DAVID STAAS: We differentiate between location-powered and location-based companies. We think of ourselves as a location-powered mobile advertising platform. For us it’s a distinction in that location-based tends to be very simple geo-targeting types of advertising, such as with geo-fences.
We enable advertisers to connect to their specific customer and find that customer in mobile, which has been an incredibly difficult challenge for agencies and brands to date. The power of location information addresses a range of issues that the mobile ecosystem is dealing with. When we think about location, we think about understanding where people have been, historical patterns and using that info to craft customized audience segments.
We pioneered this approach after we built our Wi-Fi platform nearly 10 years ago. That was the industry’s very first location signal. And because we knew the location of every Wi-Fi hotspot, we realized we could do interesting things with the advertising that we were delivering to those locations. When we finally had scale of signal locations coming in from mobile, we evolved and extended our platform to address all the location data coming in from these mobile devices that we have today.
What kind of data are you collecting?
The data comes from a public Wi-Fi network when somebody is on a smartphone, tablet or laptop. A lot of the places where Wi-Fi is publicly provided are ad-enabled as a way to fund the investment in the Wi-Fi network. When a network is ad-enabled and it’s one of our ad partners, when we get an ad request, we get all kinds of information. There are about 25 different types of information in an ad request. Two things that are key in there are the device identifiers and a very precise latitude and longitude.
Because we are working with a Wi-Fi partner that registers every hot spot with us, we know we have very pure lat/long data that’s accurate. On the mobile side though, there’s a lot more filtering and scrubbing that we do with the data since not all publishers are dealing with true lat/long data. They may have center point data, like the center of a zip code, center of a state, or a country. They may have registration data like a home zip code, which a lot of publishers are doing, and they’re just sending you that person’s registered home zip code as their current location, regardless of where they are.
In addition, we have a tremendous amount of click data, device data, time of day data, etc. and we convert all of that through our data science algorithm to understand how that information correlates. We look at how locations are related to each other, we look at how people are connected to each other, and we look at how people are connected to those location patterns as well. From there we’re able to create audiences.
What are some insights you’ve gathered from the Location Graph?
It’s very detailed data. For example, if you’re looking to assign a sports fan designation, a lot of first generation mobile ad networks will use the content somebody’s browsing. If I’m on a sports app, I must be a sports fan. When you’re dealing with location information, you have all these different lat/longs that are coming in that you need to correlate, so you need a robust database, which also has to be scrubbed to make sure that you are getting accurate information to make sure that’s exactly where the sports stadium is so when you have a lat/long you know that person is there and not at a bar next door.
You also have to look at what’s happening at the stadium. Are they at a college football game or are they at a concert? Do that billions and billions of times and what you find are interesting things around how location data changes the landscape. If you don’t have that machine learning in place where this is all automated, then you’re going to get your audiences wrong.
It’s also one thing to have raw data like lat/longs, but those don’t really mean anything until you understand how it maps to audience data.
For example, we were running a campaign for a major CPG brand that was marketing a lot of breakfast products to moms. They wanted to drive a healthier approach to breakfast foods. We started with a custom audience of moms that we created for the brand, but as we ran the campaign, subsets began to emerge. We found things like gym moms or moms leading very active lifestyles.
We also found that suburban moms engaged more highly than urban moms and by using other levers like how people travel throughout the day, we found that car pool moms versus those that stayed in tighter geographic footprints also engaged significantly with the brand. Those are the kinds of insights that we can provide, because we have the scale of data to find these segments.
What are your thoughts on what Foursquare, Twitter and Google are doing with geo-targeted ads?
I think it’s great that everyone is recognizing that location is the key to mobile. The challenge that Foursquare has is that their data is incomplete and it’s very hard for them to build a viable audience segment. If you think about check-in data, those are the bars and the restaurants, vacation spots—the exciting things. That’s just a small sliver of our lives however. To understand audience, you also have to understand when people are at the dry cleaner or the doctor’s office and all those mundane things.
Foursquare also doesn’t have a lot of data. Three billion or four billion check-ins is very small. When you’re trying to deal with relevant data, you need billions and billions of data points over the last 90 days for it to be effective for advertisers.
What Google is doing with Maps is interesting. They’ve done a really nice job of incorporating ads into local search. It’s a nice play for the local ad ecosystem, but it still doesn’t get into that broader range of audiences. Unless I’m looking for a physical place that’s near me, then how does a brand really begin to engage?
Who are your main competitors?
Right now I’d say folks like Millennial Media. What’s interesting is they’ve been in the market for a long time but they don’t have the same audience targeting capabilities. We’re seeing those dollars shift to us, but Millennial is certainly not letting that happen without a fight.
So what are your thoughts on the Jumptap acquisition?
They’ve got a lot of scale when you put the two together. But neither company has a strong solution set around location data and how do you build that robust audience capability that brands are looking for. The challenge they’re going to have is they have to bring these two platforms together and they also have to evolve those platforms, which are built on a content-based approach, and rebuild their technology to address the current state of the market.
Millennial buying Metaresolver made a lot more sense, with them moving to a more data driven, programmatic approach. I don’t quite see the value of Jumptap so it will be interesting to see how that plays out.
What’s your approach to privacy concerns about data collection?
Through partnerships with TRUSTe we never touch the device identifier. When an ad request comes to us and it has an IDFA or Android ID, etc., it all goes to TRUSTe where they convert it into an anonymous ID for us. The data we have can’t be tied back to an actual device or personally identifiable information. Step two is we also implemented the AdChoices icon across all our Wi-Fi and mobile networks. Consumers can choose to opt out and we’ll delete their data from our records and won’t capture any info from their devices.
Are you seeing an increase in the number of people who are opting out of advertising through the AdChoices icon?
We haven’t done a current analysis of how many people are doing that. I think we’re seeing rates similar to people opting out of traditional display ads, and not much of a difference in mobile.
What can you say around revenue or growth?
We’re a private company but I can say that our mobile revenue is doubling year over year and we’re on track to be profitable this year. Our Wi-Fi service is also growing but we’re seeing the primary growth from the mobile side as more brands shift their dollars there.
One interesting trend is that for CPG brands there weren’t many great mobile solutions for a long time. Geo-fencing didn’t really help since people already knew where to buy their products since they obviously knew where the groceries were. But fast-forward to audience segments and now that we can find specific segments like gym moms, we’re seeing more brands coming into mobile for the first time.
We’re benefiting from this shift of dollars from companies that were previously on the sideline and it’s fueling a lot of our growth. The big win is when you take an audience and layer data over that. You’re not just targeting anybody when they’re nearby anymore and that’s really transforming mobile.