Throughput measures the number of completed items per time unit, such as tickets per week or features per month. Compared to Velocity, which relies on story point estimates, Throughput is simpler to measure because it merely counts how many items have been marked as done. This eliminates the entire estimation overhead that story points require.
In practice, a team might track that it completes an average of 15 tickets per week. With this number, forecasts become straightforward: if 45 tickets remain in the backlog, it will take approximately three weeks. Through Little’s Law, Throughput is directly connected to WIP and Lead Time. If the team wants to increase Throughput without raising work intensity, it must reduce WIP, because less parallel work means faster completion and therefore a higher throughput rate.
The Throughput approach comes from queuing theory and was popularized in software development by the NoEstimates movement. It is particularly suited for teams that work with similarly sized items or want to deliberately forgo elaborate estimation processes.
Apply methods, don't just know them. View trainings and workshops
