National Tax Authority Collection Program—Predictive Modeling

Summary:  Booz Allen Hamilton improved the collections strategy of a national tax collection authority by prioritizing past due accounts using predictive models, resulting in significantly increased revenue collection.

Client Challenge:  The client’s use of an antiquated workload prioritization system allowed its past-due tax collection resources to waste valuable time on uncollectible cases. As a result, the client suffered a significant negative impact on its revenues, with both total collected dollars and dollars collected per FTE declining.

How Booz Allen Helped:  Booz Allen first identified and aggregated relevant data from multiple legacy sources in order to develop predictive models that would discriminate between cases that were the most and least likely to pay past-due taxes. Our analytics experts used the predictive models to perform optimization analyses that maximized the productivity of collection resources, while delivering the case mix desired by client executives. Booz Allen then designed specialized treatment strategies based on the predicted model outcomes and applied these to the client’s case assignment system.

Results Achieved:  The client’s implementation of Booz Allen’s recommendations optimized its past-due collection strategy by prioritizing and segmenting the case load. As a result, the client collected an additional $1.1B in revenue, with similar results anticipated annually.