The Psychology Behind Online Rabbit Holes [Recovery Tips Included]

Journal of Marketing Research Scholarly Insights are produced in partnership with the AMA Doctoral Students SIG – a shared interest network for Marketing PhD students across the world.

Have you ever visited YouTube to look for a lunch recipe and 40 minutes later found yourself watching one funny cat video after another, nary a pan on the stove or chopped parsley to show? Well, then you, like us, have experienced going down a rabbit hole! A new Journal of Marketing Research study by Kaitlin Woolley and Marissa A. Sharif posits that the “rabbit hole effect” can also be externally induced, and merely consuming similar media consecutively can increase the preference for consuming additional similar media subsequently. They further find that the reason we are drawn to rabbit holes is because consecutive consumption of similar kinds of media increases the accessibility of the media category in our minds, and we end up believing the category to be more enjoyable. 

Past influential research on the experience of flow and immersion has shown that consumers feel engaged and present in the ongoing experience, and it consequently increases their enjoyment. On one hand, it is possible that consuming similar media, one after another, leads to a feeling of boredom and decreased enjoyment. However, the authors find that the rich media experiences that we encounter on digital platforms are designed to hold a viewer’s attention and increase the choice of similar media for subsequent consumption, as is evident in the binge watching behavior of consumers. This tussle between variety-seeking needs and consistency-seeking needs is unresolved; thus, the authors attempt to unravel, specifically, what leads to rabbit hole–type behavior as exhibited by media consumers.

Through a comprehensive series of lab experiments with a variety of video stimuli and manipulations of consecutiveness, similarity and accessibility, the authors show the presence of a rabbit hole effect. They also document the underlying mechanism of anticipated enjoyment and category immersion as the drivers of this effect. Further, the authors identify potential disruptors (or not) of a rabbit hole effect by using advertisements that are similar or dissimilar to the focal stimuli category. They find that when viewers are exposed to dissimilar ad stimuli that disrupts their consumption, the rabbit hole effect is attenuated.

This research has implications for both marketing practitioners and for consumer well-being. It provides evidence for how to build rabbit hole–type viewing experiences even with the presence of ads, as well as how to use variety to increase the range of viewing experiences within subscription services. Additionally, this research provides suggestions on how to create appropriate disruptions to avoid getting stuck in a rabbit hole. 

We had the opportunity to talk with the authors to understand their research design considerations as well as some implications for generalizing the results of this study beyond video consumption. We share their responses here.

Q: Given adults in the United States spend about 470 minutes per day with digital media (and different trend reports suggest between two to three hours on digital video), why is the psychology and behavior related to digital media consumption not a more widely studied topic in empirical marketing research?

A: We would love to see more research on the psychology of digital media consumption.

There are potentially a few reasons why there is not more work on this topic at the moment. First, it may be because this is an emerging area and the landscape keeps changing. Given the research process can be slow (it may take years for a paper to come out), there may be people working on these topics and the papers have not been published yet. Second, especially given the changing landscape, it may be a difficult question to study, and to do so in a way that feels naturalistic. However, there are creative ways people are starting to study these topics (including running field studies through Twitter, for example).

Q: Given the heterogeneity in content type, which reduces consistency and comparability of stimuli, what do you recommend in the research design (methods and choice of stimuli) for managing this challenge?

A: Our solution to this in our article was to focus in on documenting a consistent finding using a narrow set of stimuli. For example, instead of examining all types of media one could consume, we started out with videos from YouTube. Only after we had a stable effect, and after understanding the mechanism, did we expand to other types of media (e.g., photos, stories). Once we understood more about why we find the effect to begin with, we could then better predict whether the effect would hold for different types of stimuli.

Q: While fascinating, investigating behaviors that usually happen in private contexts is not easy. Is there any contextual factor of the experience that you consider important but problematic to measure?

A: A challenge for us was how to measure how people stop the rabbit hole. We did this in our studies by looking at when people switch to a different type of consumption. This captures some forms of stopping the rabbit hole, but not all. For example, in reality, many times people stop the rabbit hole and go to sleep or are forced to exit to complete work or chores at home.  

Q: Does the effect extend to information consumption on social media? Does this contribute to information silos and echo chambers, and by extension, can the proposed effect of time delays and interruptions be used to mitigate these consequences?

A: We believe the effect is due in part to the immersive nature of video stimuli. We do believe the effects we document here could speak to rabbit holes people go down outside of video media consumption (e.g., doom-scrolling on Twitter), but we have not tested that. For example, we find that this effect does not hold for standalone short stories. It would be an interesting extension to see whether the interventions we uncover for attenuating the effect would also work to prevent people from falling into echo chambers on social media.

Q: The paper examines the role of ads as potential disruptors of the rabbit hole effect, contributing to the ongoing discussion about the evolution and future of TV. Since advertising represents an important source of profits even for the highly competitive market of streaming services (e.g., Netflix added ads to its basic plan again in November 2022), which other factors, besides category similarity, do you think would be critical when deciding the type of ads to include on these types of services?

A: It is possible that outside of category similarity, other types of similarity may help people to focus on ads when they are in the middle of a consumption experience (e.g., similar tone, color, themes), even if the content is not the same. We also find in our studies that labelling can make content seem more versus less similar; thus, merely labelling the advertisement in a way to make it appear more similar to the video content could be helpful. Marketers should also consider the goal with the advertisement, whether it is meant to lead to a sale or to increase brand awareness, because that will also influence the types of decisions made when creating ad content. 

Q: You include a plentiful variety of stimuli for media consumption in the studies with a dominant visual sense stimulation—images and videos. How would you conceptualize similarity for media consumption when auditory stimulation is dominant (i.e., podcasts)? Do you think that using the podcaster’s voice instead of a different voice would increase similarity with the content and therefore facilitate the rabbit hole effect?

A: We see a natural extension of our findings from videos to music. That is, some of our studies used music videos: We could see people falling into a rabbit hole when listening to music on Spotify for example, even if they were not watching music videos. It is interesting to think about how this may extend to podcasts. It is possible people will fall into a rabbit hole where they seek out similar types of podcast content (although this may depend on the type of content and how immersive it is—as prior research finds certain content, like watching the news, is less likely to lead to binge watching—we could see podcasts falling into a similar category). A place where we think you could get some traction is with advertisements—sometimes podcast hosts also deliver ad content on their shows (i.e., they keep the voice the same), and other times the ads are delivered by someone other than the podcast host. Keeping the person delivering the podcast content and ad content constant may be beneficial. 

Read the full article:

Woolley, Kaitlin and Marissa A. Sharif (2022), “Down a Rabbit Hole: How Prior Media Consumption Shapes Subsequent Media Consumption,” Journal of Marketing Research, 59 (3), 453–71. doi:10.1177/00222437211055403

Go to the Journal of Marketing Research