It’s a well-known fact that the rise of the Internet has led to a decline in the amount of television people watch. With only so many hours in the day to dedicate to leisure activities, it should be no surprise that growth in one should take a toll on the other. But it’s also the case that we live in the age of the multi-tasker: Smartphone in one hand, TV remote in the other, a book in our lap, and food on the table in front of us, we rarely focus all of our attention on just one thing anymore. 

All of which leads to an interesting question: Does e-commerce decline on days when TV viewership surges? Or, on the other side of the coin, will people spend more money shopping online if, for example, the big game turns into a boring blowout, or when reruns dominate the airwaves? In a new research paper, “TV’s Dirty Little Secret: The Negative Effect of Popular TV on Online Auction Sales,” Wharton operations and information management professor Shawndra Hill, Oliver Hinz of the Technical University of Darmstadt, Germany, and Ju-Young Kim of Johann Wolfgang Goethe University in Frankfurt-am-Main, Germany, dug into the data to learn the answer. 

In an interview with Knowledge at Wharton, Hill explains their results, showing that watching TV does appear to substitute for online shopping to a measurable degree. 

An edited transcript of the conversation appears below. 

Couch Potatoes vs. Online Consumers:

In this particular study, we were interested in the relationship between television viewership and online sales, and asking the simple question of whether they’re substitutes or complements. The way that we studied this is we gathered a large, comprehensive data set on viewers in Germany … and sales data from a company similar to eBay – a large auction site. 

“We found that TV viewership, in this case, is in fact a substitute for online sales of the particular online retailer.”

We also collected a lot of other data that we used as controls: weather data, because weather has been shown in the past to impact online sales; average bank account balances from a large bank to get a sense for the [retail] potential for the country on any given day; and for eBay in Germany, we got online sales as well as their advertising span to get a sense for the competition for this particular company. In short, we tried to control for anything else that could possibly explain the online sales. 

And we found that TV viewership, in this case, is in fact a substitute for online sales of the particular online retailer.

How We Spend Our Attention Budget:

We found that there’s a negative relationship between TV viewership and online sales. And one might think that’s not surprising. However, we began with two competing hypotheses. On the one hand, you might think, “Well, you have an attention budget. So, you can’t watch television and use the Internet at the same time.” And therefore, one should be a substitute for the other. However, before we performed this study, we actually ran a survey where we asked a large number of people in Germany about their online behavior, specifically whether they used the Internet while they’re watching TV. 

A competing hypothesis would be that since there are more people online and at home while watching television, online sales might actually go up. We were able to show, at least in this context, that TV viewership has a negative relationship with online sales.

“Asking viewers to participate in polls, to tweet, to post on Facebook — all of these activities potentially have the unintended consequence of impacting online sales.”

Key Takeaways and Practical Implications:

First, TV viewership, at least on days where there are huge numbers of viewers watching shows, should be factored into models of predicting online sales. For auction sites in particular, sellers may want to pay attention to the closing dates and times of their auctions, and not have them coincide with times when popular television shows are on.

There are a lot of potential implications in our results. First if people can only pay attention to a limited number of activities at any given time, that would have implications for TV shows that prompt viewers to participate in activity while the shows are going on. Asking viewers to participate in polls, to tweet, to post on Facebook — all of these activities potentially have the unintended consequence of impacting online sales. And online sellers — and companies in general — are the customers of TV shows. They buy advertising, and ultimately don’t want their sales impacted. 

The TV industry is really overnight becoming revolutionized in the amount of data that shows and networks can collect, as well as TV measurement companies. And so, pretty much anyone involved in television — producers, networks, advertisers – will be interested in the relationship between what’s happening on television and what’s happening online.

Specifically, people tend to focus on engagement – how people are responding online as it relates to social media. But also there are new ways to measure audiences in terms of their demographics and what they’re doing, if they’re buying. But in all of this, the nice thing is that we have all of this data now that we can link and view in interesting ways that can help us better understand the impact of TV on viewership and sales and social media in ways that we couldn’t even imagine before.

“For auction sites in particular, sellers may want to pay attention to the closing dates and times of their auctions, and not have them coincide with times when popular television shows are on.”

There’s a lot of hype about “second-screening,” and what people are doing while they’re watching television. And I think there’s this general belief that it’s actually a good thing for television shows and for online sellers and for sellers in general. Here, for the first time, we have data that’s saying, “Well, maybe we should take a step back and think about what the impact of television is on online sales. Maybe second-screening isn’t positive for sellers just yet. 

How the study stands apart:

What sets this research apart is really the data: data on viewership at the hourly level as well as all of the sales data for this online company. It enables us to answer these questions in the first place, and, we think, for the first time.

Next Steps:

We have a plan for building this study in the U.S. context using U.S. viewership data and looking across different online retailers, to focus more on which types of online firms are impacted the most. 

In preliminary analysis, what we found was in the U.S., on days like the Super Bowl and NBA finals, where people are watching television in large numbers, while on average sales go down, sales for pizza go up. So it’s possible that while on average, high levels of television viewership impact sales in a negative way, for some categories of products, sales might actually go up. The plan is to see if there are variants across categories as it relates to television viewership and online sales.