Inventory Controls and Revenue Optimization
Inventory Controls and Revenue Optimization tests whether you can protect the right rooms for the right demand, set overbooking levels, and use length-of-stay and availability controls without destroying revenue. Expect protection levels, Littlewood's rule, bid prices, wash factors, group cut-off dates, and walk costs.
Protecting inventory
Inventory can be partitioned (fixed blocks per segment) or nested (higher-value demand can use all rooms, lower-value demand is protected against). Nested controls are generally more efficient because they do not strand rooms. A protection level is the number of rooms held for higher-value demand. Littlewood's rule for two fare classes: accept a low-fare request if the low fare is at least the high fare multiplied by the probability that high-fare demand exceeds the protection level. The bid price is the marginal value of the next room and plays the same role in multi-class settings.
Overbooking
Overbooking balances the cost of spoilage (empty rooms) against the cost of walking a guest (compensation, transport, reputational damage, regulatory exposure). Estimate no-shows and cancellations by segment and day, not with one blended rate, and review the level as the arrival date approaches.
Group blocks and stay controls
Group blocks need a cut-off date, a wash factor, and a release rule. Forecast the net block (gross minus expected wash) and release unused rooms early. Minimum-LOS, closed-to-arrival, and closed-to-stay controls protect high-demand nights from one-night bookings that block longer, more valuable stays; use them selectively because they also refuse legitimate demand.
Worked example
High rate 200, discount rate 120, high-fare demand approximately normal with mean 30 and standard deviation 10; protection level 25. P(demand > 25) ≈ 0.69. Expected high-fare value of the marginal room = 200 × 0.69 = 138, which exceeds 120, so accept the discount booking. If the protection level were 34, P(demand > 34) ≈ 0.34 and expected value = 68, below 120, so reject it. The protection level, not the discount rate, decides the answer.
Traps
- Protecting too many rooms for demand that never arrives.
- Setting overbooking without quantifying walk cost.
- Using a blended wash factor for groups with different cancellation behavior.
- Applying closed-to-arrival on a shoulder night and losing a high-value stay.
- Ignoring segment differences in no-show rates.
- Treating upgrades as free; they change room-type availability and value perception.