By the late 1980s American Airlines had spent a decade proving that yield management — forecasting demand and dynamically fencing discount fares behind advance-purchase and Saturday-night-stay rules — could add hundreds of millions of dollars a year without adding a single seat. A hotel room is the same perishable asset as an airline seat: unsold at midnight, that night's revenue is gone forever, and it can never be resold. Hotels had nonetheless kept pricing rooms the old way, with a single rack rate and blunt discounting, because nobody had built the forecasting discipline the airlines had.
Marriott's rooms division carried the same problem at a much larger scale than any single carrier's fare grid: thousands of properties, each with its own local demand curve and competing pressure to just discount empty rooms rather than model who would show up and pay full price later. The obvious answer — cut rates harder whenever occupancy looked soft — treated every empty room as a loss to minimize today rather than inventory to be sold at the right price to the right guest.
Imported airline revenue management wholesale: forecast demand per hotel per night, fence discounts with advance-purchase rules, pricing perishable rooms.
Rooms are perishable like seats, so unsold nights are lost forever.
Forecasting demand per night lets you hold price for late full-price guests.
Fencing discounts with advance-purchase rules stops cannibalizing high-paying guests.
When your product perishes at a fixed time, forecast demand and fence discounts by segment instead of reflexively cutting price.
Marriott's own accounting has credited its yield management system with about $100 million of additional revenue a year, achieved mostly by reallocating existing room inventory across customer segments rather than by adding capacity or cost.
FOLLOW THE EVIDENCE
The sources
- Introduction to the Theory and Practice of Yield Management INFORMS Transactions on Education · pubsonline.informs.org
- Marriott International Increases Revenue by Implementing a Group Pricing Optimizer Interfaces (INFORMS), via RePEc · ideas.repec.org