the issue with traditional rm - eyefortravel... smart travel analytics the issue with traditional rm...
TRANSCRIPT
www.simon-kucher.com
Smart Travel Analytics
The issue with Traditional RM
Amsterdam, Nov 2016
Dimitris Hiotis
London office1 Plough PlaceLondon EC4A 1DE, UKTel. +44 20 7832 6700 [email protected]
The issue with Traditional RM eye for travel.pptx
RM Bingo (very old-school RM) to look for
2
B I N G O
200
B I N G O
99
Let’s start our story with a quite sophisticated RM system
…years to develop and implement
…thousand prices changed a day
% were set at user-defined parameters
1,000s of booking curves Network optimisationPrice elastic dynamic forecasts
£500 £525 £550 £575 7
Optimal price:£1,089
B I N G O
4
The issue with Traditional RM eye for travel.pptx
0%
20%
40%
60%
80%
100%
3
If something is a black box
£1,040 Max
£1,000 Min
£1,040 Max
£1,000 Min
Price can be within..
The issue with Traditional RM eye for travel.pptx
So why were 99% of prices at pre-defined parameters
Sophisticated RM may be very clever, but if people do not get it they end up controlling it too much, overriding its sophistication
..you start not trusting its outcome
..so you start putting parameters to control it
..and your parameters become the answer
4
..and such controls are needed to avoid stories like this one
Automation needs control and human intervention – else we could end up in in unrealistic price
Robotic pricing can lead to insane price points§ Two booksellers were
setting up prices based on each other
§ As one increased the price the other one followed
§ In a period of 10 days price spiralled to $23 million
The issue with Traditional RM eye for travel.pptx 5
…and here is a story of a simpler system
Airport car-park daily pricing tool
0
5
10
15
20
25
-
200
400
600
800
1,000
1,200
1,400
-60 -58 -56 -54 -52 -50 -48 -46 -44 -42 -40 -38 -36 -34 -32 -30 -28 -26 -24 -22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0
PriceOccupancy
Lead timeActual SDs Target Price
The issue with Traditional RM eye for travel.pptx 6
…and here is a story of a simpler system
Airport car-park daily pricing tool
0
5
10
15
20
25
-
200
400
600
800
1,000
1,200
1,400
-60 -58 -56 -54 -52 -50 -48 -46 -44 -42 -40 -38 -36 -34 -32 -30 -28 -26 -24 -22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0
PriceOccupancy
Lead timeActual SDs Target Price
The issue with Traditional RM eye for travel.pptx 7
…and here is a story of a simpler system
Simpler models can be quicker, as effective and set you on a path for more sophisticated RM later
Airport car-park daily pricing tool
0
5
10
15
20
25
-
200
400
600
800
1,000
1,200
1,400
-60 -58 -56 -54 -52 -50 -48 -46 -44 -42 -40 -38 -36 -34 -32 -30 -28 -26 -24 -22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0
PriceOccupancy
Lead timeActual SDs Target Price
B I N G O
20%
B I N G O
6%
…weeks to develop
…increase in occupancy
…increase in revenue
B I N G O
8
The issue with Traditional RM eye for travel.pptx 8
RM capability should be built in steps with Traditional RM setting the foundations for later sophistication
Traditional RM Sophisticated RM
Simple Price differentiation
1
Rules based RM
2
Forecast based YM
3
Network optimisation
5
Customer choice based YM
4
The issue with Traditional RM eye for travel.pptx 9
Peak Peak Peak
-
50
100
150
200
250
300
-
2
4
6
8
10
12
14
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Hour of day# taxi rides B2C waiting time
Simple price differentiation is your first step
The issue with Traditional RM eye for travel.pptx
Waiting time Demand (indexed)
Off-peak(-25%)
Off-peak(-25%)
Off-peak(-25%)
§ Taxi provider had flat-pricing across the day
§ This led to high waiting times at peak periods of demand
§ …and low waiting times at off-peak periods, where demand was not in sync with supply
§ Solution:§ Introduce peak/off-peak
pricing§ Peak +5% than today§ Off-peak -25% than today
§ Result:§ Peak – better waiting times
maintaining demand§ Off-peak increase in demand
& revenue
Project Example
1
10
A rules based price approach naturally follows
Lead time based differentiation
10%
20%
35%
50%
60%
70%
80%
OTD (on the day)
Level 7
Level 6
Level 5
Level 4
Level 3
Level 2
Level 1% of capacity or #seats
£ Price points at each level
Close out (in days to departure)
a
b
c
£10
£17
£25
£35
£45
£54
£69
£85
7 2 1
Used by train operators, coach operators and traditionally by low-cost airlines
Three key levers that essentially revenue manage sales
§ % occupancy – have price build-up based on capacity utilisation
§ Price points – Start low and go-up as you move closer to departure
§ Close out rules ensure low prices are not available close to departure
a
b
c
2
The issue with Traditional RM eye for travel.pptx
Project Example
11
Below is an illustration of how we set-up a rate of sale forecast for a tour operator
A forecast-based system requires booking curves to generate forecasts based on market dynamics
Days to departure
Main components:a. Booking curve that shows expectation of demand at current price
b. Sales to date
c. Target (capacity)
Rate of sale forecast (last 4 weeks example)
§ Expected 10 pax booked in the last 4 weeks
§ Instead I got 16 pax booked
§ Rate of sale is 1.6 (16/10) times what I expected
§ From now till departure I expect to get 50
§ If I apply this recent of rate of sale expectation is 50*1.6 = 80
§ This will translate to a forecast of 56 booked + 80 expected = 136
NOTE: Typically 2-3 rate of sale forecasts are created, each with different base (e.g. last day, last 7 days, last month) and a weighted avg is used to generate a forecast
a
b
c
0180
100
50
TodayToday –4 Wks. Ago
40
56
1016
130
booking curve
actual
forecast
Expected in the last 4 weeks
Booked in last 4 weeks
a
b
c
Project example
3
The issue with Traditional RM eye for travel.pptx 12
Forecast based system can be combined with margin rules to drive pricing decisions
Resort Egypt – Sharm
Budget month 2014_08
Budget week Week 34
Flight MXP-SSH
Day of week Saturday
Current occupancy 91%
F’cast occupancy 197%
Seats capacity 100Hotel Capacity Fore-
castContrib-
utionMargin
Pax booked
Total pax forecast
Forecastpax (to come)
Apportion Comula-tive build
up
% build
up
Pricedecision
Hotel 1 5 20% 967€ - - - 0% 69 91% DOWNHotel 2 10 0% 746€ - - - 0% 69 91% DOWNHotel 3 272 98% 597€ 45 76 31 39% 100 132% UPHotel 4 5 64% 534€ 2 2 - 0% 100 132% STAYHotel 5 30 68% 520€ 2 6 4 5% 104 137% STAYHotel 6 30 120% 340€ 20 65 45 56% 150 197% UPHotel 7 50 80% 120€ - - - 0% 150 197% UP
69 150 81
A
B C
§ System produces forecasts for each flight on the programme as a % of capacity (>100% flight will fill out) based on current rate of sale.
§ Equally, system produces forecasts for the hotels that will be sold with each flight, and the resulting margin per passenger each booking could bring
§ Based on forecasts and expected margin per person, rules based engine recommends a price change (e.g. UP – put price up; DOWN – put price down; STAY: Leave price as it is)
A
B
C
Forecast occupancy of flight à>100% flight will sell out
Forecast occupancy of hotel and expected margin per passenger
Rules based pricing based on forecasts of hotel and flight
Project ExampleTour operator
3
System produces a forecast for both the flight and the hotel components of package
Based on forecast of flight & hotel, price recommendation is made
Crucially, margin is also used to determine where to best increase price when relevant stock is constrained (e.g. flight capacity)
The issue with Traditional RM eye for travel.pptx 13
A customer choice model enhances forecasts with insights from external market
Project ExampleTour operator
Rate architecture Competitor pricesForecasts of demand (based
on RoS & booking curves)
Sales to date (pax, margin, price)
Demand data(searches)
Availability(failed searches)
Traditional RM elements Enhancements
4
Forecasts are enriched by a selection of KPIs which feed into pricing decisions:§ Price positioning vs.
competitors§ Web demand – are my
forecasts low due to low demand or high price?
§ Failed searches (i.e. do I have less bookings because my availability is restricted rather than my price is too high)
The issue with Traditional RM eye for travel.pptx 14
A network optimiser optimises the revenue mix based on forecast & predicted impact of pricing action
Pkg #3 Pkg #5 Pkg #6Pkg #2 Pkg #4Pkg #1
Price
15 50 80 100 120
450400
350300
250200
CAPACITY
Pkg #3 Pkg #5Pkg #6Pkg #2 Pkg #4Pkg #1
Price
15 50 80 100 120
450400
380
360 320
275
CAPACITY
A network optimiser will compare forecast to capacity and accounting for all interdependencies (e.g. flight to hotel and hotel to flight or even between hotel nights) will recommend price changes
Optimisation is done in multiple dimensions (outbound flight, inbound flight and each stay-night at the hotel)
Margin is crucial as optimiser will tend to price-up the lower-margin products to maximise overall margin
Project ExampleTour operator
5
The issue with Traditional RM eye for travel.pptx 15
Whatever you choose, you need a process, a set of reports and a team to yield manage
Mon Tue Wed Thu Fri W/E Mon
Market monitoring
Price management
Capacity changes
Demand shocks
Process Tool Decision
1
2
3
4
Daily pricing
B2S capacity changes
B2S re-balancing
Supply based pricing
Forecast based pricing
Buy
Sell
Trade
Weekly capacity management
Late deal alert
Price
Daily pricing Daily pricing Daily pricing Daily pricing
Daily pricing
Weekly capacity management
Supply change alerts
Drill Down tool
Project ExampleTour operator
The issue with Traditional RM eye for travel.pptx 16
Conclusion and key take-aways
§ Sophisticated RM is clever and can be quite revenue optimal
§ However, it needs to be understandable and believable to users, as else their sophistication is wasted
§ To do this you need to develop RM capability in steps
§ First start with the traditional RM principles, from simple price differentiation to rules-based RM
§ …before you move to forecast based, Customer choice and/or network optimiser
§ And don’t forget à Process, Process, Process
§ This allows you to:- Build the right RM culture in your organisation- Find the right algorithms and methods for your business
through trial & error- Only move to hyper-sophistication when you feel there is
extra benefit
The issue with Traditional RM eye for travel.pptx 17
Thank you!
Rosalind HunterDirector
1 Plough Place EC4A 1DE London
United Kingdom
Tel. +44 20 7832 67 00Fax +44 20 7832 68 00
Dimitris HiotisPartner
Tel. +44 20 7832 67 00 Fax +44 20 7832 68 00
1 Plough Place EC4A 1DE London
United Kingdom
18