Location data is buzzing at Mobile World Congress in Barcelona.
“Where you go is who you are,” said Daniel Rosen, global director of advertising at Telefónica, speaking at an MWC Mobile Media Summit (MMS) satellite event. “[It’s] hugely powerful.”
That’s coming from a carrier with its own mobile exchange, Axonix, an audience buying platform fueled by first-party Telefónica data, including location, demographics, sites visited and in-app behavior.
“Operators have access to reams and reams of very accurate independent data streams in the same way that Facebook has login data,” said Alex Newman, head of mobile for EMEA at OMD International.
Telecoms are clearly interested in getting into the ad biz, with Verizon’s acquisition of AOL as the shining example. More recently, Norwegian telco Telenor shelled out for cross-device company Tapad.
Not everyone can get their hands on first-party data signals, though, and low-quality location data is an ongoing issue. According to 2015 research from Thinknear, more than half of location data appended to bids on the open exchange is inaccurate, at least for targeting purposes, and one in 10 location-targeted ads is not accurate to within 60 miles.
But in-app location data is a different beast from location data on the open exchange, Roger Woods, Adobe’s director of mobile product and strategy, told AdExchanger.
“Apps have direct communication with the operating system, so they can get more accurate location data,” he said. “One of the problems with the exchanges is that they’re pooling data and sharing from multiple sources, and don’t necessarily have control over how all of that data is being collected.”
In any case, industry bodies like the Mobile Marketing Association and the Interactive Advertising Bureau are plugging away at recommendations and advocating for location data transparency standards.
That should help, but it doesn’t change the fact that location is not a static data point. As Mike Schneider, VP of marketing at location services company Skyhook Wireless, put it to AdExchanger in a previous interview: “Nobody stands still.”
What’s accurate now isn’t necessarily accurate, or at least not as relevant, later.
“All of this data is useful – if it’s in real time,” said Fabio Esposito, eBay’s global display programmatic lead, at the MMS Summit. “Location is changing every minute, interests are changing every minute. … What’s going on with [a] specific person at that moment in time?”
But the argument could be made that in-the-moment targeting isn’t necessarily the most compelling use case for location data.
“The more mature way of looking at location is around driving deeper engagement and loyalty, not location-based targeting,” Woods said.
Some marketers are using location as part of the service layer, a means to enrich an audience profile rather than yet another way to hit users with outbound messaging.
Take British home goods retailer Argos, which allows consumers to use its app to reserve items in-store for up to two days with no down payment. Argos was looking to get more people to actually pick up the merchandise they had reserved, so it worked with Adobe to send friendly notifications based on a user’s location reminding them to claim their stuff or offering them the option to retrieve the item or items from another closer, more convenient location. After implementing that line of messaging, Argos saw in-store pickups increase.
There are other signs that the market is maturing. On Tuesday at MWC, xAd launched what it’s calling MarketPlace, a self-serve advertising platform that allows users to buy and sell location in the same manner as they might do with keywords through Google. KFC is a launch partner.
Say there’s a home goods store and a KFC in the same mall. Advertisers could use the MarketPlace platform to target shoppers in that store right before lunch with an offer to get them to walk in the door. That works because there’s an “immediacy of need,” said xAd founder Dipanshu Sharma.
On the flip side, advertisers can use location to inform a user’s profile over time. From frequency of visits – say three or four KFC visits within a month – a brand can infer loyalty.
It really just depends on the use case, Sharma said.
“Fast food is one thing,” he said. “But you’re not going to buy a car because got a message while you were standing near a dealership.”
Which is why looking at patterns around location could have value.
Boingo Wireless, for example, through its advertising arm Boingo Media, partners with advertisers to sponsor Wi-Fi in public locations like airports, malls, subways and stadiums. Boingo uses deep packet inspection to see the types of websites users visit and where they are when they visit them. If Boingo notices a lot of people consuming similar web content in a particular region, it can use that information to pitch relevant brands on advertising in that geographical area.
The same philosophy could be applied to creating audience.
“We look at how often we see people and where,” said Derek Peterson, CTO of Boingo Wireless. “Have you connected to Wi-Fi more than once? In that case, maybe you could be part of our ‘Wi-Fi Warrior’ segment. Have I seen you at one concert or multiple concerts? At one stadium or multiple sporting events? Once we start to answer those questions, we can bring that data together with data from our DMP relationships and third-party data like age and interests.”
OMD’s Newman, for one, is ready to embrace location data as a “method to understand people and their behavior.”
“We spend a lot of time talking to clients about location data,” he said.
For plenty of others, both in Barcelona and elsewhere on Earth, location is interesting, but it’s not the be-all, end-all. Sean Huang, Quantcast’s global head of mobile, is one of them.
“There are a lot of important signals to look at when you’re trying to figure out if a customer is interested in a product or receptive to advertising or if now is really that right time,” Huang told AdExchanger at MWC. “The value proposition of mobile is that you carry it with you and there’s a potential wealth of location data if you can get to it – but it’s just one signal we evaluate.”