Upload
truongthu
View
216
Download
1
Embed Size (px)
Citation preview
The Analytics of LoyaltyThe Analytics of Loyalty –Qantas Frequent Flyerq y
Vaughan Chandler and Wade Tubman© Qantas / Quantium
Qantas Frequent Flyer
COALITION LOYALTY PROGRAM RELATED BUSINESSES
• 8 million+ members – 50% household penetrationB d t t hi
• Operates over 100 loyalty programs• Scalable marketing and analytics
l tf
COALITION LOYALTY PROGRAM RELATED BUSINESSES
• Broadest earn partnerships• Unrivalled partner network, redemption options and Qantas assets
• Program design creates considerable i d f Q
platform• Significant opportunity exists to leverage IP in adjacent industries
strategic advantages for Qantas
Vii Loyalty
Driving Growth in the Core Innovate and Expand along the Loyalty Value Chain
2
The coalition engagement model
WorldBEHAVIOURAL OUTCOMES
World leading partner profile
Customeracquisition
Improved retention
Increased Improved
Valued member proposition profile share of wallet
pmargin
p p
• Simple• Achievable• Desirable
• Reach• Breadth• Brands
• Innovation and technology• Communication channels• Leverage assets• Community• Community
3
Qantas Frequent Flyer Business Model
Billings of over $1bn
Cash / deferred
revenue of over $2bn
EBIT of over $340mn
over $2bn
4Source: 2010 Qantas investor briefing and Qantas Data Book 2011
The evolution of Qantas Frequent Flyer
Fundamental shift of business model in 2007 World
leadingCustomer Engagement and Fi i l b fi• Shift to Coalition model
•Position for growth into mass market
leading partner profile • Continual Evolution of the business, for • Record membership satisfaction
Financial benefits
example:• Platinum One• Auto‐rewards• epiQure by Qantas Frequent Flyer
• 60% increase in EBIT• 50% increase in membership• Over 3.5 million award seats redeemed and 400,000 items of merchandisep Q y Q q y
• New comms channels,
Analytics Capabilities
5
Analytics CapabilitiesLoyalty design
analyticsPredictive and event modelling
Campaign optimisation
Customer behaviour analytics
Qantas Frequent Flyer is unique
AUD* MultiplusAIMIA
(Nectar and Aeroplan)
Qantas Frequent Flyer
2010 2010 2010
Core Markets Brazil UK and Canada Australia
Billings per capita of core markets $3.50 $21.50 $41.50
Estimated Breakage Rate 23% 18% <10%
Adjusted EBITDA per capita $0.89 $2.53 $7.30
6
Market Penetration of core markets 4% 34% 37%
* converted to AUD at spot rateSource: Annual reports and QFF analysis
Systems and data are not the end goal
Organisation structure
Organisational Culture
Commercial acumen
These elements are all required to
Measurement and KPIs
Knowledge management
qbe in balance to optimise valueKnowledge management
Investment
optimise value from data
7
Process and practice
Case Studies : QFF continues to break new ground in the loyalty sector by combining internal assets with proven analytical approaches and external data
Case Studies : QFF continues to break new ground in the loyalty sector by combining internal assets with proven analytical approaches and external data
1. Setting liabilities through customer engagement modelling
2. Targeting customer retention efforts with predictive modelling
Case Study 1 - Customer engagement modelling
• Program liabilities and the release of revenue are dependent on the eventual use of each point
• Over the long term all points are either redeemed or expire unused• Over the long term, all points are either redeemed or expire unused• Modelling future points expiry (breakage) helps inform the setting of
assumptions by management• There is a balance between customer engagement and breakage, between
long term customer value and shorter term revenue
Modelling techniquesModelling techniques
Many international loyalty programs have “life of point” where individual points
Breakage is often modelled using pension
life of point , where individual points expire, say, 3 years after they are earned
g g pfund techniques
The QFF program is based on member
Breakage is modelled with a multi-state
The QFF program is based on member engagement, so customer activity levels
are the key determinant of breakage
Source: Quantium
Breakage is modelled with a multi state transition model based on the level of
customer activity
Learnings for financial servicesLearnings for financial services
Whole of business focus Holistic customer viewContinual assessment
Coordinated executive level committee – finance,
strategy, analytics, Consider customers
across multiple dimensions
Ongoing monitoring of customer activity and touch
pointsmarketing, product etc
Culture and KPIs supporting balanced
– status tier, channels, source, level of
engagement, value, activity etc
Actuarial assessment of customer engagement
implications of keysupporting balanced outcomes
activity etcimplications of key business decisions
OutcomeApproachObjective
Case study 2 - Predictive retention modelling
Targeted surprise and
Outcome
1.Market Blueprint® used to identify loyal and
Approach
Identify customers at risk
Objective
Textg p
delight campaigns to most valuable customers before the “defect”
y ydisloyal behaviour, and measure potential customer value
Identify customers at risk of disloyalty to focus retention and winbackinvestment
Text
Text
TextAbility to respond to disloyalty on an ongoing basis
Build a predictive model to rank customers on their risk of “defection”
2.Calibrate model to QFF customer variables (both internal and external) tointernal and external) to score loyalty for individual members
The loyalty / value trade off
Hig
h
ue
Focus retention efforts on those
with most potential Ideal
customer
tom
er V
alu value
LowC
us Upsell to these highly loyal
customers to increase value
Low HighLoyalty
increase value
Number of Points Redeemed on Domestic Classic Awards Redeeming rewards generates loyalty
54%
60%
eing
Qan
tas
42%
48%
of n
ext f
light
be
36%
42%
Like
lihoo
d o
30%
Points redeemed on Classic flightsSource: Quantium
Lounge Access Flag (Qantas Club and Higher Tiers)Lounge access generates loyaltybe
ing
Qan
tas
d of
nex
t flig
ht
Like
lihoo
d
No lounge access Lounge access
Lounge access - Qantas Club and high tiersSource: Quantium
Supermarket Loyalty ScoreLoyalty carries across industriesbe
ing
Qan
tas
d of
nex
t flig
ht
Like
lihoo
d
58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 100
Supermarket loyalty - by percentileSource: Quantium
Focusing retention efforts - customer value overlay
Source: Market Blueprint ®
Fitted versus ObservedOutcome : Individual customer risk of defection
eing
Qan
tas
Focus retention efforts
of n
ext f
light
be Focus retention efforts
on customers with 5-10 times the risk of others
Like
lihoo
d o
0 10 20 30 40 50 60 70 80 90 100Model Percentile
Ranked IndividualsSource: Quantium
Learnings for financial servicesLearnings for financial services
Predictive modelling Ongoing monitoringRoot cause analysis
Predictive models allow retention activities to be
Implementing retention risk measurement on CRM Multivariate models help
understand the drivers ofretention activities to be targeted on the highest
value customers at risk of churn, maximising ROI
systems allows ongoing monitoring and rapid
response at an individual customer level
understand the drivers of churn behaviour, informing
the design of potential retention initiatives customer level
C l i H li ti i f th t• Traditional customer, product
and transactional data• CRM data customer profiles activity
1. Effective use of internal data
Conclusion: Holistic view of the customer
• CRM data – customer profiles, activity, touchpoints
internal data
• Obtain a view of customers and their behaviours outside your organisation2. Exploiting external
Maximising customer
loyalty and l• Predictive modelling at an individual3 Extensive customer
y g• Monitor, understand and capitalise on
emerging trends
p gdata sources
valuePredictive modelling at an individual customer level for acquisition, cross-sell and retention
3. Extensive customer focused modelling
• Supported by customer centricSupported by customer centric management and culture
• Ongoing investment in knowledge, systems and processes
4. Customer centric organisational focus
Vaughan ChandlerHead of Strategy and AnalyticsHead of Strategy and AnalyticsQantas Frequent Flyer
Wade TubmanWade TubmanPrincipalQuantium