1 CoolaData Predictive Analytics About CoolaData CoolaData empowers online companies to become proactive and predictive without having to develop, store, manage or monitor data themselves. It is an open data infrastructure which provides deep behavioral analytics to analyze, visualize, predict and act on data. CoolaData provides a full analytical stack at a fraction of the cost of developing similar capabilities in-house. Its customers get actionable answers to business questions, resulting in improved user acquisition, reduced churn rates, better conversion and higher customer lifetime value. For more about CoolaData go to
2 Predictive Analytics What would you give to know what your customers will do in the future? Most online companies track information on past customer activity, but the ability to survive and thrive online and in mobile is increasingly tied to the prediction of future actions and transactions. Predictive analytics is a broad term which describes a variety of statistical and analytical models used to predict future events. Businesses use predictive models to identify risks and opportunities based on patterns in their data. These predictive models generate a score for each variable the higher the score, the more probable it is that a behavior will take place. Predictive analytics is in use by financial services, travel, healthcare, retail, gaming and more, but banks and financial services are at the forefront of its innovation. Banks utilize scoring models to process anything from credit history and loan applications to customer data and demographics in order to rank individuals by their probability of being a lucrative customer or a defaulter. Predictive models define the relation between a certain subject s behavior and one or more predefined known variables. Variables in this case can be gender, age, time of purchase, item purchased, location, a sequence of actions online, etc. The model s objective is to predict the likelihood that a similar subject in a different sample will exhibit the same performance. Advancements in computing speed have enabled individual agent modeling systems to effectively simulate human behavior or reactions to given scenarios. Data mining is an inseparable component when it comes to predictive analytics. We use it to identify trends, patterns and relationships. Only after the identification process is complete, can we move on to develop our predictive models. The overall goal of data mining processes is to extract information from a data set and transform it into an understandable structure for further use in our case prediction. It s important to note that many experts use the term predictive analytics broadly to describe two types of future-oriented big data scenarios: predictive and prescriptive. Predictive analytics looks into the future and helps deal with risk assessments and what-if scenarios. Prescriptive analytics focuses on analyzing what would happen based on different alternatives and scenarios, followed by a choice between best options and optimization of what s ahead. There are three phases when comes to business analytics: 1. First, descriptive analytics answers the questions what happened? and why did it happen? 2. Next, predictive analytics answers the question what will happen? 3. And finally, prescriptive analytics adds in the why will it happen? The human mind is not capable of analyzing more than 2-3 variables, let alone 50, but functions allow us to analyze an unlimited amount of variables. We can predict what our subject s future behavior will be, based on the assumption that, in general, different humans tend to respond in similar ways to specific actions. This method of using one group s behavior to predict another s is referred to as The Simple Method. So how is it done? 2
3 Take a period of three weeks (back) and divide the weeks into a two week period (green) and a one week period (red). Start from the green part which contains weeks one and two. Observe what happened during these two weeks and set up your variables according to the occurrences. Next, look at the final red week and calculate that week s actual real churn rate for each variable. Finally, go to time zero (now) and give your variables scores based on the green weeks to predict the week ahead. At CoolaData we use advanced BI technology that produces a predictive score for each customer, including logistic regression, linear regression, mark of chains and segmentation. We use these scores to build our predictive models and figure out which variables have the most effect; or, specifically, to answer questions like what s going to happen with a certain user who logged into a game and then left after making it to Level 2? The amount of variables used is what makes predictive analytics so special. The more variables included, the higher the prediction s quality (or confidence level). The regression technique s focus lies on establishing a mathematical equation as a model to represent the interactions between the different variables in consideration. Machine learning, or the ability of a computers to learn, is used in a wide variety of fields including medical diagnostics, face and speech recognition and analysis of the stock market. In some cases, the relationships between different variables can be very complex and the mathematical form of the dependencies unknown. For such cases, machine learning techniques emulate human cognition and learn from training examples to predict future events. 3
4 One of the most common predictive models is used for churn, which is useful to have once translated it into business terms. Below, we ve created a simple matrix of scenarios and actions which can be of use. Typically we ll want to work hard keeping customers on the upper right: those who spend a lot and have a low probability of abandonment. The three main types of predictive models for online businesses deal with: Churn LTV, otherwise known as Lifetime Value, is a prediction of the net profit attributed to the entire future relationship with a customer. This figure goes beyond cash and takes into account the long-term quality of a relationship with a customer, often representing an upper limit on spending to acquire new customers. For example, what will my European August revenue be? The purpose of a customer lifetime value metric is to determine the financial value of each customer, and to define, accordingly, how to channel marketing acquisition efforts. NBO (Next Best Offer) is a customer-centric marketing paradigm that is based both on the customer s interests and on the marketing organization s business objectives. It s a win/win predictive model which only recently has become technologically feasible in real-time. Using NBO enables online companies to interact with their users and answer their needs while ensuring that the action also benefits the company itself. 4
5 Big data embodies a priceless collection of experience from which to learn. Every medical procedure, credit application, Facebook post, movie recommendation, fraudulent act and purchase is encoded as data and warehoused. This delivers an amount of samples only a computer could manage to learn from. Businesses collect vast amounts of real-time customer data and predictive analytics uses this historical data, combined with customer insight, to predict future events. Each customer s predictive score informs actions to be taken with that customer. Predictive analytics uses this score to optimize marketing campaigns and website behavior, increase customer responses, conversions and clicks, and to decrease churn. Contact us now for a free trial and demo 5
Introduction to Data Mining and Knowledge Discovery Third Edition by Two Crows Corporation RELATED READINGS Data Mining 99: Technology Report, Two Crows Corporation, 1999 M. Berry and G. Linoff, Data Mining
How to Develop an Effective Sales Forecast July 2006 Introduction The ability to effectively forecast sales can have a significant and positive impact on sales and operations as well as the overall financial
Understanding Big Data with Business Intelligence Tools Analytics Help Leverage Your Massive Amounts of Information Big data is a hot topic in a variety of industries these days. Major retailers know how
How to embrace Big Data A methodology to look at the new technology Contents 2 Big Data in a nutshell 3 Big data in Italy 3 Data volume is not an issue 4 Italian firms embrace Big Data 4 Big Data strategies
IBM SPSS Direct Marketing 20 Note: Before using this information and the product it supports, read the general information under Notices on p. 105. This edition applies to IBM SPSS Statistics 20 and to
IBM SPSS Direct Marketing 21 Note: Before using this information and the product it supports, read the general information under Notices on p. 105. This edition applies to IBM SPSS Statistics 21 and to
Big Data Getting Value from Big Data: Focus on the Opportunities, Not the Obstacles Table of Contents 2 Embark on Your Big Data Journey with Confidence Getting Started, Keeping Moving 3 Big Data Hype Versus
CHAPTER 1.5 Managing the Risks and Rewards of Big Data MATT QUINN CHRIS TAYLOR TIBCO One of the biggest challenges of the term big data is deciding on a standard definition of what those words really mean.
IBM Software Thought Leadership White Paper June 2013 The top five ways to get started with big data 2 The top five ways to get started with big data Big data: A high-stakes opportunity Remember what life
An Oracle White Paper March 2013 Big Data Analytics Advanced Analytics in Oracle Database Advanced Analytics in Oracle Database Disclaimer The following is intended to outline our general product direction.
WHITEPAPER Get the Right People: 9 Critical Design Questions for Securing and Keeping the Best Hires Steven Hunt & Susan Van Klink Get the Right People: 9 Critical Design Questions for Securing and Keeping
Journal of Retailing 85 (1, 2009) 95 111 Choosing the Right Metrics to Maximize Profitability and Shareholder Value J. Andrew Petersen a,, Leigh McAlister b, David J. Reibstein c, Russell S. Winer d, V.
INTELLIGENT BUSINESS STRATEGIES W H I T E P A P E R Architecting A Big Data Platform for Analytics By Mike Ferguson Intelligent Business Strategies October 2012 Prepared for: Table of Contents Introduction...
Compliments of 2nd IBM Limited Edition Business Analytics in Retail Learn to: Put knowledge into action to drive higher sales Use advanced analytics for better response Tailor consumer shopping experiences
shop shop shop shop Retail / E-commerce Turning Big Data (and Little) Into Actionable Intelligence and Customer Profitability Part 3 in a series of 5 ebooks on intelligent customer engagement Case Study
CRISP-DM 1.0 Step-by-step data mining guide Pete Chapman (NCR), Julian Clinton (SPSS), Randy Kerber (NCR), Thomas Khabaza (SPSS), Thomas Reinartz (DaimlerChrysler), Colin Shearer (SPSS) and Rüdiger Wirth
1 The Big Data Opportunity Every now and then, new sources of data emerge that hold the potential to transform how organizations drive, or derive, business value. In the 1980s, we saw point-of-sale (POS)
COULD VS. SHOULD: BALANCING BIG DATA AND ANALYTICS TECHNOLOGY The business world is abuzz with the potential of data. In fact, most businesses have so much data that it is difficult for them to process
BUSINESS INTELLIGENCE: FROM DATA COLLECTION TO DATA MINING AND ANALYSIS Appendix W4A for EC organizations can be viewed as either transactional or analytical. Transactional data are those pieces of information
Using Credit to Your Advantage. Topic Overview. The Using Credit To Your Advantage topic will provide participants with all the basic information they need to understand credit what it is and how to make
Making Small Business Lending Profitablessss Proceedings from the Global Conference on Credit Scoring April 2 3, 2001 Washington, D.C. GLOBAL FINANCIAL MARKETS GROUP, IFC FINANCIAL SECTOR VICE PRESIDENCY,
Big Data Management and Predictive Analytics as-a-service for the Retail Industry Serendio Predictive Analytics for the Retail Industry 2 Executive Summary The biggest and most successful retailers today,
By: Rob McDougall Upstream Works Software March 2009 First Call Resolution affects every aspect of your business. Understand the ways it affects you, and the 6 benefits you ll get from improving it. Learn
BUY BIG DATA IN RETAIL Table of contents What is Big Data?... How Data Science creates value in Retail... Best practices for Retail. Case studies... 3 7 11 1. Social listening... 2. Cross-selling... 3.
SAP Solutions for Analytics Big Data Analytics Guide Better technology, more insight for the next generation of business applications Big Data Analytics Guide 2012 Big Data Analytics Guide 2012 Big Data