What is throughput?
Throughput is the number of work items a team completes in a given time period. It is typically measured as cards completed per week, per sprint, or per month. Unlike velocity, which measures estimated effort (story points), throughput counts actual completed items regardless of their size. A team that finishes 12 cards in a week has a throughput of 12, whether those cards were trivial fixes or substantial features.
Throughput is one of the three core flow metrics in kanban, alongside cycle time and work in progress. Together, they are connected by Little's Law: average cycle time equals WIP divided by throughput. This means that monitoring throughput alongside WIP gives you a direct lever for controlling how fast work moves through your pipeline.
Why throughput matters
Throughput is the most practical metric for forecasting. If a team consistently completes 15 cards per week and there are 30 items remaining in the backlog, you can reasonably predict the work will take about two weeks. This prediction does not require estimating each item individually - it uses historical completion data to project forward. The longer the history and the more consistent the throughput, the more accurate the forecast.
Throughput also reveals capacity issues that cycle time and WIP alone do not. A team can have low WIP and fast cycle times but still have low throughput if work items are arriving too slowly (starved pipeline) or if the team is undersized for the scope of work. Conversely, high throughput with rising cycle times suggests the team is completing many items but taking longer on each one, which usually signals growing complexity or technical debt.
Improving throughput
The most reliable way to increase throughput is not to work faster, but to reduce the size of work items. Smaller cards flow through the pipeline with fewer delays, fewer blockers, and fewer handoff complications. A team that breaks a five-day feature into three smaller cards will typically complete all three faster than the original single card, because each piece can move independently through review and testing.
Reducing WIP can also increase throughput counterintuitively. Teams with too much work in progress spend significant time context-switching between tasks, which reduces the effective time spent on any single item. By lowering WIP limits and focusing on finishing, the team completes more items per week even though fewer are in flight at any given moment.
How Flux supports throughput measurement
Flux tracks when every card reaches the Done column through its event-sourced activity log. By marking a column as Done on any board, the system records completion timestamps automatically. Teams can count completed cards over any time window to calculate throughput. The multi-board view extends this visibility across multiple boards, making it possible to assess throughput at the team or department level rather than just per-project.
Related terms
See also: Cycle Time, WIP Limit, Velocity.