Predictive occupancy is forecasting how full an office, floor or car park will be before the day arrives, built from booking patterns, attendance history and calendar signals. Instead of discovering on Tuesday morning that the office is over capacity, the forecast says so on Friday.
The forecasts feed practical decisions: opening or closing floors, catering numbers, desk ratios, and whether a lease renewal should be smaller. Accuracy depends on input quality, which is why offices running booking with check-in get usable predictions and offices running guesswork get weather forecasts.