The Role of (Dis)similarity in (Mis)predicting Others’ Preferences

Kate Barasz, Tami Kim, and Leslie K. John
Article Snapshot
Current average rating    
Key Takeaways

People (erroneously) predict that others dislike dissimilar things even when the dissimilar item is more highly rated and even when financially incented to make an accurate prediction

In reality, people’s actual preferences are broad and diverse; individuals report liking a range of options within a category, and often actively like (or even prefer) dissimilar options.

The discrepancy between actual and predicted preferences is driven by a more fundamental misperception: people have a tendency to believe that others’ preferences are more homogeneous than they, in fact, are.

Article Snapshot​s: Executive Summaries from the Journal of Marketing Research​​​

People assume that others’ preferences are narrower and more homogeneous than they actually are, and as such, (erroneously) predict that others must dislike dissimilar things.



Research

In this world of social media sharing—where ample opportunities exist to observe the things other people "like" or have purchased—we wondered: what inferences are people making about things that are forgone or unchosen? For instance, if you see that someone "liked" a documentary on their Facebook page, or Tweeted about an orchestral performance they recently attended, what inferences do we make about that person's preferences for thriller films or Top 40 Pop music? As we found, people make predictions systematically: similar options are liked, but dissimilar options must be disliked.

Method

We conducted five experiments, both online and in the lab, and either asked people to specify their own tastes ("actual preferences") or to predict someone else's preferences ("predicted preferences"). By comparing actual and predicted preferences, we were able to see whether people were accurate in their assessments. We found that while people were accurate in their predictions for similar items, they routinely failed to predict others' actual preferences for dissimilar goods.



FIGURE: When faced with the choice between a higher-rated dissimilar option or a lower-rated similar one, people readily choose the high-rated dissimilar option for themselves; however, even despite the quality trade-off, few observers predict that choice.


Findings

People enjoy dissimilar things themselves (e.g., vacations to both a lake and city), but predict that, for others, a preference for one precludes enjoyment of the other. We found evidence that this mistaken inference is driven by a more fundamental misperception: the belief that others have narrow, homogeneous preferences. In other words, once we observe someone choose one thing (e.g., lake vacation), we tend to believe that all of their other preferences must be consistent and alike (e.g., "he must only like outdoorsy vacations"), thus failing to recognize the diversity of others' tastes.

Implications

This research examines the inferences we all make about others' unspecified tastes—predicting how much they like an option that wasn't chosen after observing an option that was—which has applicability across many domains, from interpersonal inferences, to agent decision-makers (e.g., realtors, money managers), to marketers.  As our results and General Discussion section suggest, the findings may have particular relevance for practitioners:  just because you've observed a customer purchasing one thing, don't automatically assume they're not interested in something dissimilar.   


Questions for the Classroom

  • In what domains or industries do you think this mistaken inference might be especially prevalent? Especially problematic?

  • How might these findings change how you would craft a marketing plan?

  • How do you think consumers would react if recommendation agents began suggesting explicitly dissimilar products?


Full Article
Kate Barasz, Tami Kim, and Leslie K. John (2016)., "The Role of (Dis)similarity in (Mis)predicting Others’ Preferences." Journal of Marketing Research, 53 (4), pp. 597-607.
doi: ​http://dx.doi.org/10.1509/jmr.15.0226​


Kate Barasz is a doctoral candidate in Marketing, Harvard Business School, Harvard University (e-mail: kbarasz@hbs.edu).

Tami Kim is a doctoral candidate in Marketing, Harvard Business School, Harvard University (e-mail: tkim@hbs.edu).

Leslie K. John is Assistant Professor of Business Administration, Harvard Business School, Harvard University (e-mail: ljohn@hbs.edu).


Author Bio:

 
Kate Barasz, Tami Kim, and Leslie K. John
Add A Comment :
 

Become a Member
Access our innovative members-only resources and tools to further your marketing practice.
 

 Sign Up for AMA Journal Reader

 
Receive Article Curations from AMA Journals delivered to your inbox!


 

 Upcoming Conferences