HKSM
performance measurement
Quantitative metrics that describe how a system behaves in operation, capturing its physical and functional characteristics.
Key Points
- Focuses on operational behavior such as latency, throughput, capacity, reliability, and error rates.
- Should be objective, repeatable, and tied to acceptance criteria or nonfunctional requirements.
- Collected regularly (e.g., per iteration) to reveal trends, forecast issues, and guide improvements.
- Works best when defined with clear units, targets, thresholds, and automated collection in CI/CD.
Example
An agile team delivering an API tracks 95th percentile response time, requests per second, CPU/memory usage, and error rates each sprint. When response time rises above 500 ms at expected load, they add backlog items to optimize database queries and introduce caching until the metric consistently meets the target.
PMP Example Question
Which of the following best represents performance measurement for a new streaming service?
- 95th percentile video start-up time under 2 seconds at 50,000 concurrent users
- Team velocity improving from 30 to 35 story points per sprint
- Stakeholder satisfaction score of 8/10 after release
- A complete Definition of Done for all backlog items
Correct Answer: A — metrics describing system operational behavior
Explanation: Option A is a quantitative metric of how the system performs under load, reflecting its operational physical and functional characteristics; the others are not direct measures of system operation.
