Process Mining Basics

Quick Introduction to Process Mining for Beginners

The following is a quick introduction to Process Mining and meant as an overview for novices rather than a detailed guide for advanced users. Hence, this section focuses on practical explanation and does not claim to be complete or cover all facets of Process Mining. If you are interested in further information, feel free to reach out via the contact form.


  1. Fundamentals: What is a Business Process?
  2. What is Process Mining?
  3. What are the resulting benefits?

At first, we will go over some of the fundamentals that Process Mining is building on, which is an illustrative process and the corresponding process model. Afterwards we will dive into Process Mining and the resulting benefits for organizations.

1. Fundamentals: What is a Business Process?

Before answering the question, what Process Mining is, we first need to understand the basics that Process Mining is building on: Processes!

A process represents a collection of activities which are executed in a certain order to achieve a particular goal. Our example process, illustrated in Figure 1, describes a simple ordering process at a take-away restaurant, which you are probably familiar with. Let's take a closer look at the individual steps of this process.

© Source:
Figure 1: Example ordering process

The process starts with you calling the restaurant to order food and drinks for dinner. A person at the restaurant will pick up the phone and take your order. After that, your address and preferred payment method are noted. Only when both activities are completed and you hang up the phone, the process moves on to the subsequent steps. The chef prepares your burger and another person will grab your drink in parallel, which is followed by wrapping your order for the delivery. The final step of the process is the handover of the food. It can either be delivered to your home or you can pick it up at the restaurant yourself.

The circles in Figure 1 define the start and end point of the process. The steps in between, such as “Note address”, are called activities. These are linked via connections and gateways.

It is important to emphasize that process models are always idealized representations and in reality the sequence may vary significantly. Our example process is quite simple, but it might be the case that the drink is grabbed only after the burger is prepared and wrapped. So even in this simplistic sequence, there is room for deviation. You can now imagine how many variations exist in real-world processes. Such discrepancies are normal and might be meaningful, but often variations are an indication for inefficiencies. To collect and capture such process variants is a crucial part of Process Mining. But more about this in the section about benefits of Process Mining.

  • Activity: square with flattened corner
  • Connection: arrow
  • Gateway: AND (activities are parallel),  XOR (only one or the other activity)

There exist many other types of gateways, but these are the most prominent.

2. What is Process Mining?

Now that we have a common understanding of a process, it’s time to look at Process Mining.

We start with a definition:

“The idea of process mining is to discover, monitor and improve real processes by extracting knowledge from event logs readily available in today’s systems.” [Aalst et. al (2012): Process Mining Manifesto]
There are several terminologies which might be new to you, so we are going through them together. First there is the Event Log.

You can imagine the Event Log like a table, where process data from different IT systems are captured. The required minimal process data are an ID (e.g. an invoice number), an activity name (e.g. a short description of the task, e.g. “create invoice”) and the timestamp (i.e. the time when the task happened). You can see an example Event Log corresponding to our order process in Figure 2.

© Source:
Figure 2: Corresponding Event Log to our Ordering-Process with the minimum of required information

As we learned in the previous section, pretty much every organization runs processes. Whether it is paying an invoice, handling requests or dealing with orders, processes are everywhere. With the increasing degree of digitalization, information systems support in many areas. These IT systems, e.g. Enterprise Resource Planning (ERP), Customer Relationship Management (CRM) or Supplier Relationship Management (SRM), are constantly tracking and storing data of performed activities related to business processes. That means, in most of today's organizations, there is event data already available for the use of Process Mining. This event data can be exported into Event Logs. They are the input for Process Mining software, which uses sophisticated algorithms to extract valuable information. Such software traces back the digital footprints of every executed activity. This is how the underlying process can be rebuilt, monitored and finally improved. These insights help to identify bottlenecks, observe process deviations or redesigning the existing process. You see, because Event Logs are the foundation of Process Mining, their data quality is key for a successful implementation.  In Figure 3 you can see the practice-oriented principle.

Figure 3: Process Mining in practice

These were the basics about the application of Process Mining, now we dive a little deeper into detail. Figure 4 presents the theoretical approach.

As previously cut, there are different stages of Process Mining. The earliest stage is discovery. For this purpose, the performed process is rebuilt from the raw data in the Event Log. So, from the event data, algorithms return a process model.

As it resonated from the previous paragraph, there may be tremendous deviations from the to-be process and the as-is process detected via discovery techniques. Consequently, the second stage is conformance checking. In conformance checking real-world processes are compared with a reference model and deviations are observed. Third stage is enhancement. The underlying process is redesigned to optimize certain parameters like costs, lead time or quality.

Figure 4: Theory of Process Mining

This was a lot of information, so it’s time for a recap: The goal of Process Mining is to gain an extensive understanding of the underlying process structure. Therefore, it uses the digital footprints that are captured in various IT systems whenever a process step is executed. For example, when a person at a company pays an invoice, the IT system captures a lot of data, among others, the activity that was executed (paying the invoice), the number of the invoice and the time when this process step was executed. From this data, which is sometimes spread across various IT systems, a so-called Event Log is built for every process. This Event Log is essentially a large table that contains data such as time, process step and e.g., invoice number. Based on this Event Log, a Process Mining software can then build a process model, which illustrates all the different variables that have been recorded for a certain process. Such a process model, enables the dive into the process, e.g., identifying potential bottlenecks or unnecessary process steps. It can also be used to compare a to-be-model of the process (how it should be) with the as-if of the process (how it happens in reality) and identify deviations.

3. What are the resulting benefits?

Increase transparency

In companies a lot of processes run at the same time. With increasing digitalization and use of different IT systems, the process landscape is getting more and more complex and the underlying as-is processes are hidden and oftentimes a black-box. Process Mining is a technique to investigate this black-box, to discover the actual process and potential inefficiencies.

Improve customer experience

Process Mining is a technique which helps customers to really understand the underlying processes. Therefore, it is possible to identify and resolve existing bottlenecks. Through the improved reaction time due to process transparency, the company’s performance can be significantly elevated.

Enable data-driven decision making

For a profound decision making it is key to have quality data at hand. Through Process Mining, data availability and powerful visualization and analysis tools are enabling management to stay on target.

Support digitalization and automation

Since Process Mining techniques derive how distinct cases are handled and how operational decisions are made, automation potentials can be leveraged.

Identify non-compliant processes

The tremendous gap between as-is and to-be processes in a lot of organizations can be identified and reworked. Dashboards and alarms can be installed to reveal compliance issues in real-time.

Fraunhofer-specific advantages

Fraunhofer offers an unique approach to enable your organization to make use of this broad range of benefits. Interest aroused?
Please visit the ‘Our Offer’ tab for further information.