commercial BRMS products and The Decision Model. The surprise is the common misconception is that business logic needs a Rete engine. This turns out not to be true because,inference is integral to decision models, soRete is less important. Some clients turn off the Rete algorithm and performance is excellent.

Dave turned his first decision models over to programmers, not for commercial BRMS automation. Even so, the decision models reduced programming time by 50%. Another project turned decision models into homegrown code, but also deployed them into a BRMS. Even without an automated tool, testing decision models was a significant improvement over previous approaches, reducing weeks to days.

Panelist #4’s decision models have been deployed in a commercial BRMS as well as in a home-grown engine. There was one gap between business and IT so they created an interface by which IT consumes various forms of output, such as destined for a commercial BRMS. With proper formatting, decision models are exported in a standard form and imported directly into a commercial BRMS, for example.

Question 5: What are the most important lessons to share with others?

Panelist #4 warns that it is very easy to start creating decision models. However, don’t lose sight of creating the process models carefully denoting where decision models execute.

Gil advises that organizations can start creating decision models in MS/Visio and MS/Excel with success. Nevertheless, as the quantity of logic grows or as decision models proliferate, MS/Visio and MS/Excel don’t scale - they are difficult to maintain. At that point, organizations need an integrated repository for the artifacts.

Dave confirms that after the first decision modeling experience within an organization, the reduction in cost, increase in productivity and agility catches attention. Be prepared for the increasing interest in The Decision Model.

Nick admits that it is easy to get caught up in the various details of the decision model principles. However, the most important task is to introduce decision models to business people and work with them step-by-step in completing and validating them. When this happens, decision modeling becomes self-sustaining.

Wrap up

The Decision Model made its public debut through our book in 2009. In the past 3 years, many organizations have adopted it for single projects and for entire enterprises. The panel of these four experts covered both paths to decision model adoption. It also provides insights for readers looking to get started or seeking confirmation of their experiences to date.

For reference, Table 1 summarizes the input from panel members.

Table 1: Summary of Panelist Answers

Authors: Barbara von Halle and Larry Goldberg of Knowledge Partners International, LLC (KPI)

Larry Goldberg is Managing Partner of Knowledge Partners International, LLC (KPI), has over thirty years of experience in building technology based companies on three continents, and in which the focus was rules-based technologies and applications. Commercial applications in which he played a primary architectural role include such diverse domains as healthcare, supply chain, and property & casualty insurance.

Barbara von Halle is Managing Partner of Knowledge Partners International, LLC (KPI). She is co-inventor of the Decision Model and co-author of The Decision Model: A Business Logic Framework Linking Business and Technology published by Auerbach Publications/Taylor and Francis LLC 2009.
Larry and Barb can be found at www.TheDecisionModel.com.

[1] Certification requires training, experience, passing a written test, and passing an oral presentation. See www.kpiusa.com