Short Answer
Velocity tracks completed work per iteration, cycle time measures task duration from start to finish, and throughput counts completed items over time. These metrics help teams forecast capacity, identify bottlenecks, and improve delivery predictability.
Comprehensive Answer
Understanding how well an agile team delivers value requires looking beyond simple output counts. While velocity, cycle time, and throughput provide foundational insight, effective measurement combines quantitative indicators with qualitative assessments to reveal both productivity patterns and opportunities for improvement.
Velocity represents the amount of work a team completes during a fixed iteration, typically measured in story points or ideal hours. Teams use historical velocity data to estimate how much work they can commit to in upcoming sprints. A stable velocity suggests predictable delivery, while significant fluctuations may indicate inconsistent estimation practices, changing team composition, or external disruptions. However, velocity should never be compared across teams, as each group defines and estimates work differently. The metric serves internal planning purposes rather than cross-team benchmarking.
Cycle time measures the elapsed duration from when work begins until it reaches completion. This metric highlights how long individual items spend in active development, revealing delays that may stem from technical complexity, unclear requirements, or handoff inefficiencies. Shorter cycle times generally indicate smoother workflows and faster feedback loops. Teams often segment cycle time by work type to understand whether defects move faster than features, or whether certain categories consistently encounter delays. Tracking cycle time distributions rather than averages provides richer insight, as outliers often signal systemic issues worth investigating.
Throughput counts the number of work items a team completes within a given timeframe, regardless of size or complexity. Unlike velocity, which weights items by estimated effort, throughput treats each completed piece equally. This approach suits teams practicing continuous flow rather than fixed iterations. High throughput with stable quality suggests efficient delivery, while declining throughput may indicate growing work-in-progress limits, technical debt accumulation, or scope creep. Pairing throughput with cycle time reveals whether teams are completing more work by moving items faster or simply handling more items simultaneously.
Lead time extends the measurement window to include the period before work begins, capturing the duration from initial request to final delivery. This broader view exposes delays in backlog refinement, prioritization, and intake processes. Organizations concerned with customer responsiveness often prioritize lead time over cycle time, as it reflects the complete experience from request to fulfillment. Large gaps between lead time and cycle time suggest that work sits idle before teams can address it, pointing to capacity constraints or prioritization challenges.
Work-in-progress limits track how many items occupy the team's active workflow simultaneously. Excessive work-in-progress fragments attention, increases context switching, and delays completion. Monitoring WIP alongside cycle time helps teams identify optimal concurrency levels. When WIP rises while cycle time increases and throughput stagnates, teams are likely overcommitted. Conversely, low WIP with unused capacity may indicate overly conservative limits or upstream bottlenecks in requirement preparation.
Defect escape rate measures how often quality issues reach production or downstream phases rather than being caught during development. High escape rates suggest insufficient testing, unclear acceptance criteria, or rushed delivery. Tracking defects by severity and origin helps teams determine whether issues stem from requirements misunderstanding, coding errors, or environmental factors. Pairing defect rates with velocity reveals whether teams sacrifice quality for speed.
Sprint goal success rate evaluates how often teams achieve the objectives they commit to at iteration start. This metric assesses planning accuracy and delivery predictability. Frequent misses may indicate overcommitment, poor estimation, or mid-sprint disruptions. Conversely, consistently exceeding goals might suggest conservative planning that underutilizes capacity. Teams benefit from examining both the frequency and magnitude of variances to calibrate their commitment practices.
Customer satisfaction scores and stakeholder feedback provide qualitative balance to quantitative metrics. Delivering quickly means little if the output fails to meet user needs. Regular feedback sessions, net promoter scores, and usability assessments help teams understand whether their delivery cadence aligns with value creation. Combining delivery speed metrics with satisfaction data reveals whether faster throughput correlates with better outcomes or simply more output.
Flow efficiency calculates the ratio of active work time to total lead time, exposing how much of an item's journey involves actual progress versus waiting. Low flow efficiency indicates that work spends most of its lifecycle idle, suggesting opportunities to reduce handoffs, clarify dependencies, or adjust capacity allocation. Improving flow efficiency often yields greater gains than simply accelerating work execution.
Effective measurement requires teams to select metrics aligned with their specific improvement goals, review them regularly, and avoid gaming behaviors that optimize numbers at the expense of genuine progress. The most valuable metrics prompt conversations about how work flows through the system and where targeted changes can enhance both speed and quality.