The other form of open science service to the market research community!

Check out our new R package published in The Journal of Open Source Software.

 In recent years, during discussions with specific small and medium-sized market research providers and marketing departments, we observed that studies on preference elicitation through Conjoint analysis or (anchored) Maximum Difference Scaling (MaxDiff) rarely include measures to assess the predictive (out-of-calibration) validity in holdout tasks to ensure data quality in managerial decision-making.

The reasons for this neglect may be manifold:

• Functions for implementing and validating holdout tasks are often not fully integrated into commercial software solutions that handle Conjoint and MaxDiff studies.
• A lack of time in the fast-paced market research industry
• Insufficient training for participating market research managers
• Ultimately, professional online panel providers have no incentive to address previously unmentioned data quality issues.

 Our new R package validateHOT comes into play (see GitHub for tutorials).

Beginning with a few functions we provided to our students during Conjoint seminars at the Otto von Guericke University Magdeburg, we have now consolidated several functions into a package that offers tools for validating holdout tasks and preparing basic market simulations, among other features. The package also helps with Total-Unduplicated-Reach-and-Frequency (T.U.R.F) analysis. It integrates seamlessly with (Adaptive) Choice-based Conjoint or MaxDiff as conducted by popular proprietary packages commonly used by companies.

Curious? Check out the open-access publication for free by following this link.

We wish you all some happy coding

Have you ever wondered whether incentive alignment in conjoint analysis is worth the additional costs? Based on our latest meta-analysis, we would say … yes, it is!

Conjoint analyses are often conducted in hypothetical settings where participants are not sufficiently motivated to invest the same mental effort as in an actual purchase situation. This can lead to significant losses and missed opportunities for companies relying on the resultant biased preferences. One effective workaround is to apply incentive alignment.

Our recent meta-analysis, published in Marketing Letters with 12,980 participants, shows that incentive-aligning conjoint analysis increases the prediction accuracy of consumer choices by 12%.

In the meta-analysis, we also find that incentive alignment is more effective for durable and service goods (compared to FMCGs) and when applied with adaptive choice-based conjoint analysis. The latter mirrors our recent findings from a JAMS article.

Our meta-analysis concludes with practical guidelines for market researchers. These include a detailed discussion on the cost implications of incentive-aligning a conjoint analysis, providing valuable insights and equipping researchers with the necessary knowledge for those considering this approach.

We warmly invite you to explore our open-access article, which comprehensively overviews our research findings. This is your chance to enjoy the article for free and delve deeper into the world of incentive alignment in conjoint analysis.

Ever thought about anchored MaxDiff for product choice predictions in market research? If not, it is high time to do so.

Anchored MaxDiff upgrades traditional MaxDiff by also measuring the outside good’s utility (i.e., the no-buy alternative), which converts relative MaxDiff scores into absolute ones that can also be compared between respondents.

Our recent research published in Marketing Letters investigated in a 2 (direct anchored MaxDiff vs. indirect anchored MaxDiff) x 2 (hypothetical vs. incentive-aligned) online-experiment which method yields the highest predictive validity. We show that both anchoring methods benefit strongly from incentive alignment (similar to other preference measurement techniques, as previously shown in a JAMS article.

We also show that incentive alignment predicts general demand fairly accurately, while hypothetical MaxDiff tends to overestimate demand.

Finally, we take a look at how marketing implications (e.g., product assortment optimizations) may differ between hypothetical and incentive-aligned anchored MaxDiff.

The article provides an overview of anchored MaxDiff and how incentive alignment can now be implemented in a MaxDiff study design.

Enjoy the article for free in Marketing Letters.

Should you opt for incentive alignment or adaptive designs in your choice-based conjoint market research study?

Verena Sablotny‐Wackershauser, Marcel Lichters, Daniel Guhl, Paul Bengart, and Bodo Vogt find in their recent JAMS publication that you should!

Results from 4 conjoint experiments (n=1,150) on diverse products (from pizza to fitness trackers) show

• Adaptive CBC designs compare well to incentive-aligned CBC regarding product choice predictions.
• Combining both principles delivers superior predictions.

The paper also presents a concise review of proposed adaptive designs in CBC, along with an analysis of their popularity in terms of impact factors.

• Furthermore, the relative merits of different mechanisms to incentive-align (A)CBC studies are discussed.
• All raw data and analysis scripts are freely provided via the open science framework.
• This article thus serves market researchers well in the analysis of data sets of (A)CBC studies conducted with Sawtooth Software and other solutions within R.

Happy reading with the open-access article  published in the Journal of the Academy of Marketing Science.

A series of experiments indicate that after consuming high doses of caffeine, we are likely to fall prey to marketers’ efforts to manipulate our product choice by presenting mostly irrelevant product alternatives to foster target sales.

However, our results indicate that this adverse consequence of caffeine intake only holds for high doses of 200 mg of caffeine (e.g., 20 fl oz Monster Energy or a large coffee) and real purchases. Notably, we found no influence of caffeine consumption in hypothetical product choices and low doses of caffeine.

The freely available article can be read in Marketing Letters: https://link.springer.com/article/10.1007/s11002-023-09710-6.

The article was authored by Michael Canty, Felix Josua Lang, Susanne Jana Adler, Marcel Lichters, and Marko Sarstedt.

Happy Coffee in the Morning.

Authors