2018 Analytics Challenge Winner Presentation

Size: px
Start display at page:

Download "2018 Analytics Challenge Winner Presentation"

Transcription

1 SPEAKER NAME TITLE, COMPANY 2018 Analytics Challenge Winner Presentation Announcement and Analytic Trends 1

2 Thanks to our Challenge Partner and Emcee: Aaron Davis Chief of Analytical Services for DataLab USA Member of the Analytics Community Leadership Council 2

3 SPONSOR - DataLab USA AT A GLIMPSE DataLab USA is a market leader in data-driven addressable marketing technology & approach. We collaborate with our clients to develop fully optimized CRM and acquisition programs Fully customized data sourcing from 12,000+ attributes spanning multiple credit and demographic sources. Experts at delivering ROI through advanced analytics and predictive modeling Winner of multiple DMA Analytic, KDD, ICDM, and Kaggle Competitions 3

4 The Analytic Challenge Committee Devyani Sadh, Ph.D., CEO and Chief Data Officer, Data Square Leo Kluger, Principal Data Scientist, Chief Analytics Office, IBM John Friedmann, Manager of Analytics, Deloitte Dr. Tomeka Hill-Thomas, Lead Data Scientist / Associate Director of Workforce Analytics, Ernst & Young Marty Thomas, Senior Vice President, RPM Direct Aaron Davis, Chief of Analytical Services, DataLab USA Learn more DMACommunities@ana.net 4

5 The Analytics Community will soon be the DataScience and Machine Learning Committee of the ANA. Members are welcome to join us and give us your input for events that focus on hot topics, best practices and emerging trends in advanced analytics, AI, Machine Learning, and Deep Learning. We re always looking for presenters and authors with use cases and application-based presentations. Contact us to learn more 5

6 MISSION STATEMENT The current goal of this Community is to provide a forum to develop thought leadership and best practices related to customer-centric, datadriven marketing. This includes the synergistic integration of CRM strategy and analytics, which together yield increased growth and profitability through the optimization of customer insights, coordinated touch-point interactions and corporate infrastructure. The future description and purpose will include AI, ML, Deep Learning and IoT to keep up with emerging technologies, applications and trends. 6

7 The Role of the Analytic Challenge Launched by the DMA Analytics Council in 2006 Raise the visibility of analytics as a critical enabler of better business outcomes Allow practitioners to go head-to-head in building a model to support a realworld marketing challenge Share best practices and facilitate the exchange of ideas allowing practitioners to raise their game and increase the value of their work 7

8 About the Analytic Challenge - Constituents Corporate Partner Supplies the foundational problem and data for the competition Receives the winning solution for inmarket use Validation and benchmarking of their internal solutions versus vetted external solutions Challenge Partner Responsible for the infrastructure and resources to run the competition Works with corporate partner to construct the challenge problem Scores and evaluates competitor solutions Competitors DMA members, industry practitioners and academia. Able to benchmark their predictive modeling approaches in a public forum Winners are able to share their ideas and innovations 10

9 Analytic Trends in the Industry Shift from traditional statistical approaches to data mining/machine learning Massive parallelization; Cloud, GPUs, etc Shift to open source from comprehensive commercial offerings 11

10 Analytic Trends in the Industry Artificial Intelligence tasks and systems making analysts and marketers more effective. Automated Machine Learning Internet Of Things 12

11 Analytic Trends in the Industry Evolving role of the analyst from narrowly defined job descriptions (ie data analyst, statistician, business analyst, etc) to more wholistic ones (ie data scientist) 13

12 Successful themes in Data Mining Competitions Domain knowledge Methodical datamining process that recognizes cyclic nature of knowledge discovery Team based solution Ability to evaluate many different algorithms 14

13 This Year s Challenge The corporate partner sells products and services directly to their customers through their widely used smartphone app The partner seeks to leverage a predictive model to maximize ROI by identifying new users that are most likely to generate long-term revenue. Challenge participants were asked to predict additional 6-month revenue utilizing: 1 st Party Data Customer attributes at the time of first order Transactional data captured within the first 30 days 3 rd Party Data hundreds of attributes Demographics Financial Behavioral Lifestyle 15

14 This Year s Challenge 27.9% of customers had additional purchases within 6 months Average revenue given additional purchase: $28.40 Average Additional Revenue (27.9% X $28.40): $7.93 Solutions evaluated utilizing randomized blind sample on Mean Squared Error 16

15 Consider joining our team of experts Make Connections - Get noticed. 17 Get published in the 2019 Analytics Journal Now accepting article abstracts. Love a Challenge? Give us your tough nut analytic problem for 2019, and let us solve it for you with our next Analytic Challenge. For Consideration please send to: DMACommunities@ana.net

16 Give us your feedback and stay connected. Congratulations to our winner: Cartesian DataSciences Thanks to DataLab USA for being a great Challenge Partner! Enjoy your &THEN18 Experience!