Even within the rapidly growing world of technology, mobile gaming is impressively popular, posting awe-inspiring numbers in terms of growth and revenues. Mobile games have further demonstrated their value by achieving industry-leading on-boarding and conversion numbers.

However, mobile games are facing revenue limitations because of how little their developers know about their users. To solve this problem, developers are looking to the same source that allowed them to achieve their already impressive growth: smartphone technology.

The numbers and possibilities

As of 2013, 1.2 billion people played mobile games. While that figure is extraordinary on its own, it’s even more staggering in relative terms. More people use their mobile devices to play games than go on Facebook or any other social network (1.1 billion), and mobile gamers compose almost all of the smartphone market (1.3 billion). As far as apps are concerned, games produce 90 percent of annual revenue for Google Play and 74 percent of the App Store’s annual revenue. Moreover, in 2015, mobile games are expected to generate more revenue than console games for the first time.

Beyond being able to house sophisticated interfaces, smartphones are capable of providing tremendous insights about the people who use them. By discovering who their users really are, machine learning technology can allow mobile game developers to unlock their full market potential. In the short term, these advancements will generate immediate changes in advertising, which are then likely to transform into personalized in-game experiences that adapt to the traits, or even personas, of people who play mobile games.

Smart phones, dumb system

While some games do learn more about their users via registration or integration with Facebook Connect, their ability to make meaningful inferences about their users is narrow.  Users are fundamentally reluctant to register accurate personal data, and coverage is often a major obstacle. Moreover, many game developers forego registration to prevent the churn caused by long on-boarding funnels and privacy concerns. Consequently, game developers usually base their relationship with their users on profiling based on in-app activity, and highly technical parameters provided by Android such as: device type, operating system version and geo-location. However, vastly different people in the massive mobile gaming market often fall within the same buckets, based on superficial and simplistic data points. A 20-year-old female student and her 65-year-old businessman grandfather playing the same game on similar devices receive the exact same experience: the same version of Candy Crush Saga and the same advertisements for Coca-Cola (as cookies, used for ad targeting on the web, are not working on mobile apps)

The end result of the current system generates poor results for everyone involved: Developers receive less ad revenue, advertisers fail to interact with their market and users are faced with irrelevant, and more importantly — needlessly frequent, ads. However, a new approach to data management, and accompanying advances in technology, can take mobile game advertisements to a new level.

Building a persona implicitly through big data

An array of research has demonstrated that an individual’s interaction with their phone can provide insights into their personality: what they like, what they dislike, what they do and when they do it. Surprisingly, this sort of data can be inferred without accessing private identifiable data, but rather by analyzing how a device is used. Machine learning technology, already in place in sophisticated platforms, can explore the different ways users interact with their games, their phones and other apps to help get a sense of users’ broader interests. By listening to the story each phone has to tell about a user, developers and advertisers can make/achieve deeper insights into what users want to see/are likely to engage with and when. Industry professionals have also found ways to find insights from smartphone data, even from seemingly trivial information such as Wi-Fi connectivity patterns and battery drainage. By leveraging big data and machine learning algorithms, users can be identified within a set of personas—user profiles built on behavioral analyses, which incorporate demographic data such as age, income and interests.

Any development of user-based profiles naturally raises immediate privacy concerns.  In fact, maintaining user privacy is a key concern, if not top priority, when building mobile, or browser-based, targeting platforms. Companies looking to maintain a high moral standard should ensure that their solution complies with a strict set of guidelines and rules to ensure the protection of user privacy. These guidelines include complete transparency with end-users regarding the fact that targeting activity is taking place, and allowing the users to easily opt-out of said services should they desire. Furthermore, targeting technologies should refrain from collecting and storing any Personally Identifiable Information (PII) such as personal user IDs, physical device IDs etc. Finally, technologies should strictly comply with laws like COPPA, which regulate data collection on children and refrain from profiling sensitive personas related to religion, sexual preference, political leanings and such. Even with the risk considered, the use of personal data can bring a number of benefits to users.

Broad benefits

Persona-based targeted ads will be relevant to users and much less invasive, in comparison to current advertising methods. Recent user backlash against ads has been so severe that the Interactive Advertising Bureau recently issued an apology. As much as users might find bad ads frustrating, companies also stand to lose by creating these negative interactions. However, advertisements that are built into gameplay, and presented based on a thorough analysis of a user’s behavior and interests, will be more relevant, and less frequent, than the advertisements in place today, ultimately creating a better user experience.

David Noy is the chief commercial officer at Adience, a data-science company focused on providing the mobile industry with critical audience insights to enable a user-centric, personalized mobile experience.