Building and Deploying Customer Behavior Models

Transcription

1 Building and Deploying Customer Behavior Models February 20, 2014 David Smith, VP Marketing and Community, Revolution Analytics Paul Maiste, President and CEO, Lityx

2 In Today s Webinar About Revolution Analytics About Lityx Customer Behavior Lifecycle Classic Approach vs. Today s Approach Demonstrations and Case Studies Q&A

3 Revolution Analytics at a Glance Who We Are Only provider of commercial big data big analytics platform based on open source R statistical computing language Our Software Delivers Scalable Performance: Distributed & parallelized analytics Cross Platform: Write once, deploy anywhere Productivity: Easily build & deploy with latest modern analytics Our Services Deliver Knowledge: Our experts enable you to be experts Time-to-Value: Our Quickstart program gives you a jumpstart Guidance: Our customer support team is here to help you Customers 300+ Global 2000 Global Presence North America / EMEA / APAC Global Industries Served Financial Services Digital Media Government Health & Life Sciences High Tech Manufacturing Retail Telco

4 Exploding growth and demand for R R Usage Growth Rexer Data Miner Survey, % of data miners report using R R is the first choice of more data miners than any other software Source: R is the highest paid IT skill Dice.com, Jan 2014 R most-used data science language after SQL O Reilly, Jan 2014 R is used by 70% of data miners Rexer, Sep 2013 R is #15 of all programming languages RedMonk, Jan 2014 R growing faster than any other data science language KDnuggets, Aug 2013 More than 2 million users worldwide

5 Revolution R Enterprise is. the only big data big analytics platform based on open source R High Performance, Scalable Analytics Portable Across Enterprise Platforms Easier to Build & Deploy Analytic Applications

6 Speaker Bio Paul Maiste is President and CEO of Lityx. He has a Ph.D. in Statistics, with nearly 25 years of experience designing and delivering strategic analytic solutions for predictive modeling and marketing optimization to businesses of all sizes and across industries.

7 Thank you

8 Building and Deploying Customer Behavior Models February 20, 2014 Click to edit Master title style

9 Agenda Intro and Background Customer Behavior Lifecycle Classic Approach vs. Today s Approach Demonstrations and Case Studies Q&A 2

10 Speaker Bio Paul Maiste is President and CEO of Lityx. He has a Ph.D. in Statistics, with nearly 25 years of experience designing and delivering strategic analytic solutions for predictive modeling and marketing optimization to businesses of all sizes and across industries. 3

11 Company Background Lityx is a world-class analytic solutions and services firm with a diverse set of clients across multiple industries. We deliver a hosted advanced analytics platform, and help our clients by applying deep expertise to complex analytic solutions. Our focus is predictive modeling and optimization applications in marketing analytics and CRM. 4

12 Our track record Lityx has worked with marketers in diverse markets such as nonprofit, media, gaming, financial services, healthcare, and retail/cpg. 5

13 Poll Question #1 What analytics platform are you currently using? - SAS - SPSS - R / Revolution R Enterprise - KXEN - Other 6

14 Customer Behavior Lifecycle Modeling Customer Acquisition Customer segmentation. Predict prospect future value. Predict likely responders. Predict best product and best offer. Determine best offer timing. Relationship Growth Predict cross-sell and up-sell. Determine natural product affinities. Determine most profitable marketing offers / messaging. Increase loyalty and share of wallet. Customer Retention Predict likely churners and reasons. Determine customer potential value. Determine best retention offer. Increase loyalty. Winback lost customers. 7

15 Customer Behavior Lifecycle Modeling Customer Acquisition Customer segmentation. Predict cross-sell and up-sell. Predict prospect future value. Determine natural product Predict likely responders. affinities. Predict best product and best offer. Determine most profitable marketing offers / messaging. Determine best offer timing. Increase loyalty and share of Optimize Customer wallet. Communication Customer Retention Predict likely churners and reasons. Determine customer potential value. Determine best retention offer. Increase loyalty. Winback lost customers. Relationship Growth 8

16 Poll Question #2 What area of customer behavior modeling are you most interested in leaning about/doing more of? - Customer Acquisition - Relationship Growth - Customer Retention 9

17 The Imperative for Advanced Analytics Marketers have a lot to worry about to maintain relevant data, create and grow profitable customers, and be more efficient with existing budget. Forrester has recently said: Vendors need to create more analytic solutions that customers can use out of the box such as business-user-oriented interfaces. We Agree, BUT ALSO Let s use the opportunity to make data scientists and modelers more efficient as well! 10

18 Classic Approach Iterate Often re-code in different system for implementation Data Prep and Manipulation Implement Design Approach and Algorithm Coding Iterate Write code for performance metrics and charting Test and Validate Iterate through multiple algorithms Iterate Iterate through multiple data cleaning approaches Debug and re-run 11

19 Today s Approach Design model using business language Simply presented options for the advanced user Automated and intelligent data preprocessing Iterative processing of multiple algorithms and settings Handle computational workload Pre-computed performance metrics Automated charts and comparisons Built-in model management Automated scoring process without coding 12

20 What about the data scientists? Like Me! It s time to focus our attention on design and analysis instead of hacking, debugging, and iterating. - Without losing the computation power and modeling flexibility we require 13

21 Data, insights, predict, optimize Cloud based platform for advanced analytics Data Manager InsightIQ PredictIQ OptimizeIQ Powered By 14

22 Live Demonstration Retail Churn Modeling Apparel Industry 15

23 Poll Question #3 My expertise is best described as: - Hard core data scientist - Big Data guru - Scientific programmer/coder - Business analyst - Consultant - Marketing / Business - IT 16

24 Case Study Large Non-Profit Organization Affinity / Cross-Sell Models Client outsourced building of over two dozen affinity models to vendor using classic tools and manual process (3-4 month effort). Rebuilt all models using LityxIQ in 2 weeks, and model results (such as lift) were 5% better than manually built models. Expanded to a series of 40 models, all managed within LityxIQ. 17

25 Q&A For more information: Art Warren - [email protected] Paul Maiste - [email protected] Upcoming Virtual Course led by Paul Maiste Customer Analytics for Marketers April 21, 23, 28, 30 (9-1 PT) Register at: Coupon Code: RevoWebinar for 10% discount 18

26 More Information 19

27 Data Manager: data preparation Data Manager Easily import and manage complex data sources. Append and join datasets together. Clean, transform, create new fields. Filter and aggregate. General data preparation for using in other solutions. 20

28 InsightIQ: analysis, BI, dashboarding InsightIQ Interactive graphical analysis for creating and sharing insights through charts and tables. Business intelligence, reporting, and executive dashboards. 21

29 PredictIQ: predictive modeling solutions PredictIQ Automated model building focused on business objectives including churn, value, risk, and affinity models Includes validation, model management and version control, scoring, and implementation Business forecasting models for sales, revenue, and other business metrics 22

30 OptimizeIQ: marketing optimization OptimizeIQ Optimize marketing budget/resources across customer segments, products, channels, and other business dimensions Optimize media spend within and across channels Optimize individual customer communications to maximize profitability Easy to define objectives and business constraints for a non-technical user 23

31 Version 3.0 End Q1 Integration with Revolution RRE Big data connectivity to Hadoop - In-database analytics with Teradata - Big data modeling using GLM, Tweedie, CART, and more - Integration directly with existing Revolution R code for additional control (Ver 3.x) API connectivity 24