This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).
This Small Business Innovation Research (SBIR) Phase II project involves the examination of consumer consumption behavior across multiple on-line domains to predict those items to be most likely consumed in the next interchange and the terms under which they will be consumed. The proposed innovation utilizes a persistent key technology to examine multiple attributes of identity to establish a unified identity that links individuals across multiple domains. Once linked the unified identity serves as the basis for the aggregation of consumption behavior (purchases, content, ads clicked through, invitations extended, etc.). The aggregated data establishes the consumer?s digital footprint and serves as the basis for creating highly-predictive models. The models analyze the actual consumption behavior to establish consumption propensity and terms of consumption on an industry segment level. The results of the propensity models will be returned to the client at the time of interaction to make up sell / cross sell offers that are most likely to result in action by the consumer. The result for the client is increased revenue for the transaction and the result for the consumer is increased satisfaction through the relevance of the offer.
The broader impact of the proposed innovation involves three aspects: Accelerating economic expansion, identifying potential domestic terror threats and identifying potential on-line predatory activity. The ability for a retail or social network to identify the consumption preferences of their customers and offer those items during an interaction increases the likelihood that a customer will purchase the offered item due to its relevance. Such expansion of customer spending will assist organizations in increasing inventory turnover, improving sales and overall economic health. Identification of potential domestic terror threats through the examination of cross domain purchasing behavior of linked identities. Intelligence Services could establish purchase combinations that when combined could result in a potential treat and take appropriate early intervention action. Identification of potential on-line predators through the use of persistent key technology to highlight those individuals whose established identity on other domains is materially different from a current registration. This permits the organization to establish higher authentication requirements for those individuals and in so doing protecting itself and in the case of Social Media its members (specifically minors).
Today’s increasingly digital economy is shaped by the consumers demand for more relevant content at each interaction, reduced SPAM and increased privacy protection. The focus of the research was to determine if: 1) aggregated data establishes the consumer’s digital footprint and serves as the basis for creating highly predictive models, 2) the actual consumption behavior establishes a predictable consumption propensity and terms of consumption on an industry segment level and 3) it is possible to return the results of the propensity models at the time of a consumer interaction to make up sell / cross sell offers that are most likely to result in action by the consumer. The desired outcome for the client is increased revenue for the transaction and the result for the consumer is increased satisfaction through the relevance of the offer. The results of the research indicated the following: Prior purchase/response activity is highly predictive of response rates to offers and correlates highly to subsequent purchase activity. This allows a forecast expected conversions and revenue for a subset of customers over a variety of offer scenarios. Sufficient data attributes exist to determine with a high degree of reliability the future state of a customer (Active, Lost, and Re-activated). This allows the identification of at risk customers for differentiated treatment in order to maintain a relationship. It is possible to create a unit elasticity analysis for small subsets of customers within a client’s customer database. This allows an understanding of the discount point on a basket of goods that result in the optimal revenue from the customer subset. Sufficient data attributes exist to provide product recommendations which are most likely to be consumed during the next purchase interaction. And these results can be delivered in real time. It is possible to identify untapped demand (capacity) through linking geo-demographic segmentation systems, which act as a surrogate for affluence models, with the scoring technology. Sufficient data attributes exist to determine the most effective Communication Channel to customers. This allows communication with the consumer in the "Channel of Choice" thus helping to reduce customer fatigue, increase response rates and optimize response rates. There is a high degree of correlation between PowerVue® Score and traditional RFM models with PowerVue® Score offering superior discrimination by the nature of its continuous range of score assignment and its alphanumeric construction which allows visualization across a two-dimensional plane. This allows a more effective targeting of the consumer for offers by classifying the consumers more precisely than has been available in the past.
