Revenue Management

Demand forecasting, pricing optimization, and inventory control systems.

Scope

Demand Forecasting

  • Historical booking analysis
  • Seasonal patterns
  • Event detection
  • Competitive factors
  • Macro-economic inputs

Pricing Optimization

  • Price elasticity modeling
  • Competitive pricing analysis
  • Dynamic price adjustments
  • Fare family optimization
  • Ancillary pricing

Inventory Control

  • Booking limits by class
  • Overbooking levels
  • Nested inventory management
  • O&D optimization
  • Bid price calculation

Network Optimization

  • Route profitability
  • Schedule evaluation
  • Hub strategy
  • Fleet assignment input

Research Topics

  • RM system vendors (PROS, Sabre)
  • Machine learning in RM
  • Continuous pricing integration
  • O&D vs. leg-based RM
  • Overbooking optimization
  • Group pricing strategies
  • Ancillary RM
  • Competitive intelligence

Architecture Considerations

RM System Architecture

Data Collection
├── Historical bookings
├── Competitive data
├── Events calendar
└── Market data
        │
        ▼
┌───────────────────┐
│ Demand Forecaster │
└─────────┬─────────┘
          │
          ▼
┌───────────────────┐
│    Optimizer      │
└─────────┬─────────┘
          │
    ┌─────┴─────┐
    │           │
┌───▼───┐ ┌─────▼─────┐
│ Auth  │ │ Bid Price │
│ Levels│ │ Calculator│
└───────┘ └───────────┘
          │
          ▼
   PSS (Inventory)

Data Model

Forecast
├── FlightKey (Carrier, Number, Date)
├── ForecastDate
├── DemandByClass[]
│   ├── BookingClass
│   ├── UnconstrainedDemand
│   └── Forecast
├── TotalDemand
└── Accuracy metrics

BookingLimit
├── FlightKey
├── EffectiveDate
├── ClassLimits[]
│   ├── BookingClass
│   ├── AuthorizedUnits
│   └── ProtectedUnits
├── OverbookingLevel
└── OptimizationRun

Optimization Methods

MethodUse Case
EMSR (Expected Marginal Seat Revenue)Single-leg optimization
Network RMHub operations, connections
Bid PriceRevenue threshold per seat
Dynamic ProgrammingComplex constraints

Integration Points

SystemDirectionData
PSSBidirectionalInventory, bookings
PricingOutboundPrice adjustments
ScheduleInboundFlight schedule
Competitive IntelInboundCompetitor prices
BIOutboundPerformance data

Continuous Pricing

Integration with NDC

Shopping Request
        ↓
RM System Query
├── Demand signal
├── Competitor prices
├── Inventory position
└── Customer value
        ↓
Price Adjustment
├── Base fare
├── Adjustment factor
└── Confidence level
        ↓
Offer with Dynamic Price

Enablers

  • Real-time demand signals
  • Price elasticity models
  • A/B testing framework
  • Approval thresholds

Overbooking Management

Factors

  • Historical no-show rates
  • Booking curve position
  • Flight importance
  • Recovery options
  • Denied boarding cost

Model

OptimalOverbooking = f(
  ExpectedNoShows,
  ShowRateVariance,
  DeniedBoardingCost,
  SeatRevenue,
  GoShowProbability
)

Performance Metrics

RM KPIs

MetricDescription
Revenue upliftRM vs. baseline
Forecast accuracyMAPE, bias
SpoilageEmpty seats sold out
SpillDemand turned away
Denied boardingsOverbooking errors

Benchmarks

  • Forecast MAPE: <15%
  • Revenue uplift: 3-8%
  • Denied boarding rate: <1 per 10,000

Vendor Landscape

RM Systems

VendorProduct
PROSO&D RM, Dynamic Pricing
SabreAirVision Revenue Manager
AmadeusAltéa Revenue Management
IBS SoftwareiFlight RM

Build vs. Buy

  • Buy: Proven algorithms, faster deployment
  • Build: Full control, competitive advantage
  • Hybrid: Core RM + custom pricing layer