Profit from Data and Text Mining. A two part series.

1
Profit from Data and Text Mining. A two part
series.

• Tom Carroll
• Systems Engineer
• SPSS Inc.

2

With data mining
An entertainment company discovered that their
best spenders consisted of three distinct
segments rather than one larger one. The biggest
spenders spent over 200K but typically visited
only twice a year for high profile events


They went into action and added a third high
profile event designed around this segments
purchasing patterns and drove .
3


They continued their data discovery and
Discovered one highly profitable segment that
purchased high volume, low dollar purchases on
their website

They went into action and added a special
promotions section on their website based around
this segments purchase pattern and drove
4


They continued their data discovery and
Created business profiles of all nine segments of
their customer base. They only thought they had
four segments originally.

They went into action and built propensity to
purchase models for each segment and drove
5


This company was profiting greatly, yet just
getting started.
Data mining and SPSS Clementine From data
discovery to action to profits
6
Welcome to todays presentation. Profit with
data mining.Part one in a two part series.

• Tom Carroll
• Systems Engineer
• SPSS Inc.

7
Commonly Asked Questions

• Will I be able to get copies of the slides after
  the event?
• Is this webinar being taped or can I view it
  after the fact?

Yes
Yes
www.spss.com/events
8
Agenda

• SPSS Highlights
• Data mining today Who is profiting from it
• How and where is data mining being used
• Brief review What exactly is data mining?
• SPSS Clementine
• Demonstration
• Key Take Aways
• Secrets to Success
• Q A

9
SPSS Highlights

• Commercial enterprises
• 95 of the Fortune 1000 companies
• 10 largest pharmaceutical companies
• 85 of top consumer packaged goods companies
• 10 largest market research firms

• Government non-profit organizations
• Security agencies worldwide
• Law enforcement authorities around the world
• Utilized by all four branches of the U.S.
  Military
• All 50 U.S. State Governments

• Education
• At every major college university worldwide
• Distributed with over 100 textbooks
• 35 years of experience in scientific research
• 1 statistical software in US campuses

10
Organizations of all sizes are profiting from
data mining

• NY Times
• Office Depot
• Best Buy
• Humana
• Pep Boys
• Cox Communications
• Churchill Downs
• Proctor Gamble
• Experian

• Coach
• Phillip Morris
• Vail Resorts
• GSI Commerce
• Starbucks Coffee
• Torchmark
• Verizon
• Pfizer
• Equifax
• AOL
• Direct TV

11
Who performs data mining

• Analysts
• Business Intelligence users
• Business executives
• Small, medium, and large sized organizations
• Companies in virtually every industry

12
How and where is data mining being used

• Data Mining is enabling companies to
• Understand customers better
• Increase sales
• Increase profitability
• Increase website profitability
• Cross-sell/Up-sell
• Increase retention/loyalty
• Detect fraud
• Identify credit risks
• Acquire new customers

13
What is Data Mining?
The process of discovering meaningful new
relationships, patterns and trends by sifting
through data using pattern recognition
technologies as well as statistical and
mathematical techniques. The Gartner Group

• Data mining uses existing data to
• Predict
• Response
• Category membership
• Numeric metrics
• Group
• Cluster (group) attributes and metrics together
  based on their characteristics
• Associate
• Find events that occur together, or in a sequence
• Find outliers
• Identify cases that dont follow expected behavior

14
Data Mining and Predictive AnalyticsPowered by
Clementine

• Industry-leading workbench for data mining
• Comprehensive range of tools for all stages of
  the data mining process
• Pioneered visual approach for maximum
  productivity
• Multiple modeling techniques to predict customer
  behavior
• Easily integrate Clementine into your existing
  systems

15
What the analysts think of SPSSs data mining
workbench Clementine

• From the META Group
• "SPSS remains a leader in the current data
  mining market "
• "A rich set of data preparation functions is
  provided out of the box, with little coding
  required to collect data specifically for
  modeling purposes."
• "Clementine offers internal text mining
  capabilities making it possible to extract
  information from unstructured data."
• "Clementine can incorporate Web data into the
  modeling environment..."

16
Clementine Supports for a Proven Process

• CRISP-DM
• Cross Industry Standard Process for Data Mining
  Focused on business issues
• Consortium of data miners from various industries
  manufacturing, marketing, and government
• User-centric process
• Closed loop process
• Goal is to take action in your organization with
  the results

17
Demonstration - Spruce Place ClothingBusiness
Understanding

• Business Challenge
• Spruce Place Clothing are six months into the
  launch of their Suit Separates line
• Monthly sales performance has been 42 below
  expectations
• Unable to identify the right market for the new
  line
• Business Goal
• Leverage our customer buying behavior, service
  interactions and customer profile information to
  increase sales and brand awareness

18
Demonstration - Spruce Place ClothingSystems
Spruce Place Clothing
Out Bound Campaign Application
Inbound Telesales Application
Business Intelligence Reporting Application
Predictive Metrics
Targeted Customer Lists
Real Time Offer Decisioning
Clementine
Unified Customer View
Transaction Data
Demographic Data
Web Interaction Data
CC Text Data
19
What we are going to show today

• Merging different data sets together to provide a
  unified view of the customer
• Leverage Clementine and Data Mining to identify
  buying patterns
• Take action on those discovered patterns in three
  ways
• Provide the executives with a view into the
  target market for the new line
• Targeted lists for those most likely to purchase
  the Suit Separates Line
• Leverage real time scoring to allow telesales to
  identify those most likely to purchase products
  from the Suit Separates Line.

20
Demonstration

• If you are not automatically taken to the "Shared
  Application" screen during the demonstration,
  please click on the "Shared Application" button
  at the bottom of your screen.

21
Demonstration Summary and ROI

• Combined Transactions and Customer Profiles
• Identified simple patterns through data
  visualization
• Those who buy other casual dress products (Casual
  Pants, Shorts), buy Suite Separates
• Low to Mid range income buyers
• Generated models to predict who is most likely to
  buy Suit Separates
• Identified two major rules to predict buying
  pattern
• Casual dressers purchase Suite Separates
• Pocket of buyers who are male, under age 47 , and
  have bought Shorts and Casual Pants
• Batch scoring of all customers predicted 7,500
  possible buyers, which at 6 conversion is approx
  60K for campaign

22
More on ROISPSS Customers successes

• Banking
• Lloyds TSB
• Saved 35 million by reducing credit card fraud
• Groupe Banque Populaire (Natexis)
• generated 1.8 million additional revenue on a
  lead generation campaign
• BankFinancial
• 7 x increase in response rates, 80 reduction in
  costs
• Insurance
• Aegon
• Generated 30M additional revenue in service call
  center
• FBTO
• Decreased direct mailing costs by 35 percent,
  increased conversion rates by 40 percent

• Telecommunications
• Verizon Wireless
• Cut churn by 20, saved 33 of at-risk clients
• Swisscom
• Doubled responses on win back campaigns
• Telstra
• Doubled sales in service call centers
• Other industries
• Experian
• Generated 2.5 million in catalog revenue while
  reducing hardware and software maintenance costs
  by 80
• Center Parcs
• added 3 million to their bottom line
• Reduced mail costs by 46
• Sofmap.com (retail)
• Tripled profitability of online operations

23
Nucleus ResearchThe Real ROI from SPSS

• 94 achieved a positive ROI, with an average
  payback period of 10.7 months.
• Returns were achieved through reduced costs,
  increased productivity, increased employee and
  customer satisfaction, and greater visibility.
• Over 90 of SPSS users surveyed attributed an
  increase in productivity to SPSS, expressed in
  increased productivity from existing employees,
  or reduced the need to hire additional employees.
• 81 of projects were deployed on time, 75 on or
  under budget.

This is one of the highest ROI scores Nucleus
has ever seen in its Real ROI series of research
reports Rebecca Wettemann, Vice President of
Research at Nucleus Research
24
Clementine Key Messages

• Unparalleled productivity
• Advanced Visualization
• Ease of use
• Openness
• Breadth and depth of modeling techniques
• Scalability
• Deployment
• Multi channel integration
• Traditional source, web data, free form text

25
Key Take Aways

• Companies of all sizes are profiting
• Many types of users are performing data mining
• Performing data mining is within reach
• Proven tools are available to take you from data
  discovery to action!

26
Secrets to data mining success

• Follow CRISP-DM
• Begin with the end in mind
• Set expectations
• Limit the scope of your initial project
• Identify a steering committee
• Avoid the data dump

27
Putting it all together The SPSS Enterprise
Platform
28
Achieving stronger results with multiple sources
of data
Text data Up to 40improvement

• Attitudinal data
• - Opinions
• Preferences
• Needs
• Desires

• Interaction data
• - Offers
• Results
• Context
• Click streams
• Notes

Attitudes Up to 30improvement
Web data Up to 20improvement

• Behavioral data
• - Orders
• - Transactions
• Payment history
• Usage history

• Descriptive data
• Attributes
• Characteristics
• Self-declared info
• (Geo)demographics

29
Next webcast in the series

• Profit with Text Mining
• August 10, 2005, 1200 a.m. EST
• Duration 1 hour

30
Please give us your feedback

• Take our event survey
• You will automatically be taken to it after the
  event.

31
Questions?