Employee Turnover Sankey Diagram: Editable Template
Show where headcount stays, moves or leaves by department, and why people leave. Replace the counts with your own HR data to see where turnover concentrates.
A turnover Sankey splits starting headcount by department into people who stayed and people who left, then gives each group a second level: why they left, or whether the ones who stayed moved up.
Five hundred employees at the start of the year are shown by department. Those who left are split by reason, and those who stayed by whether they kept the same role or were promoted or transferred.
Treat this as sensitive data. Aggregate to groups large enough that nobody can be identified, and report reasons from structured exit data. The figures are invented.
How to read this diagram
- Of 500 employees, 90 left (18%) and 410 stayed.
- Support lost 30% of its people and Sales 25%, against 10% in Engineering and 12.5% in Operations.
- A better offer elsewhere is the largest single reason (35 of 90).
- 70 of those who stayed were promoted or transferred.
The data behind it
Invented figures for illustration. Not data about any real business.
| From | To | Amount (employees) |
|---|---|---|
| Engineering | Stayed | 180 |
| Engineering | Left | 20 |
| Sales | Stayed | 90 |
| Sales | Left | 30 |
| Support | Stayed | 70 |
| Support | Left | 30 |
| Operations | Stayed | 70 |
| Operations | Left | 10 |
| Stayed | Same role | 340 |
| Stayed | Promoted or transferred | 70 |
| Left | Better offer | 35 |
| Left | Relocation or personal | 15 |
| Left | Involuntary | 20 |
| Left | Retirement | 20 |
The same data as CSV
source,target,value
Engineering,Stayed,180
Engineering,Left,20
Sales,Stayed,90
Sales,Left,30
Support,Stayed,70
Support,Left,30
Operations,Stayed,70
Operations,Left,10
Stayed,Same role,340
Stayed,Promoted or transferred,70
Left,Better offer,35
Left,Relocation or personal,15
Left,Involuntary,20
Left,Retirement,20How it works
Fix the headcount baseline
Use headcount at the start of the period and track the same people.
Group and anonymise
Group by department and make sure no group is small enough to identify a person.
Classify exits
Use a short, fixed list of leaving reasons.
Enter and check
Type the flows and confirm that stayed plus left equals the starting headcount.
Frequently asked questions
How is turnover rate calculated?
Leavers divided by average headcount is common. This chart uses starting headcount for simplicity; state which one you use.
Should new joiners be shown?
They are not in the starting headcount. Show them in a separate chart or add them as a source if you want a full workforce flow.
Edit this turnover Sankey
Open the example as your own chart and enter your headcount flows.