# Revenue management and yield: pricing the same seat ten ways
Two passengers sit side by side in economy on the same flight. One paid $89. The other paid $612. They have the same legroom, the same soda, the same arrival time. Neither is being cheated. Both are exactly what the airline wanted.
This is revenue management (RM): the discipline of selling a fixed, perishable inventory at the highest total yield the market will bear. A seat that flies empty is worth zero forever. A hotel room unsold tonight cannot be sold twice tomorrow. That "use it or lose it" pressure is why travel and hospitality became the birthplace of modern dynamic pricingdynamic pricingAutomatically adjusting prices in real time based on demand, competition or user behaviour to optimise revenue, margin or conversion.Voir la définition complète →.
Three conditions make RM powerful, and travel has all three:
American Airlines pioneered these techniques in the 1980s under the name "yield management" to fight low-cost entrants. The core insight: do not set one price. Set many prices, and control how many seats you release at each one.
Yield here means revenue per unit of capacity. Airlines track RASM (revenue per available seat mile). Hotels track RevPAR (revenue per available room, occupied or not). RevPAR is the number that decides whether a hotel general manager keeps their job.
> RevPAR = Average Daily Rate (ADR) x Occupancy
A hotel at $200 ADR and 60% occupancy earns $120 RevPAR. A hotel at $150 ADR and 90% occupancy earns $135 RevPAR. The cheaper hotel is winning. That tension between rate and occupancy is the entire game.
If an airline simply offered a $89 fare, everyone would buy it, including the business traveler who would have paid $612. The airline must build fare fences: rules that make the cheap fare inconvenient for high-value customers so they self-select into higher fares.
Classic fences:
The airline is not pricing a seat. It is pricing a bundle of restrictions attached to a seat. This is segmentationsegmentationDividing a market into distinct groups of customers who share similar needs, characteristics or behaviours, so each group can be served with a tailored approach.Voir la définition complète →: sorting customers by willingness to pay and serving each group a different offer.
Every flight and every hotel night has a booking curve: the pattern of how reservations accumulate over time before departure or arrival.
Leisure bookings arrive early and cheap. Business bookings arrive late and pay full fare. RM systems forecast the final shape of that curve from historical data, then decide how many seats to protect for late, high-value demand.
This is inventory control. If the forecast says 30 business travelers will book in the final week at $600, the system will refuse to sell those 30 seats at $89 today, even though the plane is half empty and the temptation is real. It "protects" inventory for higher-yielding demand still to come.
The math behind seat protection traces to a rule called Littlewood's rule: keep selling the cheap fare only while its certain revenue exceeds the expected revenue from holding the seat for a possible high-fare buyer. Modern systems generalize this across many fare classes at once.
Forecasting errors are expensive in both directions. Protect too many seats and the plane departs with empty premium seats you could have filled cheaply. Protect too few and you sell out the cabin before the high-fare buyers arrive, leaving money on the table (this is called spoilage or dilution).
For a solid free primer, see the MIT OpenCourseWare Airline Transportation materials, which cover RM fundamentals in depth.
Some travelers never show up. They miss connections, cancel, or simply do not appear. If an airline sold exactly 180 seats and 12 no-showed, it flies with 12 empty seats and lost revenue permanently.
So airlines overbook: they sell more seats than the aircraft holds, betting on a predictable no-show rate. This is a straight financial calculation weighing two costs:
RM systems set the overbooking level where expected total cost is lowest. When the bet goes wrong, airlines first ask for volunteers with vouchers (a voluntary denied boarding), then, rarely, bump passengers involuntarily.
Passenger rights here are regulated. In the US, the Department of Transportation sets denied boarding compensation rules. The EU has its own framework (EU261). Overbooking is not a loophole. It is a modeled, disclosed practice with legal guardrails, and after several high-profile incidents most carriers now lean heavily on voluntary bumping.
Hotels play the same game with an extra dimension: length of stay (LOS). A room is not just occupied or empty. It is occupied for a certain number of nights, and which nights matter enormously.
Imagine a city hotel with heavy Tuesday and Wednesday business demand and weak Sunday and Monday. A guest wants one night, Wednesday only. Should the hotel accept?
Maybe not. Selling Wednesday alone might block a three-night guest (Monday to Thursday) whose stay fills the weak nights too. So hotels apply LOS controls:
The financial goal is total-stay value, not one-night rate. A hotel will sometimes reject a high-rate one-night booking in favor of a lower-nightly-rate multi-night booking that yields more revenue overall and fills otherwise-dead inventory.
Hotels also manage across channels. A room sold direct on the hotel's site keeps more margin than one sold through an online travel agency (OTA) that charges a commission (commonly cited in the mid-teens percent range, though it varies by contract). RM and distribution strategy are joined at the hip.
Vérification des acquis
1. Why is travel and hospitality considered the natural birthplace of revenue management?
2. A hotel with a lower average daily rate but higher occupancy can still outperform a competitor with a higher rate. What core tension does this illustrate?
3. Why does an airline refuse to sell every seat at its lowest available fare, even though doing so would fill the plane?
4. Select ALL correct answers about the conditions that make revenue management effective.
Sélectionnez toutes les réponses correctes.
5. Select ALL correct answers about yield metrics used in the sector.
Sélectionnez toutes les réponses correctes.
Modern RM is a continuous loop running across thousands of flights or room-nights at once:
1. Forecast demand by segment along the booking curve.
2. Optimize how much inventory to release at each price and each restriction level.
3. Control availability in real time as bookings arrive and the forecast updates.
4. Reprice dynamically, nudging fares up as high-demand dates fill and down as soft dates lag.
The rise of continuous pricing and machine learning has pushed beyond the old lettered fare buckets. Instead of ten fixed prices, some airlines now compute a tailored price per query. But the logic is unchanged: match willingness to pay against scarce, perishable capacity.
A caution for finance professionals: yield optimization can collide with brand trust. Aggressive dynamic pricingdynamic pricingAutomatically adjusting prices in real time based on demand, competition or user behaviour to optimise revenue, margin or conversion.Voir la définition complète → during emergencies, or fare fences customers find manipulative, create reputational and regulatory risk. The best RM balances short-term yield with long-term customer lifetime valuecustomer lifetime valueLifetime Value: the total revenue (or profit) a customer generates throughout their entire relationship with your business.Voir la définition complète →.