IBM Finance Forum - Prediktivt tänkande i den finansiella världen
Portrait uplift and optimizer deck
1. Understand, Visualize and Predict Customer Behavior
The CRM “Customer-centricity Chasm”
Customer Interaction Hub Illustrated Demo
Customer Case Study Snap Shot
Market Leading Analyst Recognition
INSIGHT ACTION
Home
2. Customer Analytics Maturity Model
• Commence at any stage Real-time Decisioning
• Benefit from a clear • Inbound contextual RTD
ACTION
roadmap to success • Adaptive learning
Real-time Scoring
• Eliminate customer lists, mass-scoring
• Up to the second outbound selection accuracy
Optimize Contacts
• Analytically match customers with right offer, at right time,
via right channel, balanced against business constraints
Predict Uplift
• Predict persuadability; eliminate sure things, lost causes, sleeping dogs
INSIGHT
• Reduce spend 40% while improving outcome 30-300%
Understand & Predict Behavior
• Predict and define ideal target customer segments
• 3D visual customer data mining & automated modeling
Explore Customers
• Uncover areas of interest and opportunities to act
• Ad hoc, thin-client customer data exploration for the business user
(Underpinned by customer data access, generic BI & reporting tools)
INSIGHT ACTION
Home
3. The CRM “Customer-centricity Chasm”
DATA INCREMENTAL CRM CRM TOUCH POINTS
Right message, right time
Customer Profile Dialogue across touch points
Optimize engagement
“Line of
sight”
Transaction
Infuse customer-specific
insight into touch points
Channel
Lines of Business Understand past behavior
Predict future needs
“Hindsight”
“Hindsight” “Foresight”
“Foresight”
What | Why | Next | Change
Customer Preference
INSIGHT ACTION
Home
4. Optimized Customized Customer Interactions
Customer We sent you
Customer completes an email
Documents
survey generated
yesterday –
and mailed
do you have
any
Thank you for questions?
updating
your
preferences!
Offer
acceptance
feedback
Ideal promo: survey
Unlimited
Please Text
update Last day of Value
your Unlimited reinforcement
preferences Text - you saved $X
promo this month
Your Business
The Results
• Improved customer experience
• Increased customer satisfaction and advocacy
• Increased customer value, wallet share
• Improved customer retention
• Single source for customer preferences
INSIGHT ACTION
Home
4
7. Optimizing Customer Interactions
Mobile Web
Kiosk /
Email INTERACTIVE
PROACTIVE ATM
Mail IVR
Contact
Center
“Outbound” “Inbound”
Channels Channels
INSIGHT ACTION
Home
8. The Portrait Suite
Fulfill.
Recommend.
Orchestrate.
Optimize.
Predict.
Understand.
Explore.
INSIGHT ACTION
Home
9. What’s it Worth?
Financial Services Telecommunications
Products held per 36% increase in churn reduction
customer up by 50%
Customer retention up >20% 20% reduction in churn
Immediate increase in customer retention
Incremental revenue up 327% Reduced marketing costs >DK100,000 PA
Reduced marketing costs by >10% Churn decreased by 37.5%
ROI from a single campaign Emerging Markets
Increased revenue by $1.2M per month
Insurance
Customer retention up 300% Incremental revenue of £48M
“Immediate increase in renewals” Profitability rose 10%
Uplift doubles customer retention Revenue per customer 314% higher
INSIGHT ACTION
Home
10. Home
Information Quality Data Quality Tools Data Integration Tools
INSIGHT
DATA
Predictive Analytics & Data
ACTION
Mining Customer Data Mining
5 Magic
Gartner
The ONLY
Quadrants
INSIGHT
vendor on all
DQ
DI
Cross-channel Comms Mgmt Multi-channel Comms Mgmt
MCCM
CCCM
CDM
STRATEGY
Document Output Customer Service Contact Ctrs
The ONLY
vendor on
all 4 Waves
PBBI – Customer Communications Management
IQ
CCCM
DOCCM
PA/DM
COMMUNICATION
11. Forrester’s View of Inbound & Outbound CM
Dialogue
Customer Analytics
Foundation
Interaction
Analytic
Fulfillment
Insight
Interaction Optimizer
INSIGHT ACTION
Home
12. Overview
Case Study: Retention (Telecommunications Example)
Case Study: Cross-sell (Financial Services Example)
Case Study: Cross-sell (Retail Example)
INSIGHT ACTION
Home
13. Portrait Uplift
Predict the “persuadables” that will react to your marketing.
Portrait Uplift empowers you to focus your marketing
INSIGHT
- PREDICT
upon only those customers that will positively change
behavior as result of your message.
Benefits • Award winning solution – Predict change in
• Eliminate behavior
marketing waste Predict which customers are Persuadables, while
weeding out Sure Things, Lost Causes and
• Reduce Sleeping Dogs.
customer ad- • Spend your marketing funds only where it Uplift Segmentation
fatigue, opt-out & will have an impact
attrition Boost impact by 30-300%
Reduce cost by 40%
• Boost return on • Increase analytic bandwidth and throughput
marketing
Automate Uplift modeling best practice
investment
Build more accurate uplift models
in a fraction of the time
INSIGHT ACTION
Home
14. “Uplift analysis is a must-consider
concept for every organization
with significant campaign
management activities”.
Gareth Herschel, Gartner, “Predictive Modeling Now more than Ever”,
Gartner Customer 360, March 2011
INSIGHT ACTION
Home
15. Uplift Segmentation for Retention Marketing
If we target, customer will:
Stay
If we do nothing,
customer will:
Churn Stay
Churn
INSIGHT ACTION
Home
16. Uplift Segmentation for Cross-sell Marketing
If we target, customer will:
Buy
If we do nothing,
customer will:
Not Buy Buy
Not Buy
INSIGHT ACTION
Home
17. Incremental Impact for Response
Positive No impact Negative
120% impact impact 1 Persuadables:
Only buy if treated
Proportion of incremental effect
100%
Put this money 2A
Sure Things:
80% back into your Buy whether you
marketing budget 2 treat or not
60%
1 2 3 2B Lost Causes:
Won’t buy whether
40% you treat or not
3 Sleeping Dogs:
20%
Buy unless you treat
0%
0% 20% 40% 60% 80% 100%
Proportion targeted
INSIGHT ACTION
Home
17
18. Customer Case Studies
Business Challenge: Targeting for customer churn campaigns but lower campaign costs
The Telenor Group is currently ranked as one of the Benefits
world’s largest mobile operators (in terms of
subscriptions). Mobile subscriptions rank at 168 • Traditional targeting reduced churn by
million worldwide with telecom operators in 13 5%, but with significant negative effects
countries. • Uplift model reduced these significantly
• Saving 40% of treatment costs
Portrait Customer analytics assisted Telenor in
delivering next generation targeting for customer • Further reducing churn to 6.8%
retention (churn) campaigns Removing negative effects • Fully automated data build from Teradata
thus decreasing both churn and campaign costs • Sun Solaris server
• 3 m customers, > 300 variables
Results
Without Uplift improved
Reduction in 36%
Overall Churn
With Uplift Introduction of
first prediction Introduction of
model refined
5% 6.8%
25 % prediction model
23 %
20 % 20 %
19 % 19 %
18 % 18 %
17 %
annual defection rate
16 %
Without Uplift 15 % 15 % 14 %
15 %
14 %
14 % 14 % 13 % 13 %
Treated 13 %
12 %
12 % 11 % 11 %
Volume 11 % 11 % 11 %
Treated volume 10 % 10 % 10 %
With Uplift reduced by 40% 9% 9%
8%
9%
5%
60% 100%
0%
months
07
7
07
08
8
7
07
7
7
08
07
07
7
07
07
r.0
.0
.0
l. 0
t.0
.0
v.
b.
b.
g.
n.
n.
n.
p.
s.
ar
ar
ai
ju
ap
ok
no
fe
fe
ja
ju
ja
au
se
de
m
m
m
Konkurrentchurn Eksternchurn
INSIGHT ACTION
Home
19. Interactive Visual Insight
Mobile
subscribers and
defection
distributed across
geographic
regions in Norway
Blue indicating high defection rate
Green indicating low defection rate
INSIGHT ACTION
Home
20. Dramatic Impact to Customer Retention
Introduced
churn model
25 %
23 %
20 % 20 %
19 % 19 %
18 % 18 %
17 %
annual defection rate
16 %
15 % 15 %
15 % 14 % 14 %
14 % 14 % 13 % 13
12 % 11
11 %
10 % 10 %
5%
0%
months
7
07
7
07
7
7
07
07
07
7
r.0
.0
l. 0
t.0
.0
v.
b.
g.
n.
n.
p.
ar
ai
ju
ap
ok
no
fe
ja
ju
au
se
m
m
INSIGHT ACTION
Home
21. Typical Telco Retention Campaign Results
Annual defection rate - Mobile retention campaign
Yes - renewal 3.5 %
Control group 33.0 %
Reject - renewal 51.2 %
0% 10 % 20 % 30 % 40 % 50 % 60 %
• Can we predict “save-ability” rather than churn risk?
INSIGHT ACTION
Home
23. Dramatic Impact to Customer Retention
Introduced
churn model Introduced
Uplift model
25 %
23 %
20 % 20 %
19 % 19 %
18 % 18 %
annual defection rate
17 %
16 %
15 % 15 %
15 % 14 % 14 %
14 % 14 % 13 % 13 % 13 %
12 % 12 %
11 % 11 % 11 % 11 %
11 %
10 % 10 %
9%
10 %
9% 9%
8%
5%
0%
months
07
7
07
08
8
7
07
7
7
08
07
07
07
07
7
r.0
.0
.0
l. 0
t.0
.0
v.
b.
b.
g.
n.
n.
n.
p.
s.
ar
ar
ai
ju
ap
ok
no
fe
fe
ja
ju
ja
au
se
de
m
m
m
Konkurrentchurn Eksternchurn
INSIGHT ACTION
Home
24. Customer Case Studies
Business Challenge: Inability to target the customer with the most relevant message
U.S. Bancorp (NYSE: USB), with $266 billion in assets, is
the parent company of U.S. Bank, the 6th largest Benefits
commercial bank in the United States. Incremental revenue increased 327%
>$500K incremental revenue per campaign
US Bank engaged Portrait Software and its team of experts Reduced baseline cost from lower volume
to produce and trial an Uplift model that had the ability to 189% increase in incremental response by
generate data models which would lead to the incremental avoiding negative effects (downlift)
performance and improved customer experience that US Campaign ROI increased from 75% to 400%
Bank was striving to achieve. in 5 years
Increased incremental cross-sell revenue by
more than 300%
$1,000,000 1000%
$900,000 900%
$800,000 800%
$700,000 700%
Net Income
$600,000 600%
RoI
$500,000 500%
$400,000 400%
$300,000 300%
$200,000 200%
$100,000 100%
$- 0%
Decile by uplift score
INSIGHT ACTION
Home
25. Highlights: Location-data Integration
Business Challenge: Find more cost-effective ways to acquire new DDA customers using direct mail
U.S. Bancorp (NYSE: USB), with $266 billion in assets, is
the parent company of U.S. Bank, the 6th largest : Location Enhanced
commercial bank in the United States. Benefits
Turned around campaign from loss to profit
US Bank use direct mail to entire postal routes to acquire 91% increase in incremental revenue
new DDA customers, but low response rates and lift versus Generate additional $300K per campaign
control made even their best targeting models unprofitable. Can either reduce volume with same net adds,
Portrait Uplift delivered more effective targeting models using or achieve more net new DDAs with same volume
existing data, and achieved campaign profitability by
enhancing with new location-based attributes such as local
product penetration rates and nearest branch drive time.
Incremental Revenue ($’000s)
+91%
Break-even
Location-enhanced
+41%
Uplift Model
Portrait Uplift
Model
Previous
champio
n
INSIGHT ACTION
Home
26. Customer Case Study: Uplift in Retail
Business Challenge: Improve incremental revenue from monthly consumer catalog mailings
Fortune 50 company with $60B+ annual revenues
that sells about 140K PCs / day. It has nearly 2B
customer interactions per year, including about 15M
direct mail catalogs per month. Benefits
Increased revenue by $1.2M/month at existing mailing
Portrait was challenged to develop an Uplift model to volumes, or $2.5M/month with more focused targeting
increase the incremental revenue delivered by monthly Validated in blind trial and now in market, with software
catalog mailings compared to existing response and and services hosted by this company’s MSP
value based targeting. Demonstrated applicability to individualized email
versioning, with $250K additional revenue in a small test
Uplift model delivers
$2.5M more revenue
($500K more profit)
INSIGHT ACTION
Home
26
27. Example reduced monthly mailing volume
In-house Model
Uplift model delivers
incremental lift by 88%
over existing model
Example Reduced Targeting ~30% of mailable universe
INSIGHT ACTION
Home
28. August Mailing Results
August Uplift Results
300%
250%
Incremental Revenue Index
200%
150%
100%
50%
0%
1 2 3 4 5
Quintile
The actual mailing for the Development audience produced $612K incremental
revenue at full volumes, but would have done over $1.3M at the recommended
40%, and over $3.1M over no mail
INSIGHT ACTION
Home
29. Portrait Optimizer
Allocate your marketing for maximum impact & mutual benefit.
Portrait Optimizer is a contact optimization solution that
INSIGHT
- OPTIMIZE
can optimally allocate your marketing spend where it
will have the greatest impact, across campaigns and
customers, for utmost ROI.
Benefits
• Optimize • Analytically weighs your business’ constraints vs. expected benefit
marketing ROI to determine the right mix of offers & channels for individual
customers
• Eliminate • Improves customer-centricity while enforcing Achieve Contact
marketing waste constraints, and improving financial outcome Optimization
Constraints include eligibility, quantity & value Best Practice
• Boost relevancy Considers preferences, propensity & persuadability
• Flexible framework
• Reduce Optimize offers, message, channel, product, incentive,
customer ad- time frame and more
fatigue & • Dual interface
opt-out Foster alignment across marketers and analyst teams
INSIGHT ACTION
Home
30. How does Offer Optimization Work?
THIS GOES IN THIS COMES
OUT
For Each For Each
Offer Customer • The right
offers…
• Allocated to the
right
customers…
• To maximize
overall impact.
… while satisfying
all the constraints.
INSIGHT ACTION
Home
31. Highlights: Contact Optimization
• Maximize return from direct marketing with business constraints
Eligibility, quantity and financial constraints
Includes customer preferences, propensities and “persuadability”
Increases customer-centricity
• Broad optimization framework
Optimizes offers, message, channel, product,
incentive, time-frame
• Seamless integration with campaign planning
Marketer view (Portrait Dialogue)
Analyst view (Portrait Customer Analytics)
INSIGHT ACTION
Home
32. Portrait and Microsoft Dynamics CRM
Infusing Customer-centric Insight
into Microsoft Dynamics CRM
33. Dynamics CRM – Interaction History
INSIGHT ACTION
Home
3
34. Outbound Messages
Received by
individual contacts
sent via Portrait Dialogue
INSIGHT ACTION
Home
35. The history of the Real-
time next-best-offer
recommendations made to
Real-time next-best-offer the Contact are listed for
recommendations are made as the Rep’s reference
determined by rules & models
built in Interaction
Optimizer
INSIGHT ACTION
Home
36. Once presented to the Contact,
the Rep can assign a disposition
of Accept, Decline, or NoAction. The
Rep can also view the details of the
Campaign as set-up in CRM
INSIGHT ACTION
Home
37. Infuse Customer-centricity into Dynamics CRM
The applications
you use today
Made intelligent
Using your own data
INSIGHT ACTION
Home
38. Dynamics CRM and PCM - Architecture
Overview
G
Bran Bran
ch A D
SharePoint Report Center ch
Branch B C
Real-time
Campaign
History
Portrait
Create
Dialogue Task
CORE CRM
Core CRM
Create
Event
CRM DB
(Accounts, Contacts, Prospects)
INSIGHT ACTION
Home
Slide 38