When people hear the words data mining, they nowadays have an idea of what it means. We often define data mining as a process of analyzing data from several perspectives and summarizing it into useful information. With support from this information, we can then make decisions that affect the success of a company. However, even when data mining is familiar to people, process mining still seems to be a new topic for many. Very often, I encounter questions like, “What is process mining? How does it work?”.
Process Mining vs. Data Mining
On a general level, we aim to do the same with process mining as with data mining. The goal is to analyze data from different perspectives and summarize it into useful information for making business decisions.
Combining Business Process Management with Data Mining
But this time, the context is the business processes of an organization. In other words, business process mining is a combination of business process management and data mining. We take the data that exists in the information systems of a company and use that to visualize the real-life execution of the company’s processes. Almost all IT systems store data in databases and create logs. Process mining refers to these logs as "event data" required for process mining analysis.
Why Process Mining?
Why not just stick to traditional BPM? Because process bottlenecks are hard to uncover by hosting BPM and process mapping workshops. People have a gut feeling of what could be wrong or inefficient, but they lack fact-based proof. They need the data to back their assumptions, and this is where process mining comes to the rescue.
Replacing Opinions with Facts
One of the main goals in process mining is to be able to see the big picture of a company’s business processes and still be able to drill down to the root causes of deviations, bottlenecks, or process variations. With process mining, we don’t have to settle for averages but dig out the reasons behind unwanted behavior. To bring this closer to organizations and industries, I’ll give a couple of examples.
Where can Process Mining be Applied?
Process mining is suitable for application in any industry and process, where the data required is available in the organization. Most companies that have an IT system supporting their businesses have this data.
Applying Process Mining to Specific Industries
In the manufacturing industry, timely and accurate delivery to a customer is the ultimate goal. If a company has several factories in different regions, there usually are differences between the reliability of deliveries. It’s quite easy to see that they exist but harder to understand exactly where or why they happen. Also, very often, people with the loudest voice gets the most attention regardless of whether their problems are the biggest. With process mining, we can compare the performance of different regions down to individual process steps, including duration, cost, the person performing the step, and many more. All event data available in the systems is suitable for use in process mining.
Banking and Financial
In the financial sector, it is crucial to follow the rules and regulations and to be able to prove you have done so. By using the event data from the systems, we can visualize even individual cases as a process flow. We can also show how often deviations and variations occur and what has been the reason causing this non-conformance. With process mining, we can both prove our actions but also pinpoint the improvement needs in the business processes.
Applying Process Mining to Specific Processes
Typically, people have some idea how the process is running - or how it is designed to run - but the reality revealed with data can prove out to be quite different.
Common findings are changes, for example, in order-to-cash and purchase-to-pay processes. For varied reasons, the cases go through several changes that affect efficiency. Often we find also several deliveries for a single order or several invoices.
Order-to-Cash (O2C) Process
In the order-to-cash process, end-to-end lead times or lead times between process steps deserve some scrutiny - check, for example, the lead time from delivery to invoice creation.
Purchase-to-Pay (P2P) Process
In the purchase-to-pay process, the challenges can start already at the beginning of the process when a PO is created without a purchase requisition or a frame agreement.
Only in 63% of the cases is a purchase requisition created. Why is that? Are there differences
between regions, for example?
Visualize Your Processes for Analyses and Improvements
To summarize: in process mining, we use the event data in the company's IT systems to bring insights into the company’s business operations. The insights are provided by visualizing the data in process context with process flowcharts and creating analyses that give information on needed improvements and deepen the understanding of what is going on in the business processes. We can show both the big picture as well as the detailed ground-floor view on process execution.
Get started with QPR's process mining tool QPR ProcessAnalyzer.