Monday 26 Sep 2016

SKIM presenting Experimental Conjoint at Sawtooth Software Conference in 2016

The Sawtooth Software conferences are renowned for their practical, practitioner-oriented focus and depth in the fields of choice/conjoint analysis, segmentation, and data collection/analysis. It is a forum to exchange ideas, network, and learn more about marketing analytics and choice modeling.

In the 19th edition of the event, that took place in Utah, September 2016, we offered a pre-conference session about experimental conjoint solutions and presented a paper on nested logit.

Pre-Conference Session: Experimental Conjoint Solutions

Sometimes standards are not enough to answer the business question at hand and you have to deviate from the roads most traveled. Jeroen Hardon shared this in a four-hour tutorial on advanced applications. He focused on a number of interesting extensions of choice modeling, including:

  • Custom statistical design techniques (using Lighthouse Studio and Excel), and the implications of these customizations
  • Showing different coding methods for faster processing (using Sawtooth Software CBC/HB)
  • Combining 2 MaxDiff exercises into one Latent Class run
  • Different ways of data augmentation for CBC and MaxDiff
  • Different solutions for the red/blue bus problem

Paper: The Art and Science of Nested Logit – Case Studies from Modeling Many SKUs

Nested logit is helpful for analyzing correlated alternatives. We describe how nested logit works and offer a practical guide. This includes using simple diagnostics to generate an initial nested logit structure, then stepwise building and testing of the nested structures. The case studies, prepared by Kevin Lattery, focused on SKUs in a shelf set but the process applies to any conjoint with correlated alternatives.

For more information about the event programme, check the complete Sawtooth Software Conference Brochure.