The Power of Analytics on the System i

1
The Power of Analytics on the System i
Customer Loyalty Profitability Tactics for t
he Gaming Industry

Bill Carpenter PreSales Gaming Consultant
2
Agenda

• SPSS The leader in System i analytics for
  casinos
  
• The Journey from Business Intelligence to
  Predictive Analytics
  
• Using Predictive Analytics during the Patron
  Lifecycle
  
• Caesar Entertainment Success Story
• Demo
• Questions

3
SPSS Corporate Overview

• Software company
• Top 25 software company
• 35 year heritage in analytic technologies
• Market leader in predictive analytics
• Operations in over 60 countries
• Proven track record
• More than 95 of FORTUNE 1000 are SPSS customers
• SPSS has been used at over 50 casinos worldwide
• ShowCase Product Suite has 80 of the market of
  Business Intelligence for System i
  
• Only provider of Predictive Analytics on the
  System i

4
What Is Predictive Analytics?

• Using historical patron data and algorithms to
  identify hidden patterns and create models to
  make profitable decisions.

Understand Patron Behavior Predictive Patron Be
havior
Influence Patron Behavior
Today I will present Data Mining on the System I
as the Predictive Analytics Solution
5
How Is Data Mining Different from BI?
Business question
6
The Patron Lifecycle

• First visit Profile patron based on initial
  play to determine initial value and tier group.

• First offer Determine how to treat patrons
  using first visit, comparative profiles and
  predictive analytics.
  
• First return Validate tier group and future
  offering.
  
• Customer migration Move to higher tiers in
  future visits and reduce spend for lower tier
  groups.
  
• Predictive Analytics -- Determine Total Patron
  Value, frequency of visit, likelihood to respond,
  and offer type.
  
• Loyalty and profitability Continuously reassess
  tier group, Total Patron Value, and retention and
  offer strategy

7
How to Increase Revenue Profitability

• More frequent visits
• Longer stay per visit
• More play per visit (Higher trip theoretical win)
  
  
• More non-gaming purchases per visit
• Cost effective marketing

8
Predictive Analytics Data Mining
Using historical patron data to predictive future
patron behavior.
9
Gaming Profitability Strategies

• Patron growth initiatives
• Increase retention rates for high rollers
• Upgrade mid-tier players to higher betting tiers
• Cross-sell other casino services
• Service upgrade with low betting customers
• Patron retention initiatives
• Create return visit promotions to high value
  patrons
  
• Extend customer relationship with high Total
  Value Patrons
  

10
Examples of Data Mining in Casinos

• Classify / Cluster patrons into groups by
  recognizing patterns
  
• Example Determine new ways to segment patrons
  not just by play, but also by total activity and
  other attributes in order to enhance marketing
  campaign effectiveness.
• Association what events are likely to occur
  together
  
• Example Determine what gaming and non-gaming
  events will patrons will engage in while visiting
  a property, and to help determine the best
  cross-selling opportunities

• Predict what may happen in the future
• Example Predict response rates to a proposed
  marketing campaign.
  
• Marketing campaign optimization for multiple
  campaigns across channels (web, email, mail and
  other)
  
• With multiple campaigns and limited budget which
  campaign to send to each patron to maximize total
  profit for all campaigns.
  
• Time Series Forecasting values over multiple time
  periods
  
• Example Forecast slots staffing based on day of
  year, shift and special events occurring.

11
Customer Success Story
12
Caesars Entertainment Background

• System i platform for all data warehousing
  needs.
  
• ShowCase Warehouse Builder
• ShowCase Query
• ShowCase Report Writer
• ShowCase Essbase and Analyzer OLAP Solution
• System i and ShowCase were the natural choices
  for the data warehouse, with little effort needed
  for raw CMS and LMS data transition.
  
• Fast and user-friendly, point and click
  functionality.

13
The Next Step Marketing Analytics

• It was determined that marketing efforts needed
  to be intelligently focused. Data Mining was the
  answer to the companys wants.
  
• What was required was not a statistical tool, but
  rather an application that would enable a higher
  degree of analytical sophistication.
  
• Caesars Selection System i Clementine.

14
Caesars Four Goals of Data Mining

• Lifestyle/Demographic Segmentation
• Attrition Modeling
• Response Modeling
• Market Value Analysis

15
Lifestyle/Demographic Segmentation

• Goals
• Required by the branding department.
• Intelligent customer acquisition.
• Customer profitability analysis by product and
  segment.
  
• Results
• All analysis were moved in-house with substantial
  savings in potential outsourced services.

16
Attrition Modeling

• Goals
• A more accurate way of predicting the likelihood
  of patron attrition at an early point.
  
• Traditional activation mailers relied on 3, 6, or
  12 month policy.
  
• We needed to predict at the earliest possible
  point a deviation in gaming behavior on a patron
  level.
  
• Results
• Attrition reduction by 10-15 depending on the
  segment.

17
Response Modeling

• Goals
• Matching the right patron to the right
  promotion.
  
• Developing a response score for every promotion,
  slot tournaments, events, hotel occupancy, etc.
  
• Results
• Reduced mail cost by over 40, with only a 3
  loss in revenue.

18
Market Value Analysis

• Goals
• Determine the theoretical potential of any
  patron.
  
• Deliver a promotion based on the true Market
  Value of selected customers.
  
• Results
• An increase in theoretical win of 32 for all
  selected players.
  
• An increase of 75 of selected players in the
  upper tiers.
  

19
Demo
20
As a Marketing Analyst I need to improve my next
campaign
21
Marketing Challenge
The response rate to GDK Casino monthly mailing
promotion of 100,000 has not increased over the
last six months.
22
Current Data Available

• Loyalty card number
• Age
• Sex
• Marital status
• Home location
• Game played (slots or table)
• Theoretical win
• Promotion response history

Data is stored in three different tables
23
System i Clementine Demo
24
Customer Segmentation
S
S
S
T
T
S
F
F
M
M
M
Y
N
N
N
Y
Y
Y
25
Deployment Generating the ROI

• Direct mail
• Email
• Call Center
• Website
• Kiosks
• Tickets

26
Questions
Take the gamble out of marketing campaigns
and
increase marketing ROI
27
Free Whitepaper

• Increase Customer Value with Predictive
  Analytics How predictive analytics benefits
  gaming organizations
  
• www.BIforSystemi.com
• or contact SPSS [phone]