Obviously, you would choose another activity to spend your time on, perhaps a different online destination. But understanding what this substitution looks like on a larger scale is difficult, in part because experts generally rely on prices to understand how consumers substitute one product or service for another.
“In an economics-type 101 market, like for apples, if prices for some apples go up, we want to know, ‘Do people substitute other kinds of apples, buy oranges, or stop buying fruit?’ Aridor says.
However, there is no obvious price for the most popular apps. Companies such as Facebook, Twitter and YouTube do not charge for access to their platforms, but instead sell audience attention to advertisers within the so-called attention market.
To better understand how app replacement works, Aridor conducted an experiment to find out how consumers would react if they were temporarily unable to use Instagram or YouTube at all. Would users substitute another app in the same category, an app in a different category, or reduce their app usage altogether?
Aridor found that people easily switched to other categories of apps, although this is partly because many apps cover multiple categories: a social networking app can also be used for entertainment, for example.
He also found that the habit-forming nature of digital apps, a force he calls inertia, played an important role in the apps people chose as substitutes. “If you can’t use Instagram, you’ll be more likely to substitute with apps you normally use,” says Aridor. Or instead of going to the trouble of downloading a brand new app, you might prefer not to replace it with another app at all.
This information is obviously of interest to tech companies and marketers, who would benefit from a better understanding of how consumers replace one app with another, Aridor says. And antitrust regulators, who have struggled to define competitive markets when evaluating potential technology mergers, could also benefit.
Understanding the behavior of application users
While many popular apps do not explicitly charge for access to their platforms, there are still many ways that the “price” these apps charge could be increased. “Facebook can put a price on user attention by increasing the number of ads in your feed. But the associated price may also include product quality or privacy concerns,” says Aridor.
For a real-life example of how this could happen, look at Twitter. Since Elon Musk bought the company in October, it has experimented with real price increases, such as charging for a certified account, and non-monetary price increases, such as dramatically reducing content retention. How Twitter users will respond to these price increases remains an open question.
“If there’s a huge degradation in the quality of a product, users may switch to niche apps like Mastodon, spend more time on other premium apps, or spend more time on their phones altogether,” says Aridor.
Beyond Twitter, there are broader questions about how people might replace a particular app if its non-monetary usage price increases.
To gain insights into this phenomenon, Aridor recruited around 400 students from the US, Hong Kong, Italy and Switzerland. The average age of these paid participants was 26. They all used Android phones to access YouTube, Facebook, Instagram, Snapchat and WhatsApp.
Aridor had them download a parental control app on their phone, which allowed him to cut off their access to Instagram or YouTube as well as monitor how they spent their time on the phone. (They had been warned in advance that they might lose access to an app, but weren’t told specifically that it might be Instagram or YouTube.) They also downloaded a browser extension to their laptops to collect data about their websites they used on their computers during the study. Aridor surveyed participants several times before, during and after they cut off app access, asking questions about how they spent their time away from digital devices and what they generally use different apps for.
In the first week, Aridor collected baseline data about the time participants spent on WhatsApp, YouTube, Instagram, Facebook, Messenger, Reddit, Snapchat, TikTok and Twitter. (Aridor categorized each app based on its assigned category in the Google Play Store. For example, Instagram and Facebook were categorized as social apps, while YouTube fell under entertainment. WhatsApp was categorized as a communication app.)
One group of participants was then restricted from either YouTube or Instagram for one week, while the restrictions for another group lasted for two weeks. There was also an unrestricted control group. Aridor collected data about the time users spent in other applications during and after the restricted periods.
Participants who lost access to Instagram or YouTube spent less time overall on their phones. Those cut off from Instagram spent an average of 27 minutes less, while those deprived of YouTube spent 44 minutes less on their phones each day during the restricted period. And those who lost access to apps for two weeks reduced their usage more than those who lost access for just one week.
During the time they were on their phones, users shifted to other apps, Aridor found. Those who lost access to Instagram mostly shifted to other apps they were already using, while those who lost access to YouTube also switched to less prominent and new apps. And, interestingly, both groups—those who lost access to Instagram, a social networking app, and those who lost access to YouTube, an entertainment app—are drawn to other social apps.
While this means that users will adopt a different app category when looking for a replacement, it’s also important to note that actual app activities don’t exactly match the Google Play Store categories. For example, while Instagram is categorized as a social networking app, 37 percent of survey participants said they use it for entertainment.
Digital addiction changes the choice of substitute applications
Aridor also wanted to understand what factors made participants switch to one alternative app over another. Specifically, he wanted to know whether inactivity minimizes how often users look for new, less obvious digital apps as substitutes.
Idleness is a general term that includes users who are “addicted” to certain apps, as well as other factors that can make a particular app a default destination, such as the time users have already invested in selecting, downloading, and configuring it .
Aridor calculated an economic model based on study participants’ past app usage to estimate how users would spend their time during the experiment if inactivity was not a factor.
The non-inertia model showed that users would be more likely to look for new substitute apps, rather than switching to prominent, already used apps. Indeed, the model predicted that without inactivity, the share of time users spent on prominent social media and entertainment apps would be about thirty percent less than what Aridor observed in the real world.
Using experiments to analyze competition
Aridor sees these kinds of “app exclusion” studies as having a lot of potential to improve our understanding of digital markets.
“In the grain market, it’s really hard to do an experiment like this,” says Aridor. “You can’t make Rice Krispies go away and see how people replace it. But with digital shopping, everyone has access to apps through their phone and so can you.”
Such experiments could, for example, give companies a more nuanced view of their competition for public attention. “If I’m an executive on Twitter or Meta, are we really competing with just TikTok or YouTube as well?”
This information would also be useful for policy makers and government regulators. One area of active policy debate is how antitrust regulators can assess whether a merger between two major digital apps could create a monopoly and crowd out competition.
Aridor points out that regulators could perform this kind of blocking study to analyze whether two companies should be allowed to merge, even if their applications are in different categories. If the study showed that users were drawn to App A when App B was closed, that would show that they are indeed competitors, even if they’re not in the same category—something regulators would want to know in order to decide whether to allow the two to integrate.
