You never need a space for every subscriber
A subscriber pays for a guaranteed space, but almost never occupies it around the clock, people travel, work from home, come and go. If you kept one space idle for every subscription you'd sold, most of that guaranteed capacity would sit empty at any given moment, earning nothing.
The real question is how many to hold back, right now
Say you've sold 250 subscriptions across 450 spaces. You never keep 250 empty for them, they're never all here at once. At each hour you hold back only enough for the subscribers arriving over the next few hours, and sell the rest.
So the useful answer isn't a fixed number; it's dynamic. At this hour, how many spaces must you reserve so every subscriber still to arrive finds one? If most are already inside and few more are expected, maybe ten. If few are inside and a rush is coming, maybe a hundred. GENIE forecasts subscriber arrivals and answers that hour by hour, freeing every space it safely can for transient drivers.
Calibrated to a service standard you set
Hold back too little and a paying subscriber meets a full lot. GENIE sizes the reserve to a standard you choose, the most you'll accept as the chance a subscriber is ever turned away, from each segment's real arrival patterns, and recalibrates as they change. More occupancy, held to a risk level you decide.
Keeping the promise to subscribers
Smart overbooking only works if subscribers keep trusting the guarantee. Because GENIE reserves to a strict standard and watches live occupancy, the chance of a real shortfall stays inside your tolerance. You can hold extra buffer at peak, exclude specific reserved spaces, and set exactly how conservative to be.
Part of one optimisation
Overbooking is strongest with the other two levers. Freeing spaces this way expands what subscription optimization can safely sell, while dynamic pricing prices the transient demand that fills them. See how it works for the full pipeline.