BiTURF

Quantifying uncertainty to enhance strategic decision-making

Abstract

Total Unduplicated Reach and Frequency (TURF) analysis is widely used in both sensory product and market research for determining optimal product assortments under externally imposed constraints (e.g., a limited product development budget or a retailer's shelf space). However, researchers and practitioners face methodological limitations when applying classical TURF analysis, including the lack of quantification of uncertainty in results or issues with sparse data. To overcome these methodological limitations, this work introduces BiTURF—TURF analysis enriched with Bayesian input data—as a proof-of-concept and compares it with classical TURF analysis using two exemplary data sets. This work thereby demonstrates how to apply BiTURF to data from studies that use anchored Maximum Difference Scaling (i.e., Best-Worst Scaling) and Check-All-That-Apply data. In a nutshell, BiTURF quantifies the uncertainty of reach and frequency estimates and enables further analyses, including probabilistic head-to-head comparisons. This information provides researchers and practitioners with a richer foundation for their decisions. Furthermore, BiTURF overcomes the methodological corset of classical TURF, which can't be used with the weighted by probability approach on Check-All-That-Apply data. The article also provides a step-by-step R tutorial and concludes with a discussion of BiTURF's limitations and future research endeavors.