Why Cross-Device Matching Will Transform the Mobile Advertising Industry
The mobile advertising industry is somewhere in its adolescence. Startups have turned into public companies, and many of the major technical problems have worked themselves out. Soon, the sector will face one of the realities of maturity: integrating with the rest of the digital advertising world.
One of the drivers leading that convergence is cross-device matching where a company can identify multiple devices owned by a single person. Companies like Drawbridge, which has raised a little over $20 million, have built massive databases of user devices that allow advertisers to reach the same user as they move from their laptop to smartphone to tablet.
We caught up with Rahul Bafna, VP of product at Drawbridge, to talk about the breaking down of the wall between desktop and mobile and what that means for the way marketers think about reach local consumers.
Cross-device capabilities allow us to think about desktop and mobile inventory as tools to drive consumer behavior — not ends in themselves. What do you think about the role of desktop ads vs mobile ads in shaping the consumer journey?
At this point mobile advertising and desktop advertising are roughly the same in terms of overall spend levels and that holds in both display and even for search. What any advertiser wants to do is reach consumers, not necessarily devices or consumers on specific devices. Like any other channel desktop or mobile just gives advertisers different opportunities to reach consumers and the creative format needs to be different to better engage users. You need to consider the context the user is in.
When a user is on a mobile device they could be checking their phone for 30 seconds and probably not be in a good environment to try and engage them in an ad, but they could be consuming some longer form content and that could be a better opportunity to place an ad.
The big first application of cross-device linking has been retargeting: the ability to message one consumer on two different devices. But how can we begin to mesh the data available online — namely, browsing behavior from cookies — with the mobile data — namely, location.
The data gathered from mobile, whether it be app usage, location, or anything else, remains separated from data collected on desktop. The only way to “mesh” those together is by establishing a common identity for a single user across multiple devices. We can analyze vast amounts of data and our proprietary algorithms, to probabilistically bridge different devices back to a single, anonymous user.
We can do this at large scale (3.5B devices, 1.2B consumers) and accuracy. A marketer can then tie mobile and desktop usage together to gain a better understanding of consumers. This can be used for retargeting across devices, but also for better targeting of ads using an advertiser’s first or third party data. The identity solution can also be used to measure cross-device conversions and offline (store visit) conversions.
Drawbridge recently announced partnerships with LiveRamp and Datalogix — two offline purchase data repositories. What impact does this purchase data have on the way marketers buy media?
There’s lot of research around digital purchases versus offline purchases. For all the hype about digital advertising whether it’s online or mobile, conversions don’t always occur on those devices, because even as late as 2014, only 94 percent of sales still occur offline.
Two or three years ago, advertisers only had same device conversion capabilities — if an ad is shown on a desktop, gets purchased on a desktop, then that purchase can be attributed back to that ad. Drawbridge rolled out a cross device conversion report a couple years ago, and a couple other companies have something similar, where if an ad is shown on a phone, but the purchase happened on desktop owned by that same user we can tie those two events together. That still covers only about 6 percent of purchases.
That purchase data is still not one-to-one. What other data signals are you looking for to solve the offline attribution problem.
There are two different types of signals used to measure the lift from a campaign: One is actual purchases, so companies like Datalogix that has access to point of sale data which they can link back to a user cookie so we can track that an ad shown to a user within a certain time period result in a purchase at a store. Of course that does require the store being able to share that data back with one of those companies like Datalogix.
The second solution is around measuring location-based attribution — showing an ad to a user on desktop or mobile and then seeing if the user, along with his or her mobile device, walked into a store and was able to visit that location (you can’t tell if they made a purchase). For certain types of stores, for certain verticals, that makes a lot of sense, like a movie theater or a restaurant. For certain verticals with very lengthy consideration cycles like automotive, that’s a pretty strong signal of intent. As the number of users and mobile devices have grown and multiple location capabilities have grown, there are a handful of companies that have come up in that space that provide those capabilities.
Today, there’s a lot of specialization, with companies focused on either on desktop or mobile. Do you think cross-device technology will reduce the need for this specialization? Will it force a consolidation between the two sectors?
I agree there is a lot of specialization there. There are some companies that started in the desktop world and struggled to build equivalent capabilities on the mobile side. Companies that just started in mobile have much better capabilities.
We are at that inflection point where users are spending about half their time on mobile now and half on desktop. If you want to reach those consumers effectively you really do have to be strong in both. But there’s a difference between being able to manage segments of users or targeting data on users versus being able to execute on it.
What you can do, if you are a media execution platform, then yes you need to be able to buy effectively on both mobile and desktop. If not, then you could license targeting data and a cross device solution. But there’s more to it than just knowing who that user is and what device they are using it on, especially in the context of a performance campaign. You need to be able to basically effectively manage media budgets across both those platforms.
Liz Taurasi is a contributor to Street Fight.