How Data Science is Changing the Way Companies Do Business Colin White BI Research July 17, 2014
Sponsor 2
Speakers Colin White President, BI Research Bill Franks Chief Analytics Officer, Teradata 3
How Data Science is Changing the Way Companies Do Business Colin White President, BI Research TDWI-Teradata Web Seminar July 2014
The Evolution of Business Intelligence Type of analysis Business question Examples of deliverables Data science Prescriptive Predictive What action should be taken? What could happen? Rules-driven actions Optimization Predictive models Forecasts Business value Diagnostic Why did it happen? Behavioral analysis Interactive BI dashboards BI Descriptive What is happening now? What has happened? Real-time dashboards PDF reports via e-mail Copyright BI Research, 2014 5
It s Really About More Advanced Analytics Type of analysis Business question Examples of deliverables Data science Prescriptive Predictive What action should be taken? What could happen? Rules-driven actions Optimization Predictive models Forecasts Business value BI Diagnostic Descriptive Why did it happen? Applies to these types of analytics as well What is happening now? What has happened? Behavioral analysis Interactive BI dashboards Real-time dashboards PDF reports via e-mail 6
Fast Time to Business Value: Requirements Type of analysis Business question Examples of deliverables Prescriptive What action should be taken? Rules-driven actions Optimization Predictive What could happen? Predictive models Forecasts Usable by business analysts, not just data scientists easier to use analysis and visualization tools Seamless extension to diagnostic and descriptive BI Iterative development, easy to deploy and maintain, and (where required) near real-time results Copyright BI Research, 2014 7
Solution: Next Generation BI DRIVERS New business insights New technologies Next generation BI Reduced costs Enhanced data management Advanced analytics New deployment options TECHNOLOGIES Copyright BI Research, 2014 8
New Business Insights: Customer Marketing Situational 1-to-1 Marketing reach individual customers with the right messages and offers Micro-segmentation Analyze all channels: web, stores, call centers, purchases, buying patterns Analyze other information for influential factors: geography, weather Customer experience management make all experiences beneficial to customer/business Customer perception management analyze trends in social channels and respond appropriately In all cases analysts need to be able to move from analyzing past events to predicting future outcomes Copyright BI Research, 2014 9
New Business Insights: Fraud Detection Copyright BI Research, 2014 10
New Business Insights: The Internet of Things Further reading: GE Document - Industrial Internet: Pushing the Boundaries of Minds and Machines Copyright BI Research, 2014 11
New Technologies: extended Data Warehouse Analytic tools & applications Traditional EDW environment Investigative computing platform Data integration platform Operational systems Data refinery RT analysis platform Other internal & external structured & multi-structured data Real-time streaming data RT BI services Operational real-time environment Copyright BI Research, 2014 12
Two Key New XDW Components Investigative Computing Platform EDW data & analyses models & rules applications Analytic tools & applications o o Used for exploring data and developing new analyses and analytic models Output used by an enterprise DW, real-time analysis engine, or standalone LOB application Investigative computing platform Data Refinery o Ingests raw detailed data in batch and/or real-time into a managed data store EDW data Operational data Data refinery Other internal & external data, RT streaming data o Distills the data into useful information and distributes results to other systems Copyright BI Research, 2014 13
The Evolution of Open Source Software R (commercial version available) RapidMiner (commercial version available) KNIME (commercial version available) Apache Mahout (algorithm library) Weka (algorithm library) Issue: How easy is it to use these products to bridge the gap between BI and data science? Copyright BI Research, 2014 14
What is Data Science? One person or a team of specialists? Physical or virtual team? Where in the organization does it report, e.g., central IT, corporate executive, or business unit management? Part of, or separate from, a BI center of expertise or data governance group? Actual skills required? Which skills are the most difficult to learn or obtain? Education, recruiting or outsourcing for filling skill gaps? Traditional BI/DW versus millennial employee skills, experience and politics Modeling & analysis skills Business expertise Data engineering skills Copyright BI Research, 2014 15
Next Generation BI = Traditional BI + Data Science Business Analyst Data Scientist Prescriptive BI Predictive BI Business & data understanding Selected hypotheses Data warehouse Diagnostic BI Descriptive BI Business requirements Modeling Data preparation Model deployment Improved understanding Raw data Copyright BI Research, 2014 16
The Role of Investigative Computing Enables data scientists and analysts to blend new types of data with existing information to discover ways of improving business processes Allows data scientists and analysts to experiment with different types of data and analytics before committing to a particular solution May employ an analytic sandbox, analytic platform or a data refinery Results may include data schemas, analyses, analytic models, business rules, decision workflows, dashboards, LOB applications, etc. Represents a shift in the way organizations build analytic solutions: o o Increases flexibility and provides faster time to value because data does not have to be modeled or integrated into an EDW before it can be analyzed Extends traditional business decision making with solutions that increase the use and business value of analytics throughout the enterprise Copyright BI Research, 2014 17
Example: Teradata Aster Behavior Path Analysis Copyright BI Research, 2014 18
Example: Alteryx + Tableau Copyright BI Research, 2014 19
Dimensional Data Example: Teradata Identify/Retain At Risk Users SOCIAL FEEDS WEB AND MOBILE CLICKSTREAM Hadoop captures, stores and transforms social, images, and call records Aster does web sessionization, path and basic sentiment analysis with multistructured data Aster pre-built operators: sessionization, n-path, many to many basket and affinity, collaborative filtering for recommendations Surveys and Customer Feedback Multi-Structured Raw Data Call Center Voice Records Customer Feedback Traditional Data Flow Data Sources POS Web Sale Mobile Cust Sale & Item Master Hadoop Capture, Retain and Refine Layer ETL Tools Call Data Raw Sentiment Data Aster Discovery Platform Analytic Results Teradata Integrated DW Analysis + Marketing Automation (Customer Retention Campaign) Source: Teradata Copyright BI Research, 2014 20
Gaining Business Value from Next Generation BI Managers don t have to be data scientists, but they need to: Understand the fundamental principles well enough to appreciate the business opportunities, communicate with technologists and evaluate proposals for data science projects Be willing to invest in data and experimentation and supply the required resources Keep the BI and data science team on track Understand how to gain competitive advantage (or parity) from data science in the context of the corporate strategy and that of competitors Maintain momentum over competitors Collaborate with, and examine data science projects in other organizations Copyright BI Research, 2014 21
Final Thoughts Organizations need to build a high quality data science team that is managed by a knowledgeable person such as a chief analytics officer Keep humans in the decision making loop Mining and analyzing personal data raises important ethical and privacy issues that should not be ignored Applying BI analytics to a well-structured problem versus exploratory data mining requires different skills and tools, but these two approaches need to be able to work together Note: Several of the ideas presented on these last two slides were summarized from information in the book Data Science for Business Copyright BI Research, 2014 22
How Data Science is Changing the Way Companies Do Business Bill Franks Chief Analytics Officer, Teradata TDWI-Teradata Web Seminar July 2014
Leverage Analytics In Diverse Ways Perform discovery analysis alongside confirmatory analysis to maximize benefits Discovery Analysis Full scope not defined Interactively evolving hypotheses Business problem is developing Aim is to identify new theories Confirmatory Analysis Examining predefined problems Assessing specific hypotheses Business problem well defined Aim is to validate a theory 24 Proprietary and Confidential to Teradata and Bill Franks. Do not distribute without permission.
Utilize New Analytic Disciplines Statistics Augment traditional analytic approaches with new approaches Forecasting 25 Proprietary and Confidential to Teradata and Bill Franks. Do not distribute without permission.
Utilize New Analytic Disciplines Statistics Augment traditional analytic approaches with new approaches Forecasting Graph Analysis Geospatial Text Analysis 26 Proprietary and Confidential to Teradata and Bill Franks. Do not distribute without permission.
Teradata Aster SNAP Framework 27 Proprietary and Confidential to Teradata. Do not distribute without permission.
Your Team Will Need To Expand & Evolve No single individual will likely know every analytic discipline Build out a team that has what you need in total + = Person 1 Person 2 Total Package! 28 Proprietary and Confidential to Teradata and Bill Franks. Do not distribute without permission.
Do You Need A Chief Analytics Officer? What is a Chief Analytics Officer & why do you need one? Hire Me! 29 Proprietary and Confidential to Teradata and Bill Franks. Do not distribute without permission.
Do You Need A Chief Analytics Officer? What is a Chief Analytics Officer & why do you need one? What about a Chief Data Officer? Hire Me! 30 Proprietary and Confidential to Teradata and Bill Franks. Do not distribute without permission.
Teradata Unified Data Architecture (UDA) Your environment must enable any analysis against any type or volume of data at any time Data Scientists Business Analysts Marketing Front-Line Workers Engineers Customers / Partners Executives Operational Systems LANGUAGES MATH & STATS DATA MINING BUSINESS INTELLIGENCE APPLICATIONS ERP UNIFIED DATA ARCHITECTURE System Conceptual View Marketing Marketing Executives SCM CRM Images DATA PLATFORM INTEGRATED DATA WAREHOUSE Applications Business Intelligence Operational Systems Customers Partners Audio and Video TERADATA DATABASE Data Mining Frontline Workers Machine Logs Text Web and Social SOURCES TERADATA DATABASE HORTONWORKS DISCOVERY PLATFORM TERADATA ASTER DATABASE Math and Stats Languages ANALYTIC TOOLS & APPS 31 Proprietary and Confidential to Teradata. Do not distribute without permission. Business Analysts Data Scientists Engineers USERS
Spread Your Bets With The Teradata UDA! Who knows what the future holds? Don t place all your chips on an architecture that assumes specific outcomes Hedge your bets with an architecture that can adapt to whatever the future holds! 32 Proprietary and Confidential to Teradata and Bill Franks. Do not distribute without permission.
Question & Answer Thank you! 33 Proprietary and Confidential to Teradata. Do not distribute without permission.
Questions?? 34
Contact Information If you have further questions or comments: Colin White, BI Research info@bi-research.com Bill Franks, Teradata bill.franks@teradata.com 35