Marrying passive and custom data for effective mobile targeting

Getting to why

Editor's note: Kevin King is senior vice president of mobile product management and innovation at GfK MRI, New York. Based in Boston, Natasha Stevens is senior vice president of digital market intelligence at GfK.

It has been the complaint of countless marketers and researchers over the years: “If only I knew what consumers are really doing, not just what they say they are doing.”

Fast-forward 10, 20 or 50 years – to today. We seem to be living in the golden future that so many past researchers dreamed of, when transactional and behavioral data for millions of consumers is readily available. Marketers have copious information about their own customers and a variety of syndicated panels and studies track purchases, ad views, Web site visits and other activities.

But in the midst of this information glut, billions of digital advertising dollars are being spent based on the most rudimentary targeting information. Many marketers are all but guessing as they allocate among thousands of Web sites, social media platforms and other digital properties using data that is not always validated.

If John Wanamaker could see how little information digital advertisers have to go on, he might count himself well-informed.

This inability to translate oceans of information into streams of practical guidance was already a problem on laptops and PCs – the world of the “seated Internet.” In this universe, dropping cookies allows publishers to trace consumer activities from site to site; but what we know about each user is often only inferred. Those who visit ESPN.com and a videogame site in different online sessions might be young males – but how much are you willing to bet on it? (A million or two?)

Mobile has taken these targeting challenges to new levels; the expanding universe of mobile carriers, phone makers, app developers and other mobile data players makes data collection and integration difficult. And cookies cannot follow consumers from browser to app; each app is its own microclimate – and data owners tend to guard their user information jealously.

For advertisers, the result is a sense of true frustration. They know they need to be running campaigns on mobile devices; in GfK MRI’s recent Mobile Now study, almost two-thirds of consumers said that the first thing they do in the morning is check their mobile phone – and mobile use is only going to continue growing. But optimizing spend against specific targets across apps and mobile sites is all but impossible. So advertisers make do with whatever they have; mobile has become their biggest blind spot.

Can we do better?

How did we get to this point, when so much information yields so little real-world guidance? And can we do better? For smart decision-making, marketers need three basic types of information about consumers: who they are (demographics), what they do (“behaviorgraphics”) and why they do it (psychographics). Passive data lends itself to the second need, simply by revealing where people go, how long they stay and what they do there – whether “there” is a store, a Web site or a movie theater. (But do not imagine that passive data collection is anywhere near straightforward, especially in the mobile world.)

Knowing the “who” comes from connections to syndicated databases, which extrapolate general identifying data from sources like GfK MRI’s Survey of the American Consumer.

To get to the why requires more subtlety and direct interaction – asking people what made them visit a certain site or buy a particular TV set or bar of soap. We cannot afford this level of inquiry on a mass scale; but if we can identify the right people to ask, we will be golden.

And, as with most research, the Holy Grail is single-source data – the ability to ask custom questions of the same people whose online habits you already know.

But the mobile environment is uniquely hostile to investigation – a closed world, by design and, to some degree, pure circumstance. With each new version of the Android and Apple operating systems, compatibilities shift and the rules change; they are the ultimate moving targets. This is the essence of the “walled garden” that vexes mobile researchers. As MediaPost has reported, Verizon, AT&T and Sprint are shrouding their user data, which limits what brands can do with device IDs; and this is just one example of the problems posed by having so many different entities owning different pieces of the mobile data puzzle.

Ad-blocking, which has become a major headache of the desktop environment, is now also coming to mobile. One carrier started blocking ads on every mobile browser, with the intention of turning ad sales into its own revenue stream. Carriers have huge power to restrict – and to measure – mobile data; there is no law governing mobile, no “net neutrality” to fall back on.

Unique set of issues

A unique set of respondent issues also awaits those who would dare to take on mobile research. Consumers often see mobile devices as more personal, and therefore more private, creating new reluctance to allow intrusions. Compensation is a worry – what to pay for this unusual access. And mobile device-monitoring technologies are still in their infancy, subject to a host of technical issues, such as battery draining, OS interference and more.

But for those who do choose to invest in measuring mobile, the payoffs can indeed be lucrative. It is a matter of how you approach the problem – which tools and partners – and the internal talent you bring to the equation.

To target more effectively, marketers and advertisers need deeper information about consumers. Today, many profiles are based on location – zip code and DMA profiling, along with proximity to restaurants, stores and other potential brand touchpoints. Understanding where people are, though, does not tell us who they are or why they are there.

Many feel that the real game in mobile advertising is apps, which are optimized for mobile devices (according to Akamai, just 12 percent of the top 100 Web sites are mobile-friendly). And while app owners may have a host of data about what features get the most visits in their applications, they can only guess about what their users do elsewhere. Also, without a detailed database about users – addresses, phone numbers – apps can provide marketers with surprisingly little detail on the who element of our ad effectiveness equation.

So how can marketers get access to the detailed contextual insights they need to target more accurately in mobile? Passive measurement of mobile use on a known respondent group is a good starting point. More and more, marketers are looking for accurate information to make smart decisions, which means they are questioning the reliability of recall data. When it comes to smartphone and PC use, people often cannot recall what they did the last hour, let alone the last month. Marketers want a clearer picture of how people are engaging with the mobile Web versus apps, how this usage is changing over time and how it might be different for competitor sites. For an e-commerce company specializing in pet food, for example, it is not enough to know what percentage of mobile Web users go to Amazon or other competitive sellers; they need to understand activity for the pages and products in their categories.

To help fill this gap, measurement companies can install technology that (under ideal circumstances) captures all types of activity on the phone, from Web browsing to app visits. But this work is not for amateurs; the technological requirements of Apple’s iOS and the Android OS make even user-authorized information capture very difficult. And more practical challenges may also arise, such as noticeable effects on battery life, which is a sure turnoff for respondents.

For KnowledgePanel Digital, GfK has adopted a battery-efficient alternative that still records all online-based activity, through a browser or app. With this passive technique, plus background data from respondents’ participation in panel profile surveys, we aim to provide a robust look at mobile Web users.

Learnings from other sources

But passive data is not always enough to answer a marketer’s most pressing questions. A bigger insight and effectiveness payoff can come when passively-collected data is supplemented by learnings from other sources. These can be syndicated databases or custom surveys, depending on the need and the rarity of respondents for a given study. Together, they can help you answer questions such as:

• What mobile apps are heavy buyers of my product category using, and how?
• Are my competition’s customers accessing content on their smartphones and tablets? When, and for how long?
• Which social media platforms do people turn to on their mobile devices before or after researching a product in my category?

Many of these questions center around mobile’s role in purchase journeys. We know from our research that 48 percent of online shopping now takes place on a smartphone or tablet. At the same time, increasing proportions of marketing and ad budgets are being allocated to mobile. So how important are specific apps or mobile sites in driving all sorts of purchases – from cleaning products to TVs to cars? What specific experiences, likes and dislikes and other influencing factors are resulting in purchase decisions?

To get at these questions sometimes requires more specific inquiry – and here is where surveys can play a role. By leaning into the strengths of direct questioning, and not relying on interviews to capture mundane information like hours of use and sites visited, we can maximize the value of the survey technique. And being able to direct questions to specific respondents, based on passively recorded behavior, means marketers can learn more deeply about their most important buyers and potential buyers.

A recent GfK study of grocery retail in the digital world showed, for example, that consumers over age 60 over-index for use of grocery digital sites and apps. While overall usage of grocery digital properties is lower than other retail categories, it did drive higher engagement than other retail categories, with more repeat visits within a four-week period than home improvement or drug store digital properties.

The new hub

Mobile is the new hub of online activity and commerce; so mobile’s role in marketing and advertising is no longer about early adopters, it is a mainstream play. But the insights to fuel better mobile investments are lagging (some would say severely). Passive data collection holds the promise of behavior-based ad targeting; and while experience will solve some problems, the “walled” nature of the mobile environment is limiting the potential of ad effectiveness, and therefore campaigns.

To stay ahead of these challenges, smart marketers and advertisers are combining resources, linking to high-quality offline data sources and fielding surveys that zero in on essential questions – the whys that are so crucial to successful marketing. The result is a level of accuracy and accountability that can give mobile its due in ad intelligence and, ultimately, investments.