In 30 seconds
- Points compare items; hours predict time.
- Points suit forecasting a team's output; hours suit cost and capacity questions.
- Converting points to hours throws away their main benefit.
Pick story points when…
- Work is uncertain and estimates in hours keep missing
- You want to forecast a team's delivery with velocity
- You want estimates that don't depend on who does the work
Pick hours when…
- Work is routine and well understood
- You bill clients by the hour
- You're checking a sprint plan against capacity
Side by side
| Story points | Hours | |
|---|---|---|
| Measures | Relative size: effort, complexity, uncertainty | Expected time |
| Depends on person | Less | More |
| Speed to estimate | Fast, by comparison | Slower |
| Main risk | Misused as productivity scores | False precision, padding |
Myth vs reality
Myth · A point equals half a day.
Reality · Fixing a conversion turns points back into hours.
Myth · Hours are more accurate.
Reality · For uncertain work, they're usually less accurate.
Myth · You must choose one forever.
Reality · Many teams use points for planning and hours for capacity checks.
The AI angle
With AI, both points and hours tied to effort become less reliable. Many teams move toward counting small items and forecasting from throughput.
FAQ
What about ideal days?
A middle ground: time-based, but assuming no interruptions.
Can we drop estimates entirely?
If items are small and similar, counting items works well.
How do stakeholders understand points?
Translate forecasts into date ranges, not points.