“Process mining is an enabler of our digital transformation”

In addition to future technologies such as voice, robotics and AI, ERGO is now also exploring the field of process mining and has even set up a dedicated team for this purpose. Victor Thoma, Head of Process Mining at ERGO, explains how he and his team are working on making processes transparent – and how this benefits customers, business units and ERGO as a whole.

Hello Victor, what exactly is process mining?

Essentially, process mining is a technology that supports us in process management and especially in process optimisation. As a company, we want to constantly improve our processes in the interest of our customers and, for example, reduce the time it takes to pay out a claim. The first step in such an optimisation is to create transparency about the current process flows and, in particular, to analyse the deviations from the ideal process.

In the past, we took a random look at individual cases on the basis of a dossier and tried to reconstruct the process. With many millions of processes per year, of course, you only gain a small insight here – and not full transparency. With process mining, however, we are now able to reconstruct and visualise each individual incident based on the data points in our IT systems. Thus, we can easily find out the reason for deviations. You might compare this to a flash light that allows us to illuminate the deepest corners of our processes.

The individual cases are easily reconstructed using three core data points, which we transfer either via a data extract or an automatic connection to our process mining tool “Celonis”. We only need a unique ID – a claim number – as well as the interesting process activities – such as a bank transfer – and the time when this activity took place for this ID. 

What distinguishes process mining from other data mining methods?

Compared to classic data mining methods, process mining does not start at the data level but at the process level. The initial question is always who looks at which data points for what purpose. Let's take the example of “Next Best Action / Next Best Offer” (NBA/NBO). The point here is to identify an offer that is statistically most likely to meet the customer's current needs. We all know the logic from online shopping with reference à la “Other customers also bought ...”. For this, you would check which topics have been interesting for comparable customers.

However, process mining is not about making visible which offer we make to the customer – but how the offer was processed. For example, was it sent through the intended channel, did the feedback end up in the right department, did the customer have another query and call the call centre? If so, we could perhaps include the answer to this query directly in the offer – and the customer would no longer have to call us. Process mining is therefore explicitly about process-related data to increase our process efficiency and customer satisfaction.

What are the long-term benefits of process mining?

Process mining is an enabler of our digital transformation. Because if we look at processes in detail, there are often further use cases for technologies such as voice, AI or robotics. This indirectly increases their use – and brings us all the closer to our technology goals at ERGO.

In addition, process mining is clearly a first step towards a “data-driven company” – for an enormously broad range of users. Process mining with Celonis is not only impressive because of the new insights, but also because it is easy to use on a broad scale. In the past, individual process indicators could sometimes be requested from IT or had to be laboriously calculated via individual queries. At the end of the initial project, each department is trained by us in the software and then has an almost real-time view of its entire process indicators. In the long run, this will significantly change the way our departments work, as every process change can be checked immediately on the basis of concrete data – and without having to rely on the supply of other departments. The department can thus monitor and constantly optimise its process much more easily and autonomously.

Thank you very much for this interview!

Interview: Ingo Schenk