Process mining: the x-ray machine for business processes

Prof. Dr. Volker Stiehl: To answer this question, I will use a comparison that process mining software manufacturers like to make, namely the comparison with X-ray machines. Just as an X-ray machine makes the invisible visible inside our bodies, process mining software makes the business processes in a company visible.

Process mining shows how the processes run step by step and in which order this happens. Thus, process mining is an approach for analyzing business processes. How does process mining achieve this goal?? By systematically analyzing so-called log entries to create visual flow models of processes. A log entry represents an event that occurs in an IT system.

For example, using an order booking as an example, a log entry could contain the following information: the date and time the order was received, the order number, the products that were ordered, and the customer information such as name and address. This information is typically captured automatically by the job posting software and stored in a log file or database.

Process Mining takes these entries based on already finished processes, sorts them according to date as well as time and in this way obtains a very concrete sequence for the order booking. Now the only question is what companies can do with these visualizations? Essentially, three types of analysis can be performed: the identification of processes, the matching of actual processes with their target behavior, and the enhancement of processes for their optimization.

General reasons for process mining

How to motivate companies to engage in process mining?

Prof. Dr. Volker Stiehl: There are several reasons why companies should look into process mining. My top four reasons are:

  • Compliance: this is probably the most attractive reason for banks and insurance companies. Just think of risk assessment when granting a loan or taking out an occupational disability insurance policy. Process mining can help companies meet compliance requirements at precisely this sensitive point. Both internal and external regulations and guidelines are monitored and possible violations are detected at an early stage.
  • Process optimization: Based on the visualized processes, companies understand much better how they really work and how their processes function in real terms. It is sometimes unbelievable, but in projects I often hear the sentence "Oh, so that's how it works for us!". Process visualizations can then be used to identify bottlenecks and weak points and to optimize processes in order to work more efficiently.
  • Cost reduction: Cost reductions are very closely linked to process optimization. They go hand-in-hand, so to speak: Once the bottlenecks and weak points have been identified, they can be eliminated in a subsequent step. For example, shortening the processing time of applications helps to reduce costs.
  • Competitiveness: In today's digital world, it is eminently important for companies to remain competitive. The onslaught of new competitors in the market in the form of FinTechs (banks) and InsurTechs (insurance companies) is now legendary. Therefore, it is especially important for established companies to strengthen their competitiveness via process mining and the associated process optimizations.

So let's summarize: Companies that don't engage in process mining run the risk of losing competitiveness, increasing costs, and potentially facing compliance issues. It is therefore important that companies recognize the importance of processes and use tools such as process mining to optimize their processes and achieve their goals..

Reasons for banks and insurance companies

Can you say something to banks and insurance companies? Are there any special incentives/problems here?

Prof. Dr. Volker Stiehl: I am certainly not an industry expert, but I see an enormous need for banks and insurance companies in particular to address the issue of process optimization via process mining. Since banks and insurance companies are service industries, their added value lies especially in the optimization of their processes. FinTechs and InsurTechs as the new challengers of established banks/insurance companies are addressing exactly this point by introducing fully automated processes in order to minimize human and thus cost-intensive interactions. This puts significant pressure on established banks and insurance companies.

  • Established banks and insurance companies also find it difficult to keep up with the pace of the new competition because of the IT landscapes they have developed over time. These legacy issues are not a challenge for FinTechs and InsurTechs, as they can work with a modern IT infrastructure right from the start.
  • I would also like to refer again to the previous question. As noted there, banks and insurance companies are subject to strict regulation and high compliance requirements, where close monitoring of processes is necessary to avoid errors or violations. Process mining offers a possibility to gain transparency and control over processes and thus to meet regulatory requirements.
  • Another challenge for banks and insurance companies is customer loyalty and offering personalized products and services, a discipline in which FinTechs and InsurTechs once again feel really at home. Process mining can help to identify weak points in processes and thus to design more customer-friendly and efficient processes. This enables them to stand out from the competition and satisfy their customers.
  • As a final aspect, I would like to address a point that is closely related to increasing digitalization. More digitized processes inevitably mean more digital footprints in the form of the events mentioned at the beginning of this article, which are reflected in log entries. We can now mine this treasure through process mining and analyze the data in such a way that banks and insurance companies gain insights into their processes and their effectiveness. Thus, resources can be better deployed and efficiency gains can be achieved.

Is process mining only now becoming relevant?

Why process mining is only now becoming really relevant? After all, the term and the possibilities have been around for a few years now?

Prof. Dr. Volker Stiehl: You are right: process mining has been around since the end of the 1990s and has been continuously developed since then. So, although the technology has been around for some time, it is only in the last few years that there has been a greater awareness of its importance. This is because many banks and insurance companies are only now, in the age of digital transformation, realizing how important efficient and transparent process design is to the success of their business. FinTechs and InsurTechs have put their finger on the pulse with their lean processes and high level of automation, demonstrating that there is enormous potential for optimizing business processes even in traditional industries such as banks and insurers.

With this in mind, process mining has become more important in recent years. In particular, the ability to automatically identify and visualize processes offers banks and insurance companies the opportunity to improve the efficiency and transparency of their processes and thus gain competitive advantages. All in all, process mining is an important tool for banks and insurance companies to be successful and remain competitive in the age of digital transformation.n.

The first steps

What steps should a company start with if it wants to get involved in process mining?

Prof. Dr. Volker Stiehl: This question can be summarized briefly as follows: Start small and select a suitable process to begin with Process Mining. Collect the relevant data and connect it to an appropriate process mining tool to visualize the process. Then analyze the data and identify bottlenecks and potential for optimization. Then implement concrete measures and check the success regularly.

So much for the plan. In this context, the question of the "suitable process" is not so important immediately catch the eye. So what is a "suitable process"?? A suitable process for getting started with process mining should be clearly defined, measurable and standardized. It should be a process that is performed regularly and where it makes sense to improve the throughput. Another criterion is that the process is sufficiently digitized and the relevant data is available in digital form. A good way is to start with a process that has a high number of human interactions and possible bottlenecks to better show the potential of process mining.

Process mining – the risks

Are there also areas where process mining is counterproductive??

Prof. Dr. Volker Stiehl: Yes, there are areas where process mining can be counterproductive. First of all, process mining is backward-looking and is mainly suitable for optimizing existing processes. It is based on the analysis of historical data to uncover weaknesses and potential for improvement. Therefore, it is less suitable for developing new business models or making radical changes in process design. When it comes to creating new digital business models and processes, process mining can't help directly because it requires already established processes and interactions.

Another problem with process mining is that it can only visualize the steps in a process for which corresponding log entries exist. However, not all systems contribute to the creation of the visualization with the necessary fine granularity through log entries, as not everything is recorded. For this reason, there are natural limitations to using process mining.

Another important aspect is that process mining can only help to gain knowledge about existing processes. The implementation of improved or new processes is then still a separate task. So, companies need to take additional steps to implement the improvements identified based on process mining insights. But what are the implementation options available to companies? In this context, one naturally thinks reflexively of "programming".

But in today's fast-moving digital age, companies simply can't afford to do that anymore. The disadvantages of pure programming of processes are too obvious: too slow, too error-prone, too difficult to maintain, too inflexible, too intransparent. At this point, I have developed a new method for implementing processes called the "process-driven approach" and avoids the mentioned disadvantages. It relies on graphical process models from the beginning and reduces programming significantly. Who is interested in this approach is welcome to contact me at my email address. I consider the methodology to be extremely interesting especially for banks and insurance companies as process-intensive industries.

In this context, another positive aspect of the process-driven approach immediately comes to mind: With the new possibilities of no longer having to program processes, but to execute them on the basis of graphical process models, the importance of process mining might even decrease in the future. If processes are increasingly developed according to the process-driven approach in the future, process mining would no longer be necessary, since so-called process engines (software that can execute graphical process models) can display the real processes directly based on the modeled processes.

These are entirely new developments that are still largely unknown and offer established companies an opportunity not only to react in the era of digital transformation, but also to take the reins of action back into their own hands and shape their respective industries as leaders. So there is still an exciting development ahead of us, especially in the area of process automation based on the process-controlled approach!