Demographics behind apps reveal so much more than numbers

April 11, 2016 / by / 0 Comment
  • SumoMe

For many of us, associating things with people may be an unconscious process, but definitely not uncommon: newspapers to fathers, glasses to smart people, cooking to housewives, dark suits to professionals, or crowns to royals—the list goes on. The coefficients vary but this is how our mind works. We assign an image, number, color, or sound to someone, even to a group of people, say age, nationality, or gender. It may fall under the umbrella of Phantasia, or, as Aristotle once said, the human being’s capacity to produce images out or perceptual appearance.  

In the context of mobile applications, such things exist as well. One could simply guess a phone’s real owner by simply looking at installed apps as clues. Having Neko Atsume installed says the person could be a gamer who loves cats. A person with Tinder could be single, and installing no social networking or chat apps may mean the person is an introvert. A Bloomberg app can suggest that an investor or a finance fiend could be its real owner.

Two experts have proven that this is possible, and they did a research on it. In a study conducted by Eric Malmi of Verto Analytics and Aalto University (Finland), together with the Ingmar Weber of Qatar Computing Research Institute, it was discovered that the apps installed on a mobile phone is enough to determine its real owner’s age, gender, income, and, rather surprisingly, marital status.

For this study, Malmi and Weber consulted 3,760 mobile phones installed with various types of applications, with a total of 8,840 apps. Only phones with apps that have been opened at least once in the last thirty days could be part of the study. Apps with less than ten users have been excluded from the study to discard all personally identifiable information.

The results are pretty amazing.

For instance, the study revealed that if the phone has an ESPN app on it, there’s an 85 percent likelihood that its owner is male. Men between the ages of 33 and above are 73 percent likely to have Clash of Clans on their phone. Ninety-one (91) percent and 78 percent of unmarried men in the same age bracket also own Grindr and Tinder, respectively, which refute the unsung notion that dating apps are mostly used by women.

On the other hand, it said that if a phone has Etsy and Photo Grid on it then it could be owned by a woman. If one has Yelp, there is 61 percent possibility that it’s owned by a person whose annual income is above $50 thousand, and if it has SoundCloud the owner’s annual salary could be below $40 thousand. Married people also tend to depend highly on real estate and rental apps, while professionals would be the type who would install LinkedIn and Fitbit on their phones.

Great discoveries, which, according to Malmi and Weber, can help brands and future app developers to identify their target audience. This, in turn, will help them achieve a higher likelihood of being downloaded once it gets live on the app market as they will be encouraged to include demographics-focused features on their design.

“Understanding the demographics of app users is crucial, for example, for app developers, who wish to target their advertisements more effectively. Our work addresses this need by studying the predictability of user demographics based on the list of a user’s apps which is readily available to many app developers,” Malmi and Weber explained in the research.

In a supplementary article, Caitlin Dewey of The Washington Post created a quiz that lets people see if the research is really capable of determining one’s mobile phone user’s gender, marital status, or income by simply looking at its applications. It’s just a 32-item YES or NO questionnaire on whether you have the specified app on your phone.

“Of the four traits, the researchers studied, gender was the easiest to determine, and income was most difficult. But when you add in the other sorts of personal data that many apps collect — think location, contacts, and phone usage — it’s easy to see how targeted ads and other personalized features can feel so creepily accurate,” wrote Dewey.

Indeed, the researchers do not guarantee sure-fire success by solely depending on their study, especially that its accuracy falls only at 61-82 percent. “Several interesting questions are left for future work. First, we note that demographic attributes are most likely not independent, and therefore, predicting the attributes simultaneously, employing multi-label prediction techniques, could improve the performance.”

Dom Einhorn, CEO of Born2Invest, a fast-rising multilingual app for business and finance news, agrees: “Of course, no one could accurately determine one’s gender, marital status, or what-have-you by just looking at the apps on their phone, as personality and behavior vary depending on the user. Other factors like, say, preferring Facebook for browser to its app version also count. But it’s a perfect introductory map for those who are planning to build their own app—it can be a guide to developing a new app in an existing niche or a blueprint for doing what we’ve done—which is developing something ambitious, hasn’t existed before.”

Just last month, App Annie announced that a new set of metrics has been added to its platform to help developers, marketers, and brands understand the app market more. The platform, which is utilized by established and startup developers across the globe, is a top tool for tracking mobile applications performance and understanding the ever-changing app milieu.

Mark Ungerer, the general manager of App Annie’s Enterprise Products, suggested that a large percentage of developers and entrepreneurs are now lost with their target audience’s needs, highlighting the importance of consulting to science-supported studies on the app market and consumer behavior.

“Given how advanced and global mobile has become, it is often surprising how little app developers, investors and brand managers know about the users they seek. How do different demographic segments engage with my brand and my competitors? What are users doing outside of my app? Today’s launch provides these answers and more, bringing a complete understanding of the customer to mobile,” he added.

In other words, developers and brands would be able to position their future product on the market by simply knowing what types of app that specific consumers put on their smartphone. Would they market a SoundCloud-like app to a married CEO? Probably a bad idea, but the research could also help these app innovators determine what hasn’t been introduced on the market. Thus: Pinterest for married men or CoC for women? Why not?



Tech Enthusiast, Not a science geek but interested in all things tech can live a day without pizza but not without my smartphone