The ability to make well-informed decisions is a cornerstone of success, especially in the dynamic and complex world of modern business operations.
Making decisions is a huge part of conducting any business, no matter how big or small it is. Some of these decisions are strategic and have a massive impact on the future of the entire organisation.
On the other hand, some decisions are more routine, repeatable, and are made almost every day during regular business operations. Nevertheless, these everyday decisions still carry great weight and importance and, due to their frequency, are essential in keeping the business processes efficient and effective.
However, with the vast amount of data involved, many different possibilities, and increased demands for process optimisation, making these sorts of decisions and making sure that they are the right ones can be exceptionally challenging.
One of the main responsibilities of business analysis is to help organisations improve their decision-making process, decipher complexities, and drive effective choices that will help companies maintain their operations at an optimum level.
Probably the best way to go about this is by using decision modelling, a potent technique that helps analysts unravel the intricacies of decision-making within an organisation and identify the key steps of this process, as well as develop a systematic approach to understanding and organising data so that important operational decisions can be made easily, quickly, and accurately.
Table of Contents
What is Decision Modelling?
Decision modelling is the process of creating a structured and, typically, visual representation of how the decisions are made within an organisation.
By using this technique, it’s possible to break down complex decision-making processes into understandable and, perhaps more importantly, manageable components.
Decision models created through this process serve as visual aids helping all involved stakeholders, including analysts and key decision-makers, to comprehend all the important factors, business rules, and considerations that impact choices within an organisation.
As it helps capture the logic behind these choices, decision modelling improves transparency, consistency, and collaboration in the decision-making, and helps automate the entire process.
Within any organisation, there are dozens of critical decisions made on a daily basis. Most of these decisions need to be made quickly and accurately, as they will often determine how profitable a certain business operation will be and contribute to its overall success.
Decision modelling serves as a tool that can help organisations forecast the potential possibilities brought by particular actions. Plus, the visual nature of this approach makes it much easier to understand how organisational knowledge and data are merged to make a specific business decision.
As a business analysis technique
, decision modelling is used for the representation of the process of making both simple and complex decisions.
Decisions that are simpler and more straightforward are typically represented by a single decision table or tree to show how a set of business rules established in the organisation work on a set of relevant data to produce a decision.
When it comes to more complex decisions, decision modelling helps break them into a number of simpler, individual components (decisions), depicting how those smaller parts combine to create a more elaborate and detailed decision.
This is possible because decision models can decompose any information needed to make even the simplest decision and describe that sub-decision in terms of business rules (both definitional and behavioural) that lead to it being made.
Different Types of Decision Models
There are several different approaches to decision modelling that yield different kinds of models and cater to different aspects of decision-making.
The most frequently used models are decision table and decision tree, with the former more frequently employed, typically to represent all the rules needed to make an atomic decision, More complex decisions will often require combining numerous atomic decisions into a network and using dependency and requirements notations.
No matter the approach, the decision model will involve three key elements: decision, information, and knowledge.
Decision Tables
A decision table is a condensed and tabular representation of all possible decision inputs and their corresponding outputs. It clearly displays a set of business rules determining possible outcomes among the number that are available. So, every decision table will feature the following key components:
- Conditions or Inputs – These are variables or factors that impact the decision-making process. Basically, they’re the representation of the information available when making a choice. The table contains one or more columns with conditions corresponding to a specific data element.
- Rules – Each row represents a rule, or, more specifically, corresponds to a specific combination of input conditions. Rules specify the relationships between the input or conditions and the decision outcome. Therefore, when all the conditions for a certain rule are true for a specific set of data, the action (outcome) for that rule is selected.
- Action or Outcome – Different combinations of input conditions will result in potential actions or outcomes (decisions). Each rule in the table corresponds to an action to be taken based on the specific set of input conditions.
Decision Trees
Decision trees are another way to represent a set of business rules. They are used to prevent decisions and their potential consequences in a branching structure.
Due to their intuitive nature, decision trees are particularly useful when illustrating sequential decisions and their associated potential probabilities.
As its name says, this model resembles a tree-like structure, with each path on a tree leaf node representing a single business rule, each tree level corresponding to a specific data element, and downstream branches illustrating different conditions that must be true to move down that branch.
As each decision point branches into various alternatives, the decision tree clearly depicts different available choices and potential outcomes associated with them.
These available outcomes are shown on the leaf nodes on the bottom or far right of the tree (depending on the visual representation) and they’re selected based on the values of data elements at the branching nodes.
Decision Requirements Diagram (DRD)
A decision requirements diagram is typically used to showcase the relationship between different decisions and information and knowledge supporting those decisions in case of more complex decision-making processes, It provides a holistic view of how different atomic decisions interrelate. A typical decision requirements diagram will feature the following key elements:
- Decisions – These are the core nodes of the decision requirements diagram, Decisions are represented as rectangles capturing a specific choice based on a set of inputs and selecting from a set of outcomes after applying predefined business rules.
- Input Data – Represented as ovals, input data parts of DRD encompass the information needed to make informed choices. They represent the data and various variables that influence the decision.
- Business Knowledge Models – On the diagram business knowledge models are shown as rectangles with their angles cut off and encapsulating the underlying logic behind the decision. This includes sets of business rules, decision trees, decision tables, and analytical methodologies that guide the decision-making process.
- Knowledge Sources – As Decision-making seldom occurs in isolation, DRDs also feature knowledge sources, shown as a document and representing external repositories of information and expertise that contribute to informed choices. These sources can include databases, expert opinions, regulations, and other resources that may influence the outcome.
Developing a Decision Model
Decision modelling, regardless of the particular model used, usually involves several key steps:
- Identifying the Decision – Of course, before developing a specific model, it’s important to know the problem at hand or define the specific decision that needs to be modelled.
- Gathering Information – This step involves collecting relevant data, rules, and knowledge associated with the decision the model is developed for.
- Selecting the Model Type – Depending on the nature and complexity of the decision, it’s necessary to select the appropriate type of decision model.
- Building the Model – Once all of the above steps are completed, it’s time to construct the model by mapping out inputs, business rules, decision models, and outcomes.
- Validation and Testing – Every model should be validated by comparing its results to expert opinions or historical data. Then, it can be tested with various scenarios.
- Refining and Optimisation – Based on the received feedback and new data, the model should be continuously refined to improve its effectiveness and accuracy.
Benefits and Strengths of Decision Modelling
Decision modelling offers numerous advantages to organisations and to business analysts using this technique on a project they’re involved with.
- Effective use of Business Rules – Decision modelling provides business analysts with a critical framework for business rules analysis, their implementation, and an understanding of how they influence the desired outcomes. Instead of focusing on business rules themselves which typically leads to an expansive set of rules that are difficult to manage and coordinate, decision modelling offers a more holistic view and opportunity to add some structure to complex sets of business rules by observing them in relationship to decisions they lead to. Decision models also help manage and understand a large number of business rules by grouping the rules based on the decisions they correspond to. This helps improve both business rules and decision-making processes.
- Simplifying Processes and Workflow – Identifying and modelling decisions independent of processes keep those processes simpler and prevent clutter. By explicitly calling out decisions, decision modelling makes processes smarter and more agile.
- Clarifying Analytic Requirements – By identifying decisions for which the analytics will be required, decision modelling helps analysts focus on the use of analytics and define how the data requirements will support those decisions. This improves and improves the efficiency of business intelligence processes, data mining, and predictive analytics. Additionally, decision modelling can help prioritise data and quickly identify which information is really needed to support key business decisions. This way, the organisation can make data provisioning more accurate and avoid the delays caused by the late discovery of vital data items.
- Improving Communication – One of the greatest strengths of decision modelling is that it can capture and communicate business decisions in a way that is clear and easy to understand not only for subject matter experts but also for all other relevant stakeholders. Decision models serve as a common language that fosters and facilitates communication and collaboration even among stakeholders with different backgrounds.
- Better Change Management – Using decision models helps analysts and organisations to adapt more quickly and easily to changes. Better change management is possible because decision modelling allows organisations to modify their decision logic without affecting the entire business process. In addition, decision modelling, by explicitly stating dependencies, provides a better understanding of underlying data and how decisions rely on one another.
Limitations of Decision Modelling
Even though decision modelling brings significant advantages, it’s still important to acknowledge that this technique has certain limitations:
- Complexity – When modelling business processes that contain decisions, the use of decision modelling means adding a second diagram, which may add unnecessary complexity to already simple processes and decisions.
- Rules Left Out – As it only requires capturing rules for known decisions, decision modelling may leave out some of the rules that are not related to the specific decision.
- Current State Bias – Decision modelling may cause stakeholders to believe that the decision models within the organisation have been standardised and lead them into decision-making based only on the current state.
- Difficulties Securing Buy-In – While it fosters better communication across different departments, by crossing organisational boundaries Decision Modelling crosses might make it more difficult to acquire sign-off from all relevant stakeholders.
- Human Element – Decision models can overlook human intuition and disregard the subjective factors that often influence decision-making.
What is Decision Modelling Notation (DMN)?
Decision Model and Notation (DMN) is a standard for decision modelling administered by Object Management Group (OMG), an international consortium of technology standards. and widely adopted across many industries.
As a modelling language and notation for precise specification of business decisions and rules, DMN focuses on standardising the ways decision models should be represented, providing a medium for sharing expertise more widely and making process management systems more accessible to everyone.
Basically, the decision model and notation (DMN) standard helps convert complex decision-making processes into easily readable diagrams.
By standardising decision diagrams and automating the underlying logic, DMN also makes business processes more replicable in different companies or in different roles within the same organisation.
So, DMN not only improves business efficiency and reduces the risk of human error, but also ensures that decision models are interchangeable across the organisation and, often, across different companies.
As a decision management standard, DMN is designed to complement and work along the two other widely used OMG process improvement standards – BPMN (supporting process management) and CMMN (supporting case management), and many organisations do indeed require all three models, especially in the case of more complex business rules.
By using the DMN standard, decision models can be constructed on three levels, and correctly implementing this standard typically requires going through all three of them.
- Decision Requirements – This is the notation for decision requirements diagrams, which illustrate the more complex decisions made in the business domain. So, this level is composed of elements of the diagram: decisions, input data, business knowledge models, and knowledge sources. The decisions are visually presented along with their dependencies on input data, business knowledge, and on each other.
- Boxed Expressions (Decision Tables) – This is a more flexible notation that allows various components of decision logic to be illustrated graphically. Unambiguous notations allow decision tables to be created in a manner that is clear, convenient, and provides a commonly understood way to express business rules. This type of decision representation contains inputs, conditions, and outputs for every decision.
- Decision Logic – DMN uses FEEL (Friendly Enough Expression Language)
to define decision logic or the conditions in the table so the decisions can be executed. By using executable expressions with formally defined semantics, FEEL defines structured logic, externally defined logic, calculations, and simple data structures.
Benefits of Decision Modelling Notation (DMN)
The adoption of the DMN standard in decision modelling offers several important benefits:
- Standardisation – DMN provides a consistent notation for decision management and modelling, helping promote clarity and understanding among stakeholders. In addition, it’s supported by multiple software products, so the organisation is not dependent on a particular vendor. Plus, the standard is owned not by an enterprise, but by an institution (OMG) and is compatible with other widely used standards from the same organisation. On top of all this, DMN brings interoperability, meaning that it can often be integrated with other business analysis tools and technologies.
- Business – IT Alignment – DMN allows decisions to be modelled and executed using the same language and notation. This means that businesses can model decisions and rules that lead to them in easy-to-read tables and that those tables can later be executed by a decision engine. So, DMN allows for better alignment between the business and IT sides in the decision-making process and a better understanding between business analysts and developers, which provides an improved ability to quickly and efficiently react to changes. Through DMN diagrams even the more complex decision logic can easily be communicated among diverse stakeholders
- Compliance and Auditing – DMN also helps in compliance by providing a clear audit trail of decision-making processes.