Top 5 Custom Visuals for Process Mining in Power BI
Author: Katharina Laumann, PR Manager and Editor at PAF
Process Mining is turning into the gold standard for intelligent process improvement. Using Microsoft Power BI as a basis of Process Mining is a logical conclusion for any company that uses the digital workplace Office 365. Power BI comes with built in visuals that help to visualize data but for Process Mining you often have very specific needs for visualization and utilization of your data. That’s where Custom Visuals play an important role. They are designed to meet those specific needs. So without further ado, here are our picks of the top 5 Custom Visuals for Process Mining in Power BI:
PAFnow Process Explorer
The PAFnow Process Explorer automatically visualizes your business process in a process flow. This helps you to evaluate and understand your as-is process performance including all paths and branches.
The PAFnow Process Explorer also enables you to filter for individual process steps or paths to gain deeper insights. There is a huge number of filter options to meet your specific needs.
There is also an integrated conformance check for process variants and various layout options in the PAFnow Process Explorer which enable you to analyze your data in as much detail as possible. For example, you can view the entire process flow, or a side-by-side comparison of individual process flows based on various attributes such as country or supplier.
Why we developed it:
- Automatic visualization for easy exploration
- Multiple views on process flows for great detail
- Great starting point for any process analysis
- Core of Process Mining with countless options to extract value from data
Get all the latest news about PAFnow and Power BI, find out what events and webinars are coming up, get helpful videos, etc. >> Register here
PAFnow Root Cause Analyzer
The PAFnow Root Cause Analyzer (RCA) helps you to identify the most likely root causes of behaviors, patterns, or problems in your process. The PAFnow RCA only searches for root causes when you want it to. Users activate the visual, for example by clicking on parts of a chart that they are curious about. The PAFnow RCA will then look for the driving forces behind the numbers in that specific part of the chart.
The obvious choice is to use the PAFnow RCA to find and eliminate the causes of problems and bottlenecks in your process. But it can also be used to find out why a specific process is going particularly well. You only change what needs changes while you also know what you can do to improve processes that are doing well but could do even better.
In that way the PAFnow RCA is an important tool when it comes to making informed business decisions, setting priorities for improvisation measures, and applying changes to your process.
Why we developed it:
- No more guessing games when it comes to anomalies
- Ranking of the most likely root causes makes it easy to take action
- Great deployment options to maximize your process optimization
- Stop wasting money on measures that aren’t needed
Power BI Sankey Chart
Sankey Chart helps you to visualize and understand the relationship of various datapoints in your process between a particular source and a responding destination entity. For example, you can see which activities happen in which order and how often this pattern occurs.
In its core, the Sankey Chart is a flow chart of values with different visual cues. For example, color and size of different nodes and paths in the flow chart indicate the amount of value that moved there.
In Process Mining this means that you can, for example, look at the amount of money that is moving to specific vendors and in return compare it to the number of items you receive from those vendors. You could even look at how resources are managed in your company and find ways to save materials or time.
Why we like it:
- Cross-filtering and highlighting options to compare the different routes from source to destination
- Great to look at the relationship between datapoints in your process
- A simplistic view of movement
- Easy to start
Power BI Decomposition Tree
The Decomposition Tree can be used either for root cause analysis by visualizing the contribution of each category to the overall result, or for simulations and predictions that support decision making. It allows you to visualize and explore data across multiple dimensions by automatically aggregating the data.
The Decomposition Tree enables you to decompose or break down a group to see its individual members and how they contribute to a selected measure. An example of a group is “number of sales”, where the individual members (or categories) are, for example, product category, product subcategory, or country/region.
In addition, the Decomposition Tree includes Artificial Intelligence (AI) capabilities: you can query it to find the next dimension to decompose based on specific criteria. For example, you can explain why sales are low in a particular subcategory. If you select “low value” in the AI function, the next dimension will be added to explain why the value is low (e.g. for geological reasons, so region/country would be the next level).
Why we like it:
- Drill down feature to understand contributions of individual categories
- Flexibility in choosing dimensions
- Cross-filtering of data and between visuals
- AI support to explain values
Power BI Box and Whisker Chart
The Box and Whisker Chart visualizes a dataset in a five-number summary (mean, median, quartiles, min/max). This is an interesting visual because it provides insights that remain hidden when we only look at summary statistics. Those can suggest that everything is fine although there actually are things that need optimization.
For example, you may look at the average value of your data points such as the lead time of a process cycle and think everything looks fine. But there might be outliers in your process that you didn’t see which negatively impact your overall performance. That’s where the Box-and-Whiskers chart can help you.
The five-number summary shows outliers, the volume of data points or clusters of data points that can have a negative effect on your process. You can use this a basis to derive necessary optimization measures which you didn’t know of before.
Why we like it:
- On-point five-number summary of important datapoints
- Includes minimum and maximum
- Easy to spot outliers
- Good customization options