Glossary & Legends

Reference

Glossary & Legends

The dictionary for the whole module: every term you'll meet, and the exact status vocabularies with the colour each badge uses. When a word or a coloured chip is unfamiliar, look it up here.

Glossary — A to L

Assignment

One employee within one review cycle, scored against one template, with an optional named reviewer. The unit the engine calculates. Stored in kpi_employee_assignments; its status runs pending → in_progress → completed.

Automation flag

Whether a metric computes by itself. Automated (from existing data), semi-automated (auto with a manual fallback), or manual (a reviewer types the value). See the automation legend.

Bell-curve (target-band) metric

A metric that is healthiest at a moderate value, not a maximum — Overtime Utilization, Leave Utilization, Leave Balance. Too little and too much both score low; the middle band scores best. See Scoring.

Category

One of the 8 groupings of metrics (Attendance, Task Performance, …). Each carries a weight in a template and its own rolled-up category score. See the category legend.

Composite Score

The single 1.00–5.00 number for an employee in a cycle — the weighted average of their category scores. It maps to a rating band and drives the salary recommendation. Stored in kpi_composite_results.

Computation log

A JSON record saved with every metric score capturing its inputs, the exact formula, and a timestamp — so any score is auditable and reconstructable.

Review Cycle

A defined evaluation period with a cadence, a start/end date, and a template. Lifecycle draft → open → calculating → review → finalized (or archived). Stored in kpi_review_cycles. See the cycle legend.

Data Sufficiency

The percentage of a template's metrics that actually produced a value for an employee ((total − null) / total × 100). ≥ 60% is treated as sufficient; below that the composite is still shown but flagged low confidence and should be excluded from rankings.

Dimension

What a metric measures at heart — Quantity, Quality, Efficiency, Reliability/Compliance, Development, or Wellbeing. Categories deliberately mix dimensions so one behaviour can't dominate.

Direction

Which way is “good” for a metric: ↑ higher better, ↓ lower better, or ◑ target-band (a healthy middle). Direction only changes how the raw value is normalized.

Finalize

Advancing a cycle to its finalized state — the results are settled and salary recommendations can be actioned. The cross-cycle history snapshot makes trends possible.

Grade / Grade Step

An employee's position in the salary structure: a grade band (e.g. Grade 3) and a step within it (e.g. Step 3 of 5). The KPI rating proposes a step change. See Salary Progression.

Leading vs. Lagging indicator

Lagging confirms a result that already happened (attendance, completion). Leading predicts future results (certifications, training, burnout signals). Every metric is tagged one or the other.

Low-confidence result

A composite produced from below-threshold data sufficiency (typically a new hire or someone on extended leave). Keep it visible, but don't rank on it.

Glossary — M to Z

Manual Entry

A value a reviewer types in for a metric the system can't measure automatically — e.g. Ticket Satisfaction, Training Hours. Entered per assignment and normalized onto the 1–5 scale like any other metric. Stored in kpi_manual_entries.

Metric

One measurable indicator (there are 29). Has a stable metric_key, a formula, a unit, a direction, an automation flag, and five threshold bands. Its per-assignment result is a metric score stored in kpi_metric_scores.

Normalization

Converting a raw metric value (a %, a count, hours) onto the common 1.00–5.00 scale so different metrics can be compared and averaged fairly. Uses within-band linear interpolation. See Scoring.

PIP (Performance Improvement Plan)

A structured plan for an underperforming employee. The salary engine flags a PIP for Needs Improvement and a mandatory PIP for Unsatisfactory ratings — it flags, it does not create the plan.

Rating Band

The named quality label a score falls into: Outstanding Exceeds Meets Needs Improvement Unsatisfactory. See the rating legend.

Raw value

A metric's result before normalization, in its own natural unit. A null raw value (no data / divide-by-zero) is excluded from the score, never scored as 1.00.

Reviewer

The person nominated on an assignment to own the evaluation. Optional; informational on the scorecard.

Salary Recommendation

An advisory proposal to move an employee by N salary steps, priced against their grade, generated from the composite rating. Awaits Approve/Reject. Stored in kpi_salary_recommendations. See the salary status legend.

SMART target

A goal that is Specific, Measurable, Achievable, Relevant and Time-bound. Each metric ships a recommended SMART target to guide reviewers setting expectations.

Step change

The number of salary steps a recommendation proposes: +2 (Outstanding), +1 (Exceeds), 0 otherwise. Capped by a configurable maximum.

Template

A reusable scorecard: which metrics are enabled and the category / metric weights. Ships with 5 role-based weight profiles. Stored in kpi_templates + kpi_template_metrics.

Threshold band

The five raw-value ranges (one per rating band) that a metric's normalization uses to map its raw value onto 1–5. Can be overridden per template-metric to role-calibrate a metric.

Weight

The relative importance of a category (or a metric within a category) in the score, as a percentage. Category weights must total 100% per template.

---

Legends & colours

Each set below is shown as a lifecycle (states left to right, success end green, stop end red) and a legend of the coloured badges.

Rating bands (the master legend)

The 1.00–5.00 quality bands used for every metric, category and composite score. These five colours are constant everywhere in the module.

Review Cycle status

The cycle lifecycle. Status advances mostly through the calculation flow; only Draft/Open cycles can be deleted.

Review Cycle cadence (type)

Assignment status

The state of one employee's scorecard inside a cycle.

Salary Recommendation status

The approval state of an advisory step change.

Rating → step change

How each composite rating maps to a proposed salary-step change (capped by a configurable maximum, default +2). Full rules on Salary Progression.

Metric direction

Metric automation

The 8 category codes