Co-founder and CEO at 10 Senses
What is process mining?
Process mining saves money
Process mining is an analytical technique that is gaining more and more popularity. It enables the achievement of large cost savings by identifying areas of improvement in the functioning of processes. Process efficiency is important especially for large enterprises, which is why many optimisation initiatives are undertaken here. Historically, however, the collection and analysis of process data has been a costly and difficult task. Process mining is a revolution in process analysis because it is a largely automatic technique. The algorithm discovers the process and visualizes it itself. It is an interactive tool that allows you to filter and manipulate the process model created in this way, to calculate and monitor statistics and KPI measures, as well as to look at the process from different angles. Thanks to this, process mining is of great importance for process optimisation activities, such as RPA or lean six sigma initiatives.
Where to use process mining
Process improvement. Improving process efficiency is critical for large companies. Thanks to process mining, it is possible to gain knowledge about their actual execution. This technology not only provides process models but also gives hard data about how exactly the process is performed. These are, for example, the number of activities or their duration. Thus, process mining allows you to identify specific areas for improvement. If there are modifications to the process, then process mining allows you to capture the effect of this change in numbers. These features make process mining tools of great importance to lean six sigma initiatives.
Process audit and compliance. Process mining also allows for a comparison of the formally assumed process flow with its actual execution. With such approach you can verify to what extent the processes are performed in accordance with the procedures. Such audits can provide a lot of important information to people responsible for the risk or compliance areas.
Process automation. One of the challenges with process automation is the fact that we often have only intuitive knowledge of the process before launching a project. As such, it is often hard to reliably estimate the value of individual automation initiatives. Process mining allows you to quickly check the activities carried out in the process and their duration. As a result, it is sensible to do a process mining analysis as a first step in the RPA journey to find out what kind of automation initiatives will bring the greatest return. This use of process mining will be of key importance for managers responsible for RPA.
Process mining offers new possibilities for process analysis
Process mining solutions allow you to analyse a virtually unlimited number of activities in the process. The peculiarity of process mining tools is that they do not require knowledge of the ideal process. It is enough to insert data from the actual execution of the process and the algorithm itself builds the process model based on identifier, timestamp and activity. These tools allow you to identify activities, measure performance time, or distinguish variants of the process execution. To better understand the process, we can, for example, filter using dates, location or more specific variables – e.g. the type of invoice. Another feature is the ability to look at the process at different levels of detail. To achieve this, we change the abstraction level of process model. For example, we can start with a bird’s eye view to see what the process looks like in general, then we can zoom in on a particular element to capture more detail. For a person analysing processes, such a tool provides several possibilities that were basically not available before the appearance of process mining.
Intuitive visualization and process analysis
Data is crucial
To perform a process mining analysis, it is necessary to use data containing at least 3 pieces of information: identifier, timestamp, and activity. Such information can be found in most of the log records from IT systems that are already present in the company. Each additional information is helpful and allows you to deepen the analysis. Process models in process mining are are generated directly from data, hence the quality and availability of this data is of key importance. Therefore, the first stage of process mining projects should be determining data maturity and, if necessary, ensuring that the data is collected, and its quality is high.
3 process mining approaches
There are generally three types of process mining analyses:
Process discovery. The goal is to analyse the process flow, build a model and visualize it. This allows you to capture the intuition of how the process is actually performed during its execution. Thanks to this, it is possible to capture special cases, identify bottlenecks or unnecessary steps in the process.
Conformance checking. If we have template of the process, we can impose it on the actual performance resulting from the data. This allows you to easily capture those activities and variants of the process implementation that are inconsistent with the expected process performance.
Process forecasting. If we already have process models, we can try to forecast its future performance. For example, we can calculate in real time how long it will take to complete the client’s case and how many activities must be performed along the way. This translates into measurable resource planning opportunities.
Talk to our expert
Are you looking for expert skills for your next data project?
Or maybe you need seasoned data scientists to extract value from data?
Fill out the contact form and we will respond as soon as possible.