Composite indicators
Path: HR Analytics Overview / Role Center
Composite indicators combine the other analytics into a single readable picture. They do not recompute anything — they read the values the individual analytics have already published, so a composite always agrees with its parts. What goes into each composite is configuration, driven by the Include in Digest flag and the Digest Sort Order on each instance.
| Indicator | Code | Combines | Read it as |
|---|---|---|---|
| Workforce Health Score | CMP-HEALTH-SCORE | Turnover, fill rate, absence, overtime, real wage, tenure | A single 0–100 score |
| Department Report Card | CMP-DEPT-CARD | The digest indicators, per unit | A 0–100 rank per department |
| Monthly Digest | CMP-DIGEST | The digest indicators | A readable period summary |
| Risk Indicators | CMP-RISKS | Independent early-warning rules | A list of triggered risks |
Workforce Health Score
- What it measures — a single 0–100 score combining turnover, staff fill rate, absence rate, overtime, real-wage growth and average tenure.
- Why it matters — one number for "how healthy is the workforce", with the components always shown so the score is never a black box.
- How it is calculated — each component is read from its own analytic's current value (never recomputed), scored 0–100 against a configurable green/amber band, then weighted. Components without data are dropped and the remaining weights are rescaled.
- How to read it — higher is healthier; open the breakdown (one row per component) to see why the score is what it is.
- Settings — the weights and bands are configuration (a health score is expert opinion, not a fixed formula), edited from the Components action on the instance. Refreshes weekly.
Department Report Card
- What it measures — for each organizational unit (with headcount at or above the Minimum Group Size), a compact card assembled from the analytics flagged Include in Digest — by default Headcount, Turnover Rate, Staff Fill Rate, Absence Rate by Dept, Overtime Hours, Payroll Fund and Average Salary (in that order). Each line shows the analytic's name and its value in the analytic's own format (%, currency, hours), compared with the company: a rate, score or the average salary is followed by the company value (7.3% (company 11.4%)), a count or amount by the unit's share of the company total (23 (33.3% of company)). The company figure is the source analytic's own total, so it covers the same employees as that analytic. An analytic appears on a unit's card only if it breaks down by organizational unit (Average Salary, grouped by job title by default, therefore shows in the Monthly Digest, not on the cards).
- Why it matters — lets you compare departments at a glance and is the content behind the Unit Manager digest and the Org Unit FactBox.
- How it is calculated — a unit's figures include all of its child units. Cards are built for the units at or above the Org. Unit Group Level (every unit when Org. Unit Grouping is Each Unit), because the card adds up the department rows the included analytics publish — give those analytics the same grouping settings. Headcount is counted exactly for the unit's subtree; other count and amount metrics (for the payroll fund, the department's latest period) are summed across the child units' published rows — a sub-unit under the Minimum Group Size is folded into "Other" by its source analytic and is then missing from its parent's sum, while rate and score metrics (such as turnover, absence rate or fill rate) and the average salary are combined as a headcount-weighted average, so a larger child unit's rate counts proportionally more than a smaller one's. The card value is a 0–100 composite rank score and its severity is taken from the worst included component.
- Change against a quarter ago — every figure also shows how it moved over the last three months, with an arrow: in percentage points for rates (▼ -1.5 pp), as a percent for amounts (▲ +4.2%), as the difference for counts, days and hours (▲ +2), or no change. For the quarter-ago figures each included analytic is calculated once more, as of three months before the reporting date, with the same settings — nothing of that run is saved — and the unit's figure is assembled from it exactly as today's; headcount is counted exactly at that date. If an analytic cannot be calculated for the earlier date, its figure is shown without the change.
- How to read it — the score ranks the unit; the card lines explain it.
- Drill-down — the Overview filtered to that unit.
- Settings — Include in Digest and Digest Sort Order on the source instances (set on the seven default analytics above when they are created); Minimum Group Size. Refreshes weekly.
Monthly Digest
- What it measures — a single readable digest assembled at period close from the analytics flagged Include in Digest, in their configured order.
- Why it matters — the once-a-month "everything that matters" summary; also the basis of the User digest e-mail.
- How it is calculated — for each included analytic it shows the current value, its trend versus the comparison value and its severity (red / amber / green). The body is stored language-neutral and rendered in the reader's language when opened, and contains no per-person data.
- How to read it — top to bottom in importance order; colours flag what needs attention.
- Settings — turning the Include in Digest flag on/off for an instance adds/removes its line; Digest Sort Order sets the order. Refreshes on period close.
Risk Indicators
- What it measures — a set of independent early-warning rules; a row appears only when a rule is triggered, and its text explains the trigger.
- Why it matters — surfaces compound risks that no single KPI shows on its own.
- How it is calculated — the rules cover a mass exit of voluntary leavers in one unit, signs of team burnout from sustained overtime, a retirement wave (a unit with many people near retirement age), overtime approaching the yearly limit, and qualifications about to expire.
- How to read it — an empty list is good news; each row is a specific risk to investigate.
- Settings — the voluntary-leaver count (Threshold 1, default 3), the mass-exit look-back window (Window Date Formula 2, default 2 months), the retirement share (Threshold 2, default 15%), the retirement age (Threshold 3, default 55), the overtime sigma and the expiry window are all configurable. Refreshes weekly.
See Also
- E-mail digests & alerts
- Copilot / AI insights — an AI narrative over the same data
- Configuring analytics