The document discusses how analytics is evolving to embrace gut instinct and business expertise through a concept called "Soulful Analytics". It advocates defining clear business goals to guide modeling and ensuring business validation of results to build confidence and buy-in. The document also provides an example of how Points, a loyalty platform company, partners with business experts to develop predictive models focused on achieving specific goals like increasing transaction volumes.
2. 2
Objectives
Understand how big data is changing the
way we make decisions.
Understand what’s Soulful analytics.
Points case study.
3. 3
Leverage our LOYALTY COMMERCE PLATFORM to
facilitate growth and innovation with loyalty programs
PARTNER with leading loyalty Programs around the
world
Deliver multiple business applications through both
PRIVATE LABEL and POINTS’ BRANDED channels
Generate revenue by TRANSACTING POINTS for
retail margin, service fee or commission
LOYALTY COMMERCE PLATFORM
• Program Integration
• Member Validation
• Debit, Credit, Payment
• Marketing / Merchandising tools
• Promotions: cross / pre / post transaction
• Proactive vs. Reactive
• Management of marketing assets
• Pricing analysis
• Fraud management
• Distribution Partner Integration
• Data correlation (internal + external)
• Content Management
Points Current Core Business
3
4. 4
Points is the only company with transaction level access to
the world’s largest loyalty programs with +500 million
members
4
200+ product deployments on 5 continents – transaction
level access is a key strategic asset and barrier to entry
5. 5
Externalizing the Loyalty Commerce Platform for the future
3RD PARTY
ACCESS
LOYALTY COMMERCE PLATFORM
• Program Integration
• Member Validation
• Debit, Credit, Payment
• Marketing / Merchandising tools
• Promotions: cross / pre / post
transaction
• Proactive vs. Reactive
• Marketing asset management
• Pricing analysis
• Fraud management
• Distribution Partner Integration
• Data correlation (internal + external)
• Content Management
PAYMENTS/WALLETS
TRAVEL
RETAIL/ECOMMERCE
5
GAMING
SOCIAL
6. 6
Big Data has changed the decision making landscape
Real time
decision making
7. 7
80% decisions made based on gut instinct!
58%
of respondents
identify “Outcome
from data” as a key
analytics challenge
US executive says2.
“Sometimes in business
there’s that gut
instinct..how to take that
information and apply it to
make business strategies work
is one of the biggest
challenges.”
1. Creating Business Value with Analytics. MIT Sloan Management Review.
2. Based on Analytics in action : breakthroughs and barriers on the journey of ROI.
Adoption rate of analytics2
Mean Decisions
How are Senior managers making decisions2
Analytics Maturity Components1
• Tools and expertise
• Information management
practices
• Analytics culture
8. 8
Points story - Right Offer, Right Time, Right Person, Right Channel
Increase transaction volume for Buy, Gift product
during a month when promotion is in the market.
With minimum number of emails.
Data Available - 700MM Data Points
•Member demographic information.
•Member transaction logs.
•Communication logs.
•Web logs.
Few known facts :
• Average transaction size for each transaction.
•Average response rate for current promotions
and total number of emails sent.
•Conversion funnel for the product.
•Current penetration in the market.
3 MM +
Members
200K
Transactions/Year
4
Average emails/month
9. 9
Business Objective
Analytics Goal
Predictive Model Development
Deployment
Business Owner
Business Analyst
Data Science
Build
transaction
volumes.
Find customers
who have a high
likelihood to buy Data Findings : geographically all
over the globe. Previous balance
between 0K – 200K, Previous
activity – Skewed on left.
Modelling decision – Throw it in a
classification model and lets look
at sorting.
Model result : Deploy to 1.5MM
customers
The Waterfall Approach
?
10. 10
Lifts do not come with revenue guarantee!
0.00%
1.00%
2.00%
3.00%
4.00%
5.00%
6.00%
7.00%
8.00%
1 2 3 4 5 6 7 8 9 10 all
Transaction rate Deciles charts
All Targeted
Offer No Offer
• Why 3 deciles?
• What does lift represent?
• What about other 60% left
out in deciles 4-10?
• Is there a long term
impact?
• REVENUE???
• Does it represent all the
customers?
11. 11
Model is good but ….
Lack of Comprehensibility
Lack of Confidence in model
Long delivery time
12. 12
Soulful Analytics : Embracing the gut instinct!
Business
Objective
Business
Goals
Success
Metrics
Analytics
Objective
Modelling
Process
Business Tactic
& ROI
13. 13
Defining Business Goals : First Step towards soulful analytics
Disintegrate business objective into specific goals
Business Analytics
Ask Questions to understand and disintegrate the problem
Provide analysis to understand the problem
Business Goal
Experience + Emotion
Data and Logic
14. 14
Defining Business Goals : First Step in the Partnership
Right Person
Customer Segments
First Time
2 txn’s Multiple
programs
>2 txn’s
Increase penetration,
moderate transaction
size
Increase
engagement in
multiple programs,
Very high transaction
size
Business
Expertise
15. 15
Right Offer
1.83%
1.37%
1.22%
1.05%
Average transaction rate by offer
Increase penetration in
customer with likelihood to
buy with “50% more” offer
Business Expertise
Increase overall
campaign profitability
Increase Number of New
Customers
Increase engagement of
repeat purchasers
Analytics team is a partner in framing the goals.
Actionable business goal supported by data
16. 16
Models are based on tactics, not strategy.
Send PDC promotions to
repeat customers
Send better and best offers
to existing customers to
increase revenue.
Increase engagement for
repeat customers
Target new purchasers
using retargeting
Predict expected
membership in other
programs.
Model for offer response
on these programs.
Rank customers based on
their likelihood to buy next
month.
Rank for all offers and find
better and best offer
response rates
Identify potential
customers for retargeting
using weblog data.
Rank on likelihood to buy
Increase Number of New
Customers