Store Segmentation

A major national manufacturer of hardware/paint supplies wanted to develop within-type performance benchmarks and best practices. The primary motivations behind this effort were to be able to deliver targeted treatments to its member base and to enable peer group comparisons resulting in the identification of a ‘best-in-class’ category of members. This client engaged Analytic Innovations to define distinct segments of members, based primarily on member feedback captured via a standardized survey instrument. Analytic Innovations broadened the applicability of the newly developed segmentation by creating a model-based tool with which new members or existing members that did not participate in the initial survey could easily be classified into segments. By overlaying performance metric to segments, the client was able to deliver strategic and actionable advice to improve member performance.

Mystery Shopper Analysis

A major national manufacturer of hardware/paint supplies wanted to member attributes strongly associated with customer satisfaction. In order to gather data for this analysis, the manufacturer initiated a mystery shopper program. Unidentified individuals shopped at selected stores and recorded their experiences via a standardized survey instrument. Analytic Innovations advised this client on the most appropriate analytic methodology to extract insightful information from these data resulting in actionable strategies. Analytic Innovations employed data mining techniques to identify and prioritize a set of factors distinct from basic executional elements that isolated positive experiences. The analysis enabled the client to identify member training opportunities most likely to yield a better customer experience.

Targeting and Feature Extraction

A leading provider of limited-edition collectibles sought to increase the targeting efficiency of their direct marketing campaigns. They approached Analytic Innovations to create traditional responder targeting models. Analytic Innovations developed a response model which improved upon their existing targeting criteria for a particular product offer. During this initial engagement, Analytic Innovations advised the client about deficiencies inherent to the development of traditional targeting models for a portfolio of products numbering in the thousands. Analytic Innovations proposed testing an innovative approach to enhancing their existing business intelligence platform. This approach is based on cutting-edge data mining technology developed by the Research division at Analytic Innovations. With this technology, Analytic Innovations is able to explore large volumes of very granular data gathered from transactional systems and extract latent product features and customer preferences that drive observed purchasing behavior. This client can use extracted product features to discover new opportunities in product development and marketing. Furthermore, customer preferences described in terms of product features, will be a key enabler for this client’s customer-centric business strategies.

Broadcast Performance Analysis and Forecasting

A major provider of news and entertainment products engaged Analytic Innovations to assist its internal research unit by answering fundamental questions about the historical performance of shows broadcast on its cable television networks and predicting future performance.

Project 1: The primary objective was to develop a prediction system, which could be used to estimate the performance of movies, thereby enabling strategic decisions regarding acquisitions and broadcast patterns. Prediction systems were required to operate under circumstances with minimal historical data availability and where prior network and external box-office data could be exploited. Analytic Innovations, working collaboratively with the client, operationalized an important business practice that has a profound impact on performance. By building upon findings from an intensive and iterative research process and by incorporating insights from this client, Analytic Innovations developed a set of performance estimators. These models, now in their third generation, are capable of being deployed in ‘limited’ and ‘full’ information implementation environments. The models have been integrated into the client’s internal business intelligence system and are an essential component for planning and forecasting.

Project 2: The primary objective was to develop a rigorous statistical foundation for understanding relationships between programming strategy and network performance. This research was conducted in phases with learnings from Phase 1 resulting in Analytic Innovations making recommendations for significant data enhancements for Phase 2. In the latter phase, Analytic Innovations used relationships based on statistical modeling to generate a variety of simulations demonstrating the impact of changes in individual programming strategies on the decay trajectory of series modeled. The client expects to incorporate these findings into the formulation of new programming strategies.

Project 3: The primary objective was to investigate the role of broadcast frequency, time of broadcast, and other factors on the performance of kid’s animation series. Analytic Innovations conducted a variety of statistical analyses to establish numerical relationships that confirmed internal hypotheses and helped define programming thresholds. These analyses established an optimal broadcast repeat rate below which the audience is still growing and above which there is a drop-off in performance. By analyzing additional data, Analytic Innovations was able to establish comparative metrics across multiple networks and across multiple genres. The findings from this research will enable this client to identify new acquisition opportunities within their own networks and to establish programming parameters to minimize loss in performance due to over-exposure within a target demographic.

Product Sequencing

A technology pioneer and leader in instrumentation products sought to discover product relationships in order to better manage their direct marketing efforts. They engaged Analytic Innovations and provided purchase data encompassing a product portfolio with tens of thousands of items. Analytic Innovations used techniques such as Product Sequence analysis and Market Basket analysis to reveal which products are typically purchased in combination and the time dependencies among these purchases (e.g. inter-purchase time, the likelihood that the subsequent purchase will occur in the same day or that there will be no subsequent purchase). These findings enabled the client to better understand product groupings, triggering products, the identification of the next product, and an estimate of the time to next product purchase.

Donor Segmentation

A leading children’s advocacy group sought to develop a long-term donor targeting strategy. A partnering agency with an expertise in communication strategy, retained by this group, formed an alliance with Analytic Innovations to supply a data-driven analytic solution, which would be used to drive the design and implementation of future marketing tactics. The solution was multi-faceted, yet could be readily implemented to target various sub-groups within their donor portfolio with appropriate messaging. Analytics Innovations employed a cohort-based lifecycle analysis to describe trends within the entire portfolio. Further, a segmentation approach was used to characterize the donor base by amount of giving. An analysis of migration patterns among these segments revealed sub-populations of donors likely to be good ‘hits’ for existing tactics. The analysis also revealed opportunities to create new segment-specific marketing tactics. In addition to these central findings, Analytic Innovations also made additional recommendations (contact management, control groups, data cleansing, studying long-term lapsed donors, etc.) necessary to maintain an active donor portfolio.