Title: Proprietary and Confidential
1Two Pricing Optimization Applications at Amtrak
ShareFares Amtraks Small Group Fare
OneTouch Amtrak Group Optimization Model
2ShareFares - Background
- Almost 96 of Amtrak customers in the northeast
corridor travel by themselves or with one other
person - Current Fares discourage groups larger than 2
- We d charge from 68 to 120 one-way off-peak
travel between Washington and New York on or
Regional trains. - The one-way fare for groups of 3 or more exceeded
200. - Most families are more likely to take the bus or
drive. - ShareFares includes a deeply discounted fare for
the 3rd through 6th traveler in each group - 90-off for the 3rd, 4th, 5th, and 6th rider
- Even with the 90 discount, Amtraks incremental
revenue has been significant, because it includes
the first two riders at full far.
3Pricing Example NYP to WAS
- The first two riders pay the prevailing price.
- Starting with the 3rd rider, the average price
per rider drops significantly, especially for
larger groups. - For a group of all adults, a group of six save
60 (274) from what we are charging today. - Before ShareFares, a group of two adults and four
children paid 304. With ShareFares they pay
182, a 40 discount.
4Dilution Has Been Minimal
- Almost 96 of Regional reservations were for only
1 or 2 riders. - Dilution was minimal, due to offer only applying
to groups of 3 or more. - Dilution was further reduced with a 7-day APR.
5ShareFares has Been Very Successful
- In its first 12 months, it generated over 4
million in incremental revenues - Displacement is controlled with Revenue
Management - Dilution is minimal due to small customer base of
groups of 3 or more and use of 7-day APR - This was accomplished with little advertising (we
couldnt afford it!) - With dilution and displacement under control, we
could afford to let demand build by word of mouth
and PR - It is an Active Promotion We encourage or
reservation and sales agents to offer it whenever
it fits - We recently have changed the discount to 75 to
measure price elasticity
6Group Pricing - Background
- In FY05, Amtrak received approximately 14
million in group business - This is a little over 1 of total revenue (1.3
billion) - However, Revenue Managers spent 15 to 20 of
their time on pricing groups - Their efforts were primarily defensive,
protecting higher-demand departures from selling
out due to to many through-seats sold to short
distance groups - Group decisions were done manually, resulting in
a bias to keep all groups away from higher demand
departures - Requests took 24 to 48 business hours to be priced
7Solution OneTouch Amtraks Group Optimization
System
- OneTouch solved three problems
- Speed of Response Now most group requests are
priced within 30 seconds - Better utilization of resources
- Reservation center operators do not have to call
back customers to quote the price - Revenue Managers now spend less than 1 hour
monthly reviewing pricing decisions (versus hours
each day) - Pricing is optimized, increasing group business
8OneTouch fixed Manual Pricing Bias Against Groups
- Most Revenue Managers saw all groups as a risk
not worth taking unless there was no risk of
selling out - Groups traveling longer distances (with higher
fares) were automatically refused, or priced at
highest bucket - Revenue Managers did not consider that many times
the average price at the peak leg was well below
the lower buckets of the longer distance markets - OneTouch both forecasts non-group demand to
determine displacement risks as well as optimizes
the peak leg seats when displacement is likely - The length of market (and value of fare) plays a
bigger role than bucket in optimization decision.
9Overview of Optimization Process
- Leg Value Computation.
- Demand forecasting at leg level.
- Group Evaluation
- Displacement
- Displacement Value
- Group value
- Group Denial or Acceptance
- Future Enhancements
10Leg Value Computation
- A leg demand has a composition of different
markets - High demand short haul market
- Low demand long haul market
- Demand farther out in the booking cycle
- Demand closer in to the booking cycle.
11What is it Worth?
40 5 Legs
90 12 Legs
80-10 Legs
150 16 Legs
TOL
CUM
PGH
CLE
CHI
CHI
HMI
12Group Evaluation - Displacement
Displacement 15, Non Displacement
10 Displacement Value 76 15 Non
Displacement Value Lowest fare for the Market
10 Group Value Displacement Value Non
Displacement Value
13Future Enhancements
- Expected Demand based on the time of arrival of
the group request - Value of the leg as a function of days before
departure - Step-wise displacement