Apps and advertising: Now marketers can invest more wisely

Research by Professor of Information Systems Sang-Pil Han provides a model for mobile application analytics that helps media buyers know how much and where to spend for optimal ad return on investment.

“If you are a parent with teenagers, you may be worried seeing your kids spend hours glued to their smartphones and mobile devices,” says Sang-Pil Han, professor of information systems. But, if you are an app developer or advertiser, you might want to capitalize on such app engagement.

That’s why Han created a way of analyzing how long people spend using various app categories and, based on that factor, how to calculate which apps might be a marketer’s best bets. The result of Han’s research is a model for mobile app analytics that could help media buyers make wiser choices with in-app advertising buys.

Small screen, big bucks

In September 2016, researchers at emarketer.com predicted, “Mobile ad spending will grow 45 percent this year to reach $45.95 billion. By 2019, mobile will represent more than a third of total media ad spending in the U.S.”

Given how much time we spend on mobile devices, it’s easy to see why. According to the analytics firm Flurry, people in the U.S. now spend five hours a day on mobile devices. Flurry tracked some 940,000 apps across 2.1 billion devices to come up with that figure. And that’s just a portion of the app ecosystem. As of March 2017, Android users could download some 2.8 million apps, and Apple Store buyers could choose among 2.2 million.

Adding to the complexity of the app-sphere are these realities. First, every click from a text, transaction, or command becomes a data point that advertisers could examine. Second, just because someone uses an app, it doesn’t mean he or she will use it long enough to notice the advertising. Finally, every user is different.

“We were fascinated by the entire spectrum of time-use distribution,” says Han. “Among thousands of subjects, we found some people spend more than 10 hours a day using their mobile devices, whereas we also found a small segment that spent less than 10 minutes a day on their devices.”

To ferret out what keeps users engaged with apps, Han and his fellow researchers evaluated the apps based on two constructs: baseline utility and satiation.

If you use The Washington Post app several times each day but rarely click on The New York Times app, Han and the team surmised that you’re getting more baseline utility from The Post app than The Times app. But, he continues, “We measure utility in two dimensions. One is frequency, and the other is time-consumption.”

So, for instance, if you visit The Post six times a day for 10 minutes each but visit The Times once for an hour, you’re actually demonstrating that you get the same utility from the two.

The researchers also used satiation as a value in their analytics model. “Satiation is how quickly or slowly you get bored or done using a particular app,” Han explains. “When using banking apps, most people are quickly satiated. They spend two minutes checking a balance or making a transfer and then they are done. But when they open a gaming app like Candy Crush or Angry Birds, they typically spend more than two minutes. They may spend 20 minutes or an hour.”

How did Han and his colleagues track utility and satiation? The same way media has been measured for decades: with a little help from Nielsen.

The Nielsen Company has been tracking media usage among consumers since 1950, and its famous Nielsen TV ratings were initially based on self-reported viewing habits. By the 1980s, Nielsen largely relied on a set-top metering device that electronically tracked what TV channels people were watching.

“Nielson now has a panel of people who downloaded a specialized metering application onto their mobile devices,” Han explains. “It monitors which app a person visited and how many seconds he or she stayed in that particular app. Then the time-use and the time-staying information are transmitted to a program on the Nielson side, and that is the data we had access to.”

Reach and frequency

Anyone that took a course in advertising might remember the term “reach and frequency,” which means how many people see or hear an ad and how often those people see or hear it. Han and his team applied a similar concept to large categories of apps, and they used the same classification system used by app markets like Google Play.

Under communication, you’d find things like Skype and Snapchat, while a tools category might include alarm clocks, flashlights, and unit converters. The team also looked at map and navigation apps, web search, social media, photo apps, banking apps, entertainment, gaming, and more.

Communication, scheduling, and tool apps get used extensively by many different people, Han says, and those people tend to linger to a degree. Less popular categories include music, entertainment, social media, and gaming apps, and those who use them use them for long periods of time. As noted earlier, banking apps get used and closed quickly, so Han says they have high satiation levels. Games, on the other hand, stay open and in use for hours, which means satiation is low.

For media buyers, Han’s research illuminates where marketers should allocate their advertising dollars based on utility and satiation, but those answers depend on marketing goals.

“Let’s say a media buyer wants her online advertising pitch to appear 10 times for a given user or audience in 30 minutes. If this is the case, she may want to spend money on an app category where the satiation level is relatively low,” Han says.

But, if that same media buyer wants to maximize ad exposure to the general population, she may want to buy apps with high utility even if they aren’t apps people use for a long time, he notes. “Depending on targets, the notions of satiation and baseline utility can be handy,” he adds.

Passing it on

Another thing Han covered in his research was interdependence of apps. Han invites the reader to imagine hiking in a beautiful area, snapping a photo, and wanting to share that picture with friends or family.

You would use your camera app, then maybe touch up the composition with a photo app and, ultimately, pass it along using a social app like Facebook or a communication app like WhatsApp.

“Say you want to expose your ad on communication apps because that’s where target customers spend most of their time, but sometimes the price tag of such apps may be prohibitively high,” Han says. “As an alternative, you may spend your advertising budget on a very closely related app.”

In this particular case, photo networks or photo apps would work, because after the user visits one app, he or she is very likely to visit a related app at the same time. An ad buyer can allocate their scarce monetary resources to reflect such progression of user activity within the mobile environment.

What Han and his team didn’t do with this research was track utility and satiation by demographic groups. But, advertisers can buy such data from Nielsen, and “they can look at how much time the consumer segment of their interest spends on particular types of apps,” Han says.

In fact, market researchers can crunch their own numbers using Han’s model, as it’s available in a methodology paper he recently published. The paper also includes a section on mobile media planning with a simple optimization framework. “You put in your own input, your budget, and how many times you want your ad exposed in a given population,” Han explains. Using his model, media planners can then run the numbers to know how much and where to spend it for optimal ad-investment return on investment.