Current Trends in Process Mining – Insights from Jan

Process mining is transforming the way organizations understand and optimize their operations. In this insightful interview, we explore the latest trends, challenges, and opportunities shaping the field of process mining today with Prof. Dr. Jan Mendling from Humboldt University of Berlin.

Our discussion highlights:

• Why process mining is essential and how it creates tangible value for businesses.
• The evolution of process mining toward object-centered and complex approaches.
• Key weaknesses in current solutions and the growing need for modern, robust alternatives.
• The challenges of working with classical event-logs and how they impact organisational efficiency.
• Predictions for the future of process mining vendors and the market, shedding light on upcoming innovations. Process mining has established itself as an important approach to process management.

Process mining has established itself as an important approach to process management. But why are we talking about processes at all?

Processes are the key to better productivity. In the 1980s, companies invested massively in information technology for the first time. Economists were surprised that this did not result in any major productivity gains. This productivity paradox was resolved in the 1990s. The processes had been forgotten. This means that IT investments only lead to an increase in productivity if new processes or existing processes are improved on the basis of this IT. In order to achieve such improvements, you first need to understand the existing processes in detail.

In your research activities, you have been working intensively for years on the use of process mining approaches in organizations. What does a modern application have to offer these days in order to provide sustainable benefits?

Let’s consider that under four points. Firstly, a process mining system must help to connect different data sources quickly and easily. Secondly, the representation of the process must illustrate complex dependencies in an understandable way. Thirdly, the process knowledge of employees must be integrated into the analysis. Fourthly, simple questions can be asked with the help of generative artificial intelligence.

From your experience: What are the biggest difficulties for companies in successfully introducing and using process mining?

Process mining requires the connection of data, which often requires someone from IT, the expertise of the specialist department and the methodological competence of process experts. Consultants can help with expertise and technical challenges. The specialist expertise has to be available in-house.

How do the weaknesses of traditional process mining applications manifest in practice and what impact can this have on decision-making?

Traditional process mining methods create so-called sequence graphs from the event data. These are inaccurate and show complexity that does not exist in reality. This can lead to bottlenecks being shown that are not actually bottlenecks. Modern methods can do better.

Are these the reasons why object-centered and multi-dimensional process mining are currently on everyone’s lips?

Yes, that is correct. There are various new approaches in this area. However, some of them are also quite complicated. This is a challenge for the specialist department. It is worth comparing different approaches.

Is this trend already visible among process mining providers in the market?

Many providers are faced with innovation dilemma. It is largely new providers who are implementing these ideas and highly financially strong providers who are supplementing their traditional approaches.

Is it possible to use process mining without event logs? Are there still challenges that stand in the way?

Yes, it is possible and it also makes a lot of sense. In the past, companies often threw away a lot of detail from the rich data in their source systems in order to generate reductionist event logs. During analysis, they then tried to reconstruct facts that were explicitly mapped in the original source data. Many new approaches are therefore based on the idea of transferring detailed data into knowledge graphs and using them to provide flexible and complex evaluations.

Final question: What advice would you give companies when choosing a process mining application?

I come back to the first observation. Investments in IT pay off when you look at the corresponding processes. For process mining, this means that you have to look at the decision-making processes. Which people in which areas of the company need to continuously develop their processes? With the help of a quick proof-of-value project, a steep learning curve can be established in the specialist department. This experience should then be used to aim for improvements in the management processes and select the appropriate tool. 

Did you know that you can subscribe our LinkedIn newsletter here in order to get the latest updates directly via Email?