May 19, 2026
6 min read
Forecasting and predictability
Throughput Forecasting vs Velocity Forecasting
A practical comparison of throughput forecasting and velocity forecasting, and how to tell which one fits the team and the work better.
Why teams compare the wrong things
Teams often debate throughput and velocity as if they were rival beliefs. The real question is more practical: which signal is this team actually able to trust when it needs to forecast delivery?
If work is sliced inconsistently, throughput gets noisy. If point calibration drifts, velocity gets noisy. In both cases, the problem starts before anybody reaches for a spreadsheet.
Forecast model
Throughput and velocity forecasting are both useful when the team understands what each one is actually measuring.
Two signals
The real question is not which label sounds smarter. It is which signal fits the team's work and planning reality better.
Throughput forecast
This model is often cleaner when work items are similar enough that finished count tells a useful delivery story.
Velocity forecast
This model can still help when relative sizing is stable and the team understands what point totals do and do not mean.
Choose by context
The better system is the one that explains movement with less distortion, not the one that wins a framework argument.
Honest comparison
Teams forecast better when they compare models against real delivery behavior instead of picking by habit.
What throughput forecasting is really measuring
Throughput forecasting looks at how many work items the team finishes over time. That makes completed items the main delivery signal instead of turning relative estimates into the forecasting unit.
This is often easier to explain outside the team because the input is observable. It also fits well when the team is already trying to slice work into reasonably comparable pieces and cares about flow more than point math.
What velocity forecasting is really measuring
Velocity forecasting looks at how many story points the team tends to complete over time. It can still be useful when the team estimates consistently and treats points as a stable relative scale rather than as disguised hours.
The weakness is that velocity is one step removed from observable output. A finished-point total can look steady even while the meaning of a five-point story is shifting underneath it.
Where each one usually breaks
Neither approach is magic. Each one breaks when the planning system underneath it is messier than the metric makes it look.
- Throughput gets distorted when tiny items and oversized stories are treated as comparable units.
- Velocity gets distorted when point calibration drifts between backlog shapes, teammates, or time periods.
- Both become weak when the backlog is not ready and hidden dependencies keep changing what 'done' really means.
- Neither metric repairs capacity assumptions that were unrealistic from the start.
How to choose without making it ideological
Throughput is often the better fit when the team trusts how it slices work more than how it calibrates points. Velocity can still work when story points are well understood locally and the team uses them carefully as a relative planning signal.
Some teams use both. Throughput supports broader delivery ranges. Velocity helps inside sprint planning where relative sizing is already part of the local conversation. The healthier move is not to declare one metric universally better. It is to understand what each one is actually saying and what it may be hiding.
TL;DR
- Throughput forecasting measures completed items, while velocity forecasting measures completed story points.
- Throughput usually works better when work is sliced consistently and the team wants a more observable delivery signal.
- Velocity can still work when point calibration is stable and the team uses points carefully as a relative planning tool.
- Both approaches fail when the backlog is weak, capacity is ignored, or the input data looks cleaner than reality.
- Throughput and velocity are both just signals, but the better one is the one that fits your work mix and explains delivery behavior more honestly.