Before choosing which personnel metrics you are going to measure, decide which system you will use to capture them. Many HR teams start the wrong way round: they ask for a dashboard with the “most used” KPIs and then discover that their ERP does not export the employee’s termination date or that turnover is calculated by hand in Excel every quarter. This article explains how to set up those indicators in order, along with the usual stumbling blocks that appear when configuring them.
Step 0: define the data sources before touching the dashboard
The most widely used HR KPIs (turnover, absence, time to fill vacancies, cost per hire, training hours, internal promotion rate) are fed by three or four sources that rarely live in the same place:
- Payroll or HR ERP: hires, terminations, salary, contract type.
- ATS or recruitment system: vacancy opening and closing dates, candidates by stage.
- Time tracking or clock-in system: actual hours, absenteeism, sick leave.
- LMS or training records: courses, hours, certifications.
Before configuring anything, check that there is a unique employee identifier that travels between those systems. Without that common key, any data cross-referencing becomes a manual job every month.
Step 1: set the exact definition of each KPI
The most common mistake is using the same name for different formulas. A real example: two managers calculating turnover. One counts voluntary and involuntary terminations over average headcount; the other counts only voluntary ones over headcount at period end. The results can differ by several percentage points without anyone being wrong. Before automating, write the formula and the period in a shared definitions sheet:
- Turnover: total terminations ÷ average headcount for the period × 100.
- Voluntary turnover: only terminations by employee decision.
- Absenteeism: justified hours not worked ÷ scheduled hours × 100.
- Time to hire: from vacancy approval to offer acceptance.
- Cost per hire: internal and external process costs ÷ number of new hires.
Documenting this definition is what separates a useful KPI from a figure that each director interprets in their own way.
Step 2: build the calculation layer
With the sources connected and the formulas defined, build an intermediate table (in the ERP, in a BI tool or even in a well-structured spreadsheet) that aggregates the data by month and by department. This layer prevents the dashboard from querying payroll tables directly, something that slows down the system and complicates access permissions.
Here a limitation appears that is worth acknowledging: if your headcount is small (fewer than 30-40 people), turnover and absenteeism KPIs fluctuate so much that a single dismissal moves the needle by several points. In those cases, measure quarterly trends instead of monthly ones, or combine the figure with the reason for the termination.
Step 3: configure the visualization and thresholds
Not all KPIs deserve to be at the top level of the dashboard. A practical setup for management: turnover, absenteeism and time to hire in the first row; cost per hire, training hours per employee and internal promotion rate in a second layer of detail. Add alert thresholds based on your own history, not on generic industry benchmarks: what is normal in hospitality is not normal in a technical consultancy.
Step 4: review and adjust the cycle
Schedule a quarterly review of the definitions. Organizational changes (new locations, collective agreements, changes to working hours) alter the formulas without the dashboard reflecting it. According to industry studies, a significant share of HR dashboards end up being abandoned due to a lack of maintenance of the sources, not because of poor design.
Configuring these KPIs does not by itself guarantee better decisions: it delivers consistent data so that decisions can be discussed with comparable figures.