In this edition:

Training. Business Analytics Seminar
Case Study. Predictive Analytics for Retention and Acquisition
Services. General Information about Prediction Impact, Inc.

Data is your most valuable asset. It represents the entire history of your organization and its interactions with customers. Predictive analytics taps this rich vein of experience, mining it to offer something completely different from standard business reporting and sales forecasting: actionable predictions for each customer. If you predict it, you own it.

Predictive Analytics for Business and Marketing is a concentrated training program with an active hands-on component. In two days we cover:

• The techniques, tips and pointers you need in order to run a successful predictive analytics and data mining initiative
• How to strategically position and tactically deploy predictive analytics and data mining at your company
• How a predictive model works, how it's created, and how much revenue it generates

No background in statistics or modeling is required. The only specific knowledge assumed for this training program is moderate experience with Excel.

For course details, topic outlines, instructor bio, venue information, registration, on-site training options, and the $100 early-bird discount, see:

http://www.predictionimpact.com/predictive-analytics-training.html

If you would like to be notified of future public training sessions, email
training@predictionimpact.com.

The client is a media company generating a quickly growing revenue of $30 million and hosting a leading online dating service. The number of international paying subscribers is roughly 145,000. Since the growth in subscribers results from a mere 20% margin over a high monthly attrition rate (as is expected for this subscription service), a relatively small increase in retention provides a healthy return. For example, a 5% decrease in attrition results in an estimated $862,500 in customer lifetime values per year, plus a compound effect.

We employed predictive analytics to discover and assess the most effective metrics and indicators to predict customer attrition, as well as customer acquisition from unpaid to paid subscriptions (i.e., conversion). Both metrics that measure product usage trends, as well as demographic indicators, proved predictive of future customer attrition and conversion events. For example, customers who sign up with "Hotmail" and "Yahoo" email accounts are far less likely (20 - 25% as likely) to convert to a paid subscription than users with email addresses that may be more "permanent," such as ".net" or "EarthLink" email addresses. This could very well turn out to be an important predictor for many other online subscription-based services as well. The insight speaks directly to business strategy, such as employing incentives for customers to provide permanent email addresses, or partnering with certain email service providers. Resulting predictive models predicted customer events even more effectively by combining multiple indicators, enabling tactical deployment, such as, for example, a retention campaign strategy with a positive estimated ROI.

Acting in a more general business intelligence role, we provided several related services for this client:

• To support the above initiative in predictive analytics, as well as more general analytics goals, we developed a marketing data mart that serves as the foundation for a cohesive business intelligence system.
• We discovered a substantial, ongoing loss of subscriptions due to a credit card processing error, and advised as to a tiered business procedure that may recover an estimated $1 to 3 million in annual revenue.
• We produced general marketing reports that analyze attrition and acquisition trends over various customer segments.
• We gathered reporting requirements across multiple departments, designed a customer attrition poll for marketing intelligence, evaluated analytics software and services for purchase, designed and advised as to additional marketing metrics moving forward, and established an ongoing financial reconciliation process for internal audits, quality control, and Sarbanes-Oxley compliance.
• We helped screen candidates for a FTE Director of Business Intelligence.

Based in San Francisco, Prediction Impact, Inc. (www.PredictionImpact.com) provides consulting services over a range of disciplines, including predictive analytics, data mining, business intelligence, and text mining. Prediction Impact consultants have gained a great deal of commercial experience, working with large banks, manufacturers, retail firms, and online businesses, and enabling these organizations to learn from their data. Most Prediction Impact consultants have earned a Ph.D., demonstrating a deep technical familiarity with modeling and data mining methods.

Response Prediction
Customer Prediction
Business Prediction

For more information, contact Eric Siegel, Ph.D., President and Senior Consultant for Predictive Analytics:

phone: (415) 385-1313
email: eric@predictionimpact.com

If this newsletter was forwarded to you, you may subscribe here.