Capacity Planning: Trend, seasonality and events
Objective. Inspect the trend, seasonality and events shaping the forecast as sanity-checkable assumptions, not as guaranteed adjustments.
This screen answers: what underlying pattern and which known events are shaping the forecast? It opens up the forecast so you can check its drivers make sense for this customer.
See it in the demo
What this screen answers
The customer question is: what is driving the shape of my forecast - an underlying trend, a seasonal pattern, specific events - and are those right for my operation. A CSM uses it to validate the forecast's logic; a rep uses it to show the forecast is explainable, not a black box.
The task
Inspect the decomposition - the trend and seasonal components - and the events applied to the forecast period. The job is to sanity-check: does the seasonality match the customer's real calendar, and are the right events in or out.
How to read it
These are the forecast's drivers and assumptions, shown so you can inspect them. A seasonal uplift or an event adjustment is still part of a projection - it has error bars like everything else. The classic misread is to treat an event adjustment as a guaranteed bump, or to assume an event that has not been modelled is somehow covered. If the customer's real calendar has an event the forecast does not, the forecast will miss it.
Honest limits
Which events, holidays and calendars apply must be confirmed with the customer - demo data is seeded. The decomposition explains the projection; it does not make it certain, and nothing here is executed. An unmodelled event is a known blind spot to raise, not a hidden safety margin.
Interpretation check
Trend, seasonality and events are the forecast's assumptions to sanity-check; a modelled event is still a projection and an unmodelled one is simply missing.
Practical task
Inspect the trend/seasonality/events for a scope and write one line on whether they match the customer's real calendar - naming one event you would confirm is correctly in or out.
Self-check
The forecast shows a seasonal peak. Is that peak guaranteed?
No. It is a modelled component of a projection with error bars. It is a sanity-checkable assumption, not a certainty.
The customer has a major event the forecast does not include. What follows?
The forecast will miss it - an unmodelled event is a blind spot, not a covered case. Raise it so the calendar is corrected before the plan is trusted.
What is the CSM's job on this screen?
Validate that the seasonality and events match the customer's real calendar and work mix, rather than accepting seeded patterns.
Verified vs demo. Screen is live in the demo at /#/fcdrivers; exact controls and figures are SME-validation-pending.