
Business Process Model and Notation (BPMN) is widely recognized as an important document. de facto Standard in business process modelling. It is however an extended time to improve the level of sophistication of practice in business process modeling. Most business analysts make lots of errors, which make their BPMN models more complex and difficult to comprehend and maintain. This is why it is crucial to be aware of the most common mistakes that we call anti-patterns And how to eliminate them from affecting the best practices . Practitioners learn best through an experiential approach. In this article, we’ll look at a variety of real-world business real-time process management examples that illustrate the most common patterns in business process modeling. We will also demonstrate how you can improve the examples using best techniques.
The business process examples presented are specified at an analysis levels, i.e. without the properties that provide technical details for automation of processes. To make things simpler it is recommended that only a few of BPMN elements are utilized and is generally enough for business analysis.
Anti-Pattern # 1: Inconsistent Naming
The first thing that must be examined should be reviewed is BPMN elements names. It is commonplace that modelers employ nouns to name processes or employ a different names for element of the exact kind in various places. Many philosophers have said that “the way we use influences our thinking”. This applies particularly to the BPMN contexts and it’s quite common that models misuse BPMN components that are reflected in inconsistent the naming. Table 1 outlines some common patterns in naming, and their patterns that are best practices.
Table 1.– Some Naming Patterns and Best Methods
Anti-Pattern | Best Practice |
---|---|
Noun-based activity names – signifies that the component is an incident, data object or process area, as opposed to an activity. | Strong verb with domain-specific nouns – is a way of achieving a specific purpose after working. |
Gateway identified as an activity means that a gateway is an activity, and that is what decides on the selection. | The gateway that is not named – it’s simply a branching element that doesn’t perform any function, therefore it shouldn’t be named (except to refer to it). |
“and/or” in the activity name “and/or” in the activity’s name can indicate multiple activities being recorded within one activity. | No conjunctions in names – raise name abstraction level or split into two subsequent/alternative activities. |
Long name for activity – signifies that specifics of the an activity are highlighted rather than the main goal; it also orients diagram toward textual documents. | Name short and documents The name should be a clear indication of the purpose and the details of the activity are recorded in the form of comments or documents. |
For a skilled business process modeler, the task of figuring out names can appear to be simple. In reality, however, there are naming problems that are inconsistent in nearly every company which uses BPMN as a standard for describing business processes. A typical business process model that includes anti-patterns as shown in Table 1 as well as then refactoring it is shown in Figure 1.

We have only looked at a handful of most frequent naming problems – in real life, you could also encounter problems with the inconsistent names of data objects, events participants, messages and so on. For companies that are using BPMN as a standard, it’s strongly recommended to create the naming guidelines and to conduct periodic reviews of BPMN models to ensure these guidelines are being followed.
Anti-Pattern #2 Big Process Diagrams
Another important aspect to be reviewed is the complex process diagrams. Although BPMN facilitates modeling processes at various dimensions, many professionals are accustomed to “everything on the same page” style. This reduces understanding and makes it hard to comprehend and analyse the process. Psychologists have concluded that the most effective amount of visual elements to concentrate on in a single diagram is 5+2. In reality, this rule is not often observed. We believe it’s very practical to increase the number to 7+4, however, after consulting with various organizations about the business model, we’ve witnessed extreme instances of diagrams for business processes that have up to 400 (!) tasks. Large diagrams can make it difficult to study and identify the problems which means the benefit of visualization is diminished. As shown in Figure 3 an illustration of a huge diagram is filled with inconsistent issues, including the absence in confirmation seminar activity that is in a similar way to the existing cancellation seminaractivity as well as a loop that does not end for registering participants following the decision Deadline and inconsistent amounts of detail when announcing the seminars (the four initial tasks) and for cancelling the seminar (one compressing sub-process) and there is no clear distinction between early and late registration , and other. But, it’s difficult to identify such problems when you’re analyzing an extensive and complex business process diagram . You are pleased that you successfully read and understood the diagram and are not unsure of the accuracy of the diagram.

One of the most effective practices that helps get rid of complicated diagrams that hold all of the vital information about the business process involves using multiple layers of details. Figure 3 illustrates an improvement to the model that is shown in Figure 2 which makes use of subprocesses, with additional diagrams that are assigned to each one. With this approach it is possible to decompose an extremely large (and often inconsistent!) business process into a straightforward and consistent system that is structured in a variety of levels of depth. A business procedure with three levels of details based on the principle of having at least 10 tasks in one level can comprise 103 = 1000 tasks.

Anti-Pattern #3 The Inconsistent Utilization of Gateways
Gateways are crucial because they determine the branching of the business process flow based on the data or an coming event(s). They also establish an entry point for the event that ends the subprocess for the process outside. In reality, we encounter numerous issues the use of gateways. Sometimes, gateways are not used or multiple flows are able to go through one activity. It can affect comprehension and readability. Our opinion is that the common practice of using a query as an identifier for gateways, and using Yes/No on branching sequence flows is not correct as it makes it difficult to connect the gateway’s names to the names for the end events in a previous sub-process (Yes and No aren’t appropriate names for final events). We also observe an atypical misperception of the best way to use an event-based gateways as well as workflow patterns that relate to it. The following table lists three commonly used patterns of gateways that are not a good idea are discussed along with the appropriate best practices. Figure 4 shows parts of a process before and after applying the most effective methods.
Table 2 – Common Gateway Usage Anti-Patterns as well as Best Methods
Anti-Pattern | Best Practice |
---|---|
Multiple outgoing and inbound sequence flow This is difficult to comprehend how many flows need to exit or enter. | Always make use of gateways to branch/merge to improve the clarity of diagrams and clearly identifies those points of control. |
Gateway based on events that has an outgoing sequence flow not connected to the event that can be confusing when alternative sequence flow should be used. | Use the Deferred Choice model – all sequence flows following an event-based gateway should be linked to events. Use timer events to simulate scenarios where the expected event does not occur. |
Gateway is not synchronized with the preceding subprocess and is inconsistent between subprocess information and its use within a larger process. | Apply the Internal Business Error pattern with an synchronized end/branch name – making it easy to consistently make use of gateways and sub-processes. |

Anti-Pattern #4 Unconcordant Utilization of Events
Events are a fundamental element of BPMN Every process starts with a beginning event, is followed by one or more events at the end and could react to several intermediate events occurring throughout the course of. Even though events themselves aren’t too difficult, deciding the level of process to place an event that initiates the process or selecting the proper scope for an intermediate event can be incorrectly done. The table below present two typical anti-patterns associated using events, as well as the best ways to improve their effectiveness. Figure 5 provides the example for a process fragment that has anti-patterns, and after modifications based on best techniques.
Table 3 – The typical Events, Event Usage Patterns, and Best Methods
Anti-Pattern | Best Practice |
---|---|
Event outside and inside The process of outside and inside repetition results in it being redundant. A formal interpretation requires the event to occur twice. | Initiating event that is part of de-scription for processIn this way, it’s simpler to read a diagram and comprehend why the sub-process needs to be completed. |
Repetition of events Repeating events makes dia-gram more complicated and requires maintenance. | A subprocess that has an connected events that allows you to clearly define the nature that an occasion. |

Anti-Pattern #5: Poor Diagram Layout
Not least, in the real world there are a lot of issues regarding the layout of business process diagrams. You will notice numerous anti-patterns, such as different sizes of similar elements, irregular space between them, lengthy crossed lines and bended lines as well as other. These anti-patterns within layouts don’t alter the essence of the business process , but they make a diagram difficult to comprehend and read. Figure 6 illustrates an average layout that includes a number of these patterns. The diagram is able to be modified to conform to the most effective techniques of layout as illustrated in Figure 2 which was previously used to provide a variety of detail. The most important problem found in the figure is known as “slalom”– the lack of a consistent flow direction, which in turn makes it extremely difficult to distinguish between the primary and alternate scenarios.
In Western society the flow of a typical sequence should be directed either top-down or from left to right. The preferred direction is top down. the first direction since it helps to manage the using space. The activity’s name text flows naturally from left to right, and the is then arranged in a top-down direction. The absence of a clear direction for flow is also a sign that the modeler may not have did not capture all the details all at once. The ideal approach is to focus on the primary ‘happy-day first, and then consider other scenarios later.

Summary
We have examined five real-world instances that present original business process model examples that suffer from certain anti-patterns. We have also suggested best practices to eliminate the anti-patterns, and modified the original models to create higher-quality Business process model that conform with the most effective practices. The cases discussed here are based on uncoordinated naming, big process diagrams and gateways that are not used in a consistent manner as well as irregular use of events and poor layout of diagrams. We believe that a case basis approach to analyzing common anti-patterns and solving them is the most effective method to instruct business best practices for process modeling. The anti-patterns presented are not meant to be considered an exhaustive list but as an example of a beginning initiative. We hope that this article will stimulate further research and publication on improving BPMN model in both depth (other instances) and in depth (more specific information) directions.