Process Mining Camp: Fireside Chat with Léonard Studer

As a warm-up for Process Mining Camp, we asked some of the speakers for an up-front interview. Previously, we already spoke with Bram, Frank, Mieke, and Wim.

Léonard Studer

Today, we give you the last interview, which is with Léonard Studer, a Project Manager at the City of Lausanne.

Léonard was setting up an internal control system using process mining at the City of Lausanne. At camp, he will talk about the benefits of process mining in a resource-constrained environment.

Interview with Léonard

Anne: Can you still remember where and when you first heard about process mining? What exactly caught your attention and fascinated you about the topic?

Léonard: Well… let me think about it! In 2010, I led a project to set up an Internal Control System (ICS). Among other tasks, in this project, we had to conduct interviews to model the actual processes (before identifying risks that we wanted to control internally).

These interviews were not very efficient in elucidating the processes for two reasons: First, people usually tell you about their idealized model of the process and quite never of the actual process. Next, in my context, people were not very knowledgeable in process matters. This meant that I had to teach them about processes, which was consuming too much resources in the context of the ICS project.

Facing this situation, I started to look into alternative ways to model business processes. As I practiced data mining for more than 15 years, one day I wondered if data mining techniques would help… So, I googled the web with words such as “data mining” and “business process” which lead me to the sites linked to Wil van der Aalst and his group and… here I am. By the way, the ICS project itself had been stopped but not the process mining activity!

For me, apart from the “applied data mining” aspect, one thing that fascinates me in process mining is discovering how people relate to each other in their everyday work. Who is the central person in a operational process? Usually, this is not the boss! Who should delegate work? Who calls whom to complete a task? What happens when a key person is away?

Anne: You mentioned some of the typical problems of manual process discovery sessions. It’s quite hard to get an objective picture, and this is where process mining can help. I have always been eager to point out that process mining is not a replacement for talking to people, but that it can objectify and speed up the as-is discovery. How was your experience when you presented your process mining results to people who actually work in these processes?

Léonard: My first study was on some Human Resources data logs. I presented the methods of process mining to a project manager who is responsible to set up a new ERP for the HR department. She was interested in these process mining methods but…

Almost all meanings are included in this “but…” and reality bites!

This new ERP implied a lot of changes for HR people. The project manager acknowledged that the process mining results could help her to better fine-tune the new processes. However, like with every noteworthy change, some people are uneasy and keep negative concerns with the changes introduced. This is what happened with the project introducing the new ERP. When I presented the process mining results, the level of uneasiness of concerned people was high (too high) and the project manager didn’t want to introduce another destabilizing factor. So, we agreed to not yet use the process mining results.

So, my HR process mining results remain “academic” which is not a bad thing per se. It allows me to better position process mining methods, to raise their visibility, and to work on lowering the potential fears that such methods could trigger.

I also presented the Process Mining methods and results to some colleagues who are in charge of the software development team, and to some colleagues who offer an internal counseling service on organizational matters (optimization, change management, new tools like BI or RM etc). Here, the interest was undoubtedly much more positive. My presentation of process mining methods even triggered some changes in one major project in our organization.

Anne: This sounds like there can be quite some politics involved. For other people, who are trying to introduce process mining and run into similar situations, how would you recommend they approach the situation to get the best possible outcome?

Léonard: Would you welcome somebody who you don’t know and who tells you things that will be intrusive in your work? As your process mining project or study will induce changes and could be rather intrusive, it can trigger anger and fear as a reaction. Don’t forget that process mining is about people and their everyday tasks.

With this in mind, I would recommend to consider a process mining project or study in three different steps.

  1. A first step would be political and relational in nature. You have to be allowed in your right to investigate, hence you should concentrate on gaining a strong sponsorship at the “C” management level. On the relational side, you should try to reduce your distance to the people carrying out the activities of the investigated process. Either the sponsor or the process actors will tell you a lot about the process that is not necessarly included in the audit logs - things like why it is like it is, or to what extent you can change something.

  2. The second step will be about the technical process mining work using wonderful tools such as Disco. For a part of this step, you’ll work alone, freely investigating the data logs to “get an idea” of what is inside - you aim to learn what the process is about. For another part of this second step, you should work closely with a few experts of the process investigated. You should co-construct or co-discover the results that will be presented later - this could be a rather short but intensive session.

  3. In the third step, you will concentrate—together with the process experts and the sponsor—on what and how to present the salient results you got to the involved process actors. You should concentrate to how the investigated process actors will react when they will be presented with the results.

These three steps are not fully independant. For exemple, during the first step, it could be a good tactical move to have a first quick look to the audit logs in order to lower the risk of not being able to complete the process mining investigation, or to get strong arguments to convince the “C” level guy to sponsor you. During the second step, you could discover sensitive facts about which you should inform your sponsor. In the third step, you could be enforced to better understand something you discovered.

It is also very useful to initially set up an ethical chart for your process mining study - What are the privacy concerns? What will you do about suspected unlawful activities? Who will access the results? etc.

As I said before: Don’t forget that process mining is about people and their everyday tasks.

Anne: Absolutely! This is very important to keep in mind. Thanks a lot for these interesting insights and recommendations, Léonard. We see you on Monday!

Come to Process Mining Camp 2012!

Would you like to hear more from Léonard about his experiences? Are you interested in sharing first-hand knowledge with fellow process miners? Register now to reserve your seat at Process Mining Camp on 4 June in Eindhoven. Tickets are free, and there are just a few left…

Anne Rozinat

Anne Rozinat

Market, customers, and everything else

Anne knows how to mine a process like no other. She has conducted a large number of process mining projects with companies such as Philips Healthcare, Océ, ASML, Philips Consumer Lifestyle, and many others.