Skip to main content

KPI tracking

Purpose. Track the measures that show whether the work is having the intended effect — with a baseline, so "improved" has a meaning.

Project details​

FieldDetail
Project
Client
Baseline period
Reporting period
Owner

Delivery KPIs​

KPIDefinitionBaselineTargetCurrentTrendOwner
Deployment frequencyProduction deployments per week
Lead time for changeCommit to production
Change failure rateDeployments requiring rollback or hotfix
Mean time to restoreDetection to service restored
Pipeline durationCommit to deployable artefact
Test coverageChanged-line coverage
Escaped defectsDefects found in production per release

Workload KPIs​

KPIDefinitionBaselineTargetCurrentTrendOwner
AvailabilitySuccessful responses / total
p95 latency95th percentile response time
Error rate5xx / total requests
ThroughputRequests or jobs per minute at peak
SaturationQueue depth or pool utilisation
Recovery test resultMeasured RTO in the last drill

Business KPIs​

KPIDefinitionBaselineTargetCurrentTrendOwner

Period tracking​

PeriodDeployment freq.Lead timeChange failureMTTRAvailabilityNote
Week 1
Week 2
Week 3
Week 4

Status key​

StatusMeaning
On trackAt or ahead of target
At riskBehind trajectory but recoverable
Off trackBelow baseline or unlikely to meet target
Not measuredNo data available — this is itself a finding

Review​

DateParticipantsHighlightsActions

Data sources​

KPISourceCollection method
Deployment frequencyCI/CD platformPipeline run history
Lead timeVersion control + CI/CDCommit timestamp to production deploy
Change failure rateIncident records + deploymentsManual tagging at deployment
MTTRIncident recordsOpen to resolved timestamps
Workload KPIsCloudWatch / observability platformMetrics and logs