Olga Kacprzak, manager of process mining CoE at Poland’s fifth largest universal banking group mBank, speaks with PEX Network about her process mining journey and offers tips and advice for any practitioners looking to begin their own journey. Kacprzak boasts more than 10 years of experience working with project management in areas such as IT projects and system implementations.
Applying a data-driven approach to process mining
Kacprzak notes that initially mBank applied traditional Lean and Six Sigma approaches to process optimization, but came to understand that a data-driven, process mining approach would be more effective. The organization drew this conclusion after noticing that staff did not have full visibility over processes and were only knowledgeable about the processes they were involved in.
“It is not always possible for staff to understand processes, especially when you have large, complex ones, sometimes including multiple departments and up to more than 100 activities,” Kacprzak explains. “So it felt like we were only basing our process optimization decisions on intuition and we wanted to base them on the data, so we started to learn and practice the process mining approach.”
Kacprzak notes that the key advantage of this approach and the reason she advocates process mining for other organizations is the competitive advantage offered by enhanced process visibility. She notes that while traditional Lean and Six Sigma approaches are effective, they do not offer visibility over what processes look like in reality. While she advises that brands use them, she notes that more is necessary to secure that competitive advantage and process mining represents a higher level of process optimization.
Three steps to successful process mining
Drawing on her experience at mBank, Kacprzak offers her three top tips for organizations to ensure process mining is implemented successfully.
Data science is crucial
Involving the data science team is a top priority, as many of the organizations she has discussed process mining with did not do so, and ultimately failed. They struggled with data issues and often did not have the information necessary to build accurate event logs. For this reason, Kacprzak stresses the importance of including data scientists to assist with data pre-processing.
Be patient
In the field of process mining, patience is a virtue as planning and building a successful initiative takes a lot of time and effort. For most organizations it is simply not possible to achieve big process mining wins in just a few months and practitioners should remain patient for the results to come.
Know your processes
It is important for organizations to know their processes inside out and know what systems they reside in before they decide to optimize them. This will help them know where to begin, as Kacprzak explains.
“If there are workflow systems, it is much easier to start process optimization there because these systems are naturally developed for processes, so you have the better data there,” she advises. “With transactional systems, it is much more difficult to extract data from them for the process mining approach." It is possible, especially when using task mining approach or tools which record what people do on computers, but it is not straightforward and complicated. She adds: "If you want your organization to see the results of process mining it is better to start with workflow systems because it is easier."