Predictive Analytics
Services & Case Studies
Our Analytics Services include full customer cycle modeling.
• Predictive/Adaptive modeling
• Full customer cycle Modeling including acquisition, retention,
upsell/crossell, risk, reacquisition etc. • Predictability analysis
• Multi-perspective data mining • Dynamic data mining
• Customer profiling and segmentation • Media/Campaign effectiveness analysis • Data collection strategies
• Data quality assessment
• Analytics readiness assessment
• Analytics opportunity assessment and road mapping • Analytics tools evaluation
• Implementation
Analytics and Predictive Modeling Case
Studies
New Student Acquisition—Improving
ROI in For-Profit Education
The Need
A multi-campus vocational/technical institute knew it was overspending to get qualified students enrolled and was losing too many between enrollment and start date
The Solution
• A set of models that modeled the full student lifecycle from acquisition of optimal leads through enrollment, commencement of classes and graduation. Solution separated out low performing lead sources, leading to millions of savings in student acquisition cost.
• A dynamic, cutting-edge model to adjust start probability based on student-counselor interaction, leading to better use of admissions counselor time, better forecasting to capacity and a significant improvement in the start right of high quality enrollees
Single-line Retailer - Does Our Radio
Promotional Strategy Move Product?
The Need
A medium-sized chain of nutritional products stores wished to
understand if radio advertising drove incremental sales for featured products and for all products.
The Solution
• Cross-analyzing ad frequencies, station formats and day parts with product sales led to findings that radio moved the needle for one gender segment, but not the other, and that the ad spend would be more effective if it better fit the seasonality of the business.
• The analysis established a platform for testing whether the better choice was to reduce overall ad spend, or reinvest it into the most productive time slots and station formats.
Grocery Chain - Using Buyer Behavior
Analysis to Direct Growth Strategy
The Need
A leading grocery chain was struggling to get actionable information from their data, which was housed in silos and limited their ability to make
decisions.
The Solution
• In just eight weeks, we were able to integrate the data, create an effective customer segmentation and increase the accuracy of customer behavior predictions by 15%.
• The model outputs were used by the client to alter their expansion plans, saving millions in potentially bad product and store set
Major Toy Retailer - Does Loyalty
Equate to Profitability?
The Need
A large toy retailer was losing share to larger, full-service general merchandisers and wanted to better understand customer loyalty.
The Solution
• Combining attitude & usage research data, syndicating industry
information and transactions data samples, we identified that “highly loyal” customers not only spent less than the average toy consumer in their stores, but were also price sensitive and therefore susceptible to poaching by lower-priced retailers.
• The client used the insights to reevaluate their ad targeting strategy to begin to stem the tide.
Catalog Customer Acquisition — Using Lifetime
Value Metrics to Guide List Purchasing
The Need
A consumer catalog retailer with a multi-million customer housefile was using a single customer lifetime value assumption to measure the
performance and guide the purchase of rental lists for new customer acquisition.
The Solution
• We combined our database management skills with our understanding of direct marketing financials to build a multi-dimensional lifetime
value reporting and assessment tool.
• This allowed for scoring of list sources and estimation of the future lifetime value of the prospects presented. The company was able to better eliminate non-productive list source and improve their
Insurance
The Need
A major international insurer sought to improve customer lifetime value through. Their existing business model failed to meet even modest sales goals, let alone achieve cross-selling objectives.
The Solution
• Rich, actionable profiles to effectively identify the right customer segments, and predict behavior within them.
• Models identified those most likely to purchase a given type of policy. The client realized a 26% improvement in sales
conversion over previous best practices, lowered cost of acquisition and increased retention rate.
Health Care
The Need
The USA 's leading provider of health care risk models was
exposing itself to unnecessary risk because its actuarial approach was allowing high risk policyholders to slip through. They needed a solution to better predict which individuals could be classified into the top 0.5% of the most likely to lead to high medical costs.
The Solution
• Our models enabled senior management to improve predictions with 53% accuracy—a significant improvement over the old
method.
• The models provided a direct impact to the company’s bottom line - a net savings over $11 million.
Technology Distribution
The Need
A Fortune 500 technology solutions company sought to reduce high selling costs associated with constant re-quoting of
opportunities
The Solution
• Our model drew out the characteristics associated with successful and profitable quotes and was able to identify 85% of quotes that would go on to close from just 30% of those scored
Identity Theft Protection
The Need
A leading provider of identity theft protection services was
experiencing a high attrition rate among its subscribed member base
The Solution
• We built an early-warning model that identified over 90% of potential attriters within the top 50% of scored cases
• This allowed the company to focus retention team resources on the most problematic cases, and projected to avoid $10 million in annual subscription losses