Process Mining: Interacting with the Flow Chart
This article describes how to read the flow chart as well as how to interact with the visual
Last updated 2 days ago
Reading the Flow Chart
▶ Start and End ■ are fixed boxes marking the beginning and end of the process.
Each box (node) is one Activity. Its shade shows how busy it is: the darker/more colored the box, the higher its case volume relative to the busiest step (set the color in Format pane, see section 3).
The number in the top-right of each box is the case count and percentage of total cases that passed through that step.
If some cases repeat a step (rework/looping back), you'll see a small "↺ %" figure showing how often that happens.
Arrows (edges) between boxes show the flow from one step to another:
Thicker, more solid arrows = more cases take that path.
The label on each arrow shows the case count and the average (or median) time that transition takes.
Arrows colored in red shades highlight transitions that take unusually long compared to the rest of the process, these are your bottleneck candidates.
Hover over any box or arrow to see a quick summary tooltip.
Click a box or arrow to select it and dim everything unrelated, this makes it easy to trace a single path. Click empty space to clear the selection.
💡Hold Ctrl/Cmd and click multiple boxes to select several activities at once.
Scroll to zoom, click and drag to pan around the diagram. Use "Reset zoom" (next to the Variants slider) to snap back to a fitted view.
Focusing on the most common paths (Variants)
Real-world processes often have dozens of rare variations. The Variants slider at the top lets you control how many of the most common paths are shown, from just the single most common path up to every path that occurs.
The label reads "N of M (P%)", N = paths currently shown, M = total distinct paths in your data, P% = the share of cases those N paths represent.
Drag the slider, or use the −/+ buttons, to reveal more or fewer paths.
Starting with a small number of variants is a good way to see the "backbone" of your process before adding in the noise of rare exceptions.
If you've used the Sample control (below), the info bar also shows how many cases are represented in the current sample and how many are actually shown after variant/filter reduction.
The toolbar at a glance
Across the top of the visual you'll find:
Active features are highlighted in blue/teal so you can see what's currently turned on.
Filtering the process
Hiding activities
Click Activities ▾ to open a list of every activity with a checkbox next to it. Uncheck any you want to hide from the diagram, or use Hide all / Show all to toggle everything at once. The list also shows each activity's case count and average position in the process, click a column header to sort by it. The button label updates to show how many activities are currently hidden, e.g. "Activities (8/12) ▾".
Right-click menus
Right-click an arrow: choose Include or Exclude to keep only cases that took that path, or remove them.
Right-click one or more selected activities: choose to include only cases that visited those activities, or exclude cases that did.
Every filter you apply shows up as a removable chip (e.g. "+ Include: Approved (240)") in a strip below the toolbar. Remove one with its ×, or clear everything with Clear all.
Important: these Include/Exclude filters aren't just a local view change, they apply as a real filter on the underlying data, so they will also filter any other visuals on the same report page that use the same Case ID field.