Just how many adorable cat videos or pictures of fancy desserts can a person view in one sitting?

According to new research from Kaitlin Woolley ’12, associate professor of marketing in the Samuel Curtis Johnson Graduate School of Management, they’re kind of like potato chips: You can’t consume just one.

“You’d expect people to be satiated – if I watch one episode on this topic, I should be ready to move on,” Woolley said. “You’d think people would get less enjoyment from viewing something similar after already seeing one or two related videos, but in our studies, people wanted to keep watching more and more.”

Woolley and Marissa Sharif, an assistant professor of marketing at the Wharton School of the University of Pennsylvania, are co-authors of “Down a Rabbit Hole: How Prior Media Consumption Shapes Subsequent Media Consumption,” which published Oct. 11 in the Journal of Marketing Research.

Woolley and Sharif conducted numerous studies over the past four years to understand the drivers of what they termed the “rabbit hole effect” – being sucked into repeated viewings of similar media, instead of switching topics or activities altogether.

Woolley said considerable research on binge-watching – particularly of serial TV shows – helped inform their work, although theirs is a different phenomenon.

“If I’m watching a series,” she said, “the shows are building on each other, and they often end with a cliffhanger, so you need to get to the next episode to see what happened. It’s not necessarily because of the similarity in terms of the content, which is what we were looking at.”

For their paper, Woolley and Sharif conducted a total of 13 studies, of 150 to 200 participants each, in which participants were shown multiple videos or images. Through various forms of manipulation, the researchers measured how likely people were to either watch another similar media sample or something different.

“We ran so many studies because it was a very surprising effect to us,” she said. “After the first few studies, we were a little surprised, so we would run more studies to understand if it was a stable effect, and what was driving it.”

They found that three factors – similarity, repetition and consecutiveness – drove the rabbit hole effect. When these three characteristics of media consumption are combined, they found, consumers become immersed in the category and expect to derive continued enjoyment from similar media.

Watching videos on a particular topic makes that topic top of mind, Woolley said, which keeps people clicking on the next cat video, and then the one after that.

“You watch one video on a particular topic and the topic itself is not that accessible, but then as you watch another one, suddenly that category becomes more accessible,” she said. “Then it’s harder to switch because you are immersed in the topic, and there may be some perceived cost to switching.”

In addition to the cognitive belief that there’s a potential negative effect to switching, there’s also the motivational component. “People anticipate that options within that same category are going to be more enjoyable [than another category],” she said. “Because the category is more accessible, and people feel immersed in it, people think similar options will be more enjoyable.”

In one of the studies, Woolley and Sharif manipulated the labeling, showing participants the same two videos but labeling them “educational videos” for some and simply “videos” for others. The people who watched the labeled videos were more likely to click on the next video in the category.

“You’d think that people might be able to make those connections without those labels – they are watching the same videos in both cases,” Woolley said, “but having the overarching similarity pointed out to them by the category label leads people to say, “OK, this is the particular content that I’m watching.’ It’s activating the category in their minds, and then they’re choosing in line with that accessible category.”

The researchers found, however, that building a break into their viewing actually pulled people out of the rabbit hole. In their final study, they had people watch two videos of a certain topic, then choose which videos they’d like to watch in a couple of days. When people were asked two days later what they would like to watch, they tended to choose something completely different.

“One of our key takeaways,” Woolley said, “is that, for consumers, it’s good to create a situation where you’re going to be interrupted. Before you sit down, find a way to have some interruption coming. Otherwise, you could tend to fall down the rabbit hole, without necessarily realizing it.”

This work was supported by Cornell, the Wharton Dean’s Research Fund and Wharton’s Behavioral Lab.