Log a task outcome

This is where Execute becomes Measure. A completed task with no recorded result is an open loop: the system cannot learn from it, and neither can you, so the next plan repeats guesses instead of building on evidence. Logging too early is just as costly. A number pulled before the measurement window closes is fiction, and a plan changed on the strength of fiction spends the next week's budget on the wrong work. The window exists so the result you record is a fair one.

Only when the measurement window is ready: the task is done and its window has produced a real number. Not at the moment of completion.

• The task is checked done (see Mark work honestly: in progress or done).
• The measurement window from the task's contract has actually closed, and you have a real value, source, and data-as-of time.
• You know which kind of task this is: manual, integration-backed, or completion-only.

1. 1Open the "Log a task outcome" panel on the Marketing Plan (when a result is waiting, the "Results needed" item in the section nav jumps to it). It says: "Choose a completed task to review or record its result." and lists your completed tasks.
2. 2Pick the completed task. For a manual task, the card reads "Task completed. Result still needed." Choose "Add Result", enter the actual value for the metric the card names, and click "Save result". Keep your own note of where the number came from and when; the card already names the source and window.
3. 3For an integration-backed task, the card shows "Task completed. Measurement pending." with when to expect the result: do nothing. The baseline was captured when you marked it done, and the result lands when the window closes. Provider read failures stay visible ("Measurement temporarily unavailable." while the app retries); wait rather than guessing a number.
4. 4For a completion-only task, there is no result affordance at all. The done state is the proof. Don't attach a metric where none belongs.
5. 5A saved manual result shows on its card as "Recorded" with the metric and value. Measured results can also appear under "Recorded task results", each with when it was recorded, its verdict, and its source. Integration-backed results say whether they were matched to the task's exact published page or post or are a workspace metric observed during the window. Distribution evidence (reach, views, impressions) is never business impact (leads, revenue); label it as what it is.

Manual tasks waiting on you say "Task completed. Result still needed." Integration-backed tasks in their window say "Task completed. Measurement pending." and show when the result is expected. Completion-only tasks show no result affordance. Recorded results accumulate as the week closes out.