Advertisement
Adtech

Does Mobile Marketing Actually Work in the Real World?

Wading through the murkiness of location data

Ad networks are gung-ho but marketers are wary Photo: Getty Images

When you walk into the mobile advertising firm xAd’s headquarters in New York, you'll notice a pair of digital screens showing maps of the United States. On the left, a map plots out more than 150 billion monthly pings from smartphone owners opening advertising-supported mobile apps and sites across the country. Meanwhile, the map on the right is zoomed in to show a small sliver of midtown Manhattan. Each time someone clicks to refresh a piece of mobile content, a dot lights up showing their exact location—down to the closest street corner.

A list of locations runs alongside the screens, pulling in data on smartphone-toting users' location. At 9 a.m. on a Thursday, the list sluggishly pulls in information from hotels and commuter hubs; by noon, it will flip through data collected from quick-service restaurants and retailers.

Several unnamed retailers are reportedly using this beta map product—dubbed Footprints—to home in on Black Friday traffic this year. The idea is to help marketers visualize where and how consumers use their mobile devices based on a latitude-longitude coordinate.

Is Mobile’s Contextual Promise Paying Off?
The new hyper-targeted ad format is the latest move from the mobile advertising industry to capitalize on the growing interest in location-based marketing, which BIA/Kelsey estimates will bring in $4.5 billion this year. By 2018, per BIA/Kelsey, that number will grow to $15.7 billion.

But while the tactic seems intrusive, you may be surprised to learn that xAd throws out 80 percent of inventory from publishers because the location data isn't accurate. Similar numbers from mobile firm ThinkNear back up xAd's claim—a mere 34 percent of ad requests based on latitude and longitude data are accurate within 100 meters of a consumer's location. 

"[With] one of the big players in this space, we saw 85 percent of all impressions [in a campaign] served were not within the location that we specified for our client," noted Jon Hook, head of mobile at Mediacom International. "I'm not talking about minor discrepancies—these are significant discrepancies that we're seeing."

Ironically, Hook said that the campaign was deemed a success, indicating that brands may be getting more ripped off from location-based advertising than they realize.

Part of the problem is that the number of location-based ads available to marketers has grown significantly over the past year as big players like Facebook, Twitter and Foursquare try to inch in on the lucrative market.

With that though comes more skepticism on where networks get their data from and which type is best for individual campaigns. First-party data (information collected directly by the ad network) is preferred for advertisers since it doesn’t pass through another companies’ hands. But, third-party data (which is supplied by another company) is more readily available, helping campaigns scale.

Michael Lieberman, CEO of Joule U.S., echoed Hook's concerns about not knowing where data comes from, even with first-party data that ad networks claim is the cleanest type.

"I don’t have a way to validate it," Lieberman said. "I don’t see the lat-long coming through necessarily, I don’t know how it was generated—I’m just told by the network that they targeted someone based on GPS data in that area."

The fogginess with mobile data has prompted more than 50 vendors, agencies and brands (xAd and Joule among them) to work with the Mobile Marketing Association to create a set of standards that are expected to come out in the first half of 2015. As mobile budgets grow, the goal is to help brands make educated ad buys.

"What we’re doing now is setting about to establish some strict definitions of location data and what the associated use cases are," said Leo Scullin, head of global industry initiatives at the MMA.

Continue to next page →

Advertisement
Advertisement
Adweek Blog Network