Essential R&D Performance Metrics: Measure Business Value, Delivery Speed, Quality and Operations

Business Value

Total Users : All users.

New Users : Number of users registered within the last 7 days.

Active Users (✓): Users who accessed the product within the last 7 days.

Transaction Volume : Number of successful transactions within the last 7 days.

Delivery Speed

Rate (✓): Average story points completed per person‑day. Rate = Total sprint points / Total sprint person‑days

Story Points : Total points of all user stories in a sprint. Total story points = sum of story points of each Epic

Lines of Code : New lines of code added during the sprint.

Code Production Rate : Lines of code produced per person‑day. Code production rate = Sprint LOC / Sprint person‑days

Requirement Delivery Time : Average days from Epic creation to completion. Avg days per Epic

Task Count : Number of tasks in the sprint.

On‑time Completion Rate (✓): Ratio of tasks completed on schedule. On‑time rate = On‑time tasks / Total tasks (standard: completion date ≤ due date + 1 day).

Unplanned Task Ratio : Ratio of tasks added after sprint start to total tasks. Unplanned ratio = Unplanned tasks / Final total tasks

Burndown Chart : Visual representation of remaining work over time (X‑axis: time, Y‑axis: work to complete, measured by estimated person‑days).

Engineering Quality

Technical Debt : Work left due to temporary solutions, estimated repair time from SonarQube.

Severe Technical Debt : Technical debt that poses security risks or potential problems.

Debt Ratio : Technical debt relative to the time required for a full rewrite. Debt ratio = Technical debt / Full rewrite time (30 min per line of code)

Technical Debt Index (✓): Composite of technical debt, severe debt, and debt ratio.

Unit Test Coverage (✓): Coverage of unit tests.

Line coverage = LC / EL

Condition coverage = (CT + CF) / (2 × B)

Overall code coverage = (CT + CF + LC) / (2 × B + EL)

CT = conditions evaluated true at least once

CF = conditions evaluated false at least once

LC = executable lines covered by tests

B = total number of condition statements

EL = total executable lines

Build Failure Rate : Ratio of failed builds. Failure rate = Failed builds / (Successful builds + Failed builds) (weekly statistics).

Commit Count : Number of code commits (weekly).

Build Frequency : Builds per commit. Frequency = Build count / Commit count

Build Health (✓): Composite of build failure rate, commit count, and build frequency.

Deployment Duration : Average time from deployment start to success.

Deployment Efficiency : Ratio of ideal to actual deployment duration.

Deployment Frequency : Number of deployments. Frequency = Successful deployments + Failed deployments (weekly, across all environments).

Deployment Success Rate : Success rate = Successful deployments / (Successful + Failed deployments) (weekly).

Defect Count : Total bugs in the sprint.

Defect Lifetime : Average time from defect creation to closure.

Defect Density (✓): Bugs per story point. Density = Bug count / Story points

Defect Escape Rate (✓): Bugs that escape to production. Escape rate = Production bugs / Total bugs in the development cycle

Severe Defect Ratio : Ratio of severe defects to total defects. Severe ratio = Severe bugs / Total bugs

UAT Defect Ratio (✓): Bugs found in UAT relative to total defects. UAT ratio = UAT bugs / Total bugs

Branch Health : Degree of compliance with branch management standards.

Review Acceptance Rate : Ratio of approved reviews to total reviews. Rate = Approved reviews / Total review submissions

Operations Assurance

Page Views (✓): Number of full page loads within 7 days. Each user‑page view counts once; repeated views by the same user are accumulated.

Online Incident Count (✓): Number of errors logged by monitoring.

Error Rate : Probability of an error occurring online.

Alert Timeliness Rate : Proportion of alerts that are timely.

Average Response Latency (✓): Average response time.

MTTR : Mean time to recovery.

Jitter Alerts : Count of alerts that fire, recover, and then fire repeatedly.

TPS (✓): Transactions per second.