Plan from the supplied facts
In this fictional design, a feature can be deployed while exposure is off. The model has no automatic rollout: exposure requires a separately planned switch and audience. Count exposed audience from the supplied exposure column, not from code presence.
| Feature | Code deployed | Exposure | Exposed users |
|---|---|---|---|
| A | Yes | Off | 0 |
| B | Yes | Named cohort only | 20 |
| C | No | Off | 0 |
Work through the choice
Two deployed features do not imply two exposed features. B has a limited audience; A has none. Neither count supplies evidence that users completed their tasks.
How many features have code deployed, and how many have nonzero exposure?
Show the worked answer
Two deployed, one exposed. The exposed count is B only; the table supplies 20 exposed users but no task-completion count.
Material to use in your plan
Use separate columns for deployed revision, exposure setting, eligible cohort and observed task outcome. Name the metric when reporting a release milestone.