Last week RTi ‘s vice-president Marketing Services, Howie Firestone, joined approximately 200 other researchers and academicians from around the world at Sawtooth Software’s 17th onference. Sawtooth Software is a leading provider of software to the marketing research community. The software products they provide enable researchers to design and analyze complex conjoint exercises.
While Sawtooth Software focuses primarily on conjoint analysis, the presentations addressed a variety of research and marketing science techniques and applications. Highlights of the conference included presentations on:
- Conjoint Design: Presentations covered the spectrum of conjoint techniques – Discrete Choice Based Conjoint (CBC), Adaptive Choice Based Conjoint (ACBC), and Menu Based Choice (MBC) and Best-Worst Measurement (MaxDiff). The presentations offered innovative ways to execute these studies in order to better engage respondents (thru shorter and more interesting exercises) and generate higher quality data. The techniques included adaptive techniques that build on responses to earlier questions, tournament like exercises, and “Share Back” techniques that tell respondents how their desired product compares to other respondents. RTi plans to add these techniques to the toolkit and apply them where appropriate.
- Best Worst Measurement (MaxDiff): While Best-Worst Measurement falls under the broad category of Conjoint Design, there were a number of interesting presentations related to Augmenting MaxDiff exercises via Q-Sort techniques, and applying MaxDiff scaling to Brand Equity measurement and Concept screening. Using anchored MaxDiff exercises which provide an absolute (positive – negative) dimension to the MaxDiff (which is relative measure) exercise will allow RTi to replace straight rating scales for these studies, thereby providing more discriminating and better quality data while shortening the questionnaire and reducing respondent fatigue.
- Shapley Value: While Shapley Value is primarily considered a Line Optimization tool, the principles of Shapley Value can be extended to Regression Analysis (a.k.a. Average over Ordering) and Customer Satisfaction research (Source of Dissatisfaction analyses).
- Segmentation Analysis: We were glad to see that the techniques we employ for segmentation analysis (Cluster Ensembles via Sawtooth’s CCEA product) was highly regarded by the practitioners who presented at this conference. They showed some interesting ways of augmenting the ensemble which should enhance the learning from these studies.
RTi will incorporate the key learnings from the conference to improve the research programs we conduct.