top of page


Insights Association
North Central Chapter

Fall Conference

September 4-6, 2024

Detroit, MI USA

Transformative Insights: The Journey from Research in Sustainable Innovation

Megan Peitz

In a world where strategic innovation is pivotal for lasting impact, our presentation, "Transformative Insights," explores the profound role of research in shaping long-term business strategy and fostering social transformation. Through a compelling case study, we showcase the partnership between Numerious and Mill—a company dedicated to preventing food waste and building a sustainable future.

Join us in person!

Click below to register.

INFORMS Annual Conference
October 20-23, 2024
Seattle, WA USA

Building Designs for Individual-Level Estimation:
Considerations, Implications and New Tools for Choice-Based Conjoint

Trevor Olsen & Megan Peitz

Choice-based conjoint (CBC) experiments are widely used to understand consumer preferences and willingness to pay for different product features. One important consideration in designing CBC experiments is the balance of attribute levels across the design. Implementing this strategy seeks to give every level an equal chance to influence the respondent’s decision in the conjoint design and can work in the majority of cases. However, the authors of this paper were interested in revisiting the work of Huber and Zwerina (1996) to determine if utility balanced designs, a design strategy that trades off on level balance while optimizing which alternatives are paired against each other within tasks, could result in better predictions at the individual level. This paper sets out to explore several different methods of optimizing designs and offers access to an open-source package, built in Julia by the Numerious team, to leverage these different design strategies in the future.

The results from this paper show that utility balanced designs perform well in predicting data from both utility balanced and non-utility balanced designs, and that respondents do not seem to be fatigued by utility balanced designs. This would suggest that utility balanced designs could be a successful strategy depending on the attributes and levels being tested. However, we must caution the user of utility balanced designs as some design strategies may result in sparse data at the interaction level. We also believe that further research is needed to understand the differences in willingness to pay estimates between utility balanced designs and traditional, level balanced designs. 

Join us in person!

Click below to register.

bottom of page