Understanding Metrics

A metric is a quantifiable data point used to track, analyse, and evaluate the status of a specific element of performance. Unlike KPIs, which are curated indicators, metrics are broader in scope and volume — they are the building blocks from which insight is constructed.

Metrics serve several key purposes:

• Support KPIs with evidence

• Enable operational monitoring

• Feed AI models and automation logic

• Reveal root causes and performance trends

✅ Example:

KPI: % of incidents resolved within SLA Supporting Metrics: Average resolution time; resolution time by priority; % of incidents breaching SLA; time to assign; time to resolve by resolver group

These underlying metrics allow deeper diagnosis, faster action, and continuous refinement.

Metrics are not:

• Indicators of success on their own – they require context

• Always useful – just because something can be measured doesn’t mean it should be

• A substitute for analysis – collecting data without interpretation provides no value

An overabundance of poorly structured metrics leads to:

• Data fatigue

• Inconsistent reporting

• Loss of trust in dashboards

To ensure metrics add value to the performance chain, SPARA defines the following attributes:

1. Relevance
The metric must inform a KPI, support decision-making, or enable operational control.

2. Accuracy
Metrics must be drawn from reliable sources and consistently calculated.

3. Timeliness
They must be updated at a cadence that reflects the decision cycle they support (real-time, daily, weekly, etc.).

4. Segmented
Where possible, metrics should support filtering by relevant dimensions (region, team, category, etc.).

5. Accessible
Metrics must be retrievable and usable by those who need them, not trapped in inaccessible systems.

Metrics underpin the entire SPARA Measurement Chain. While they are often operational in nature, they directly influence:

• KPI dashboards

• CSF tracking logic

• Intervention triggers

• Governance thresholds

• Strategic review and forecasting

Metrics also feed the digital systems that enable performance orchestration:

• Business intelligence dashboards

• AI-enabled alerting and insights

• Process mining and automation

SPARA classifies metrics by their use and intent:

Descriptive Metrics

What happened?

• Example: Number of incidents logged last month

Diagnostic Metrics

Why did it happen?

• Example: % of incidents resolved at L1 vs. L2

Predictive Metrics

What’s likely to happen next?

• Example: Incident volume trendline based on past 90 days

Prescriptive Metrics

What should we do?

• Example: % of changes flagged by AI as likely to fail

Many organisations suffer from inconsistent or ungoverned metrics. To avoid this, SPARA recommends building a Metric Catalogue, aligned to:

• KPIs and CSFs

• Data ownership

• Calculation rules

• Frequency and audience

A good catalogue helps avoid metric duplication, ensures consistency, and supports onboarding of new team members or toolsets.

Objective: Improve change success rate

• CSF: Change risk assessments must be accurate and complete

• KPI: % of changes implemented without incident

• Metrics:

  • % of changes assessed as high risk

  • % of changes with missing impact analysis

  • % of changes with peer review completed

Each metric strengthens our ability to manage performance, not just observe it.

• Data over-collection – leads to overload and obscures insight

• Stale metrics – aged data drives poor decisions

• Unaligned metrics – not linked to performance logic

• Manual extraction – leads to inconsistency and human error

SPARA positions metrics as managed assets, not passive outputs.

Metrics are the technical foundation of performance architecture. They give structure to KPIs, support CSF analysis, and enable data-driven operations. Without well-defined and maintained metrics, organisations are flying blind — and unable to explain or improve what they measure.

With the full SPARA Measurement Chain now defined, the final chapter will explore how to integrate the layers into a usable, consistent, and evolving system of performance control.