Throughput Calculator

Throughput is the rate at which work items are completed and delivered over a defined time period. It is measured in units of output per time unit, such as stories per sprint, features per week, or tasks per day. In the Kanban Method and Agile flow management, throughput is one of the three fundamental flow metrics alongside work in progress (WIP) and cycle time. Together, these three metrics are linked by Little's Law, the mathematical relationship that forms the foundation of modern flow-based project management.

The PMBOK Guide addresses throughput within the Monitor and Control Project Work process, where it serves as a key performance indicator for delivery velocity. Unlike velocity in Scrum, which is measured in story points and can be gamed by inflating estimates, throughput counts actual completed items. This makes it a more objective and comparable metric across teams and projects. When stakeholders ask "How fast are we delivering?" throughput provides the clearest, most honest answer.

In throughput accounting, a concept developed by Eliyahu Goldratt as part of the Theory of Constraints, throughput takes on a financial dimension: it is defined as the rate at which the system generates money through sales. While this financial definition is more relevant to operations management than software project management, the underlying principle translates directly: throughput measures value delivery rate, and anything that constrains throughput constrains the entire system's performance. Identifying and elevating that constraint is the core management discipline.

Average WIP (Work in Progress) is the mean number of items actively being worked on during the measurement period. This includes all items that have been started but not yet completed. WIP is the primary lever for improving flow: reducing WIP reduces cycle time and increases predictability.

Throughput is the average number of items completed per time period. For a stable system, this is calculated by dividing total completed items by the number of time periods. For forecasting purposes, use the median or 85th percentile throughput rather than the mean to account for variability.

Average Cycle Time is the mean elapsed time from when work begins on an item to when it is completed. Little's Law states that for a stable system, the product of throughput and cycle time equals the average work in progress. This relationship holds regardless of the probability distribution of arrival or service times, making it one of the most robust and widely applicable results in queueing theory.

• Counting partially completed items as throughput — Only items that have passed through your entire definition of done should count. Counting items that are "90% done" inflates throughput and gives a false picture of delivery capability.
• Using throughput alone for forecasting — Raw throughput averages do not account for variability. Use Monte Carlo simulation or percentile-based forecasting to provide probabilistic estimates with confidence levels.
• Increasing WIP to increase throughput — Beyond a certain point, more WIP reduces throughput due to context-switching overhead and queue congestion. The relationship is counter-intuitive but well-documented: limiting WIP often increases throughput.
• Not accounting for item size variability — If some items take 1 day and others take 3 weeks, throughput counting becomes noisy. Break large items into smaller, more uniform sizes for more stable and meaningful throughput measurements.

For the PMP exam, throughput is primarily tested in the context of Agile and Kanban methodologies within the Executing and Monitoring and Controlling domains. Know that throughput measures the count of completed deliverables per time period, not the effort or story points. This distinction matters because throughput is an objective output metric while velocity is a team-specific planning metric that cannot be compared across teams.

Little's Law is fair game on the exam. You should be able to rearrange the formula to solve for any of the three variables. The most common exam pattern gives you average WIP and throughput and asks you to calculate cycle time, or gives you throughput and cycle time and asks you to calculate WIP. Remember the three forms: WIP = Throughput x Cycle Time, Throughput = WIP / Cycle Time, and Cycle Time = WIP / Throughput.

Understand the Theory of Constraints as it relates to throughput. The five focusing steps are: identify the constraint, exploit the constraint (make sure it is never idle), subordinate everything else to the constraint, elevate the constraint (add capacity), and repeat. The PMBOK Guide does not explicitly name these steps, but the principle that overall system throughput is limited by the bottleneck appears in quality and resource management questions. When you see a question about which process step to improve to increase overall output, the answer is always the bottleneck step.