“Attribution Solution Under Foot” – featuring Jason Uechi
It’s a problem practically as old as advertising itself: how to measure the effect of advertising spend on real conversions and revenue. And while the rise of digital advertising has multiplied the available points of measure, closing the loop in a world where 90% of purchases are still made offline can leave marketers struggling to quantify an abstract currency, such as a click through rate, against real world dollars.
Today there is also much discussion about the power of mobile: consumers spend more time online with their mobile device than at their computer. The smartphone is perhaps the most personal piece of technology ever invented. For many, your phone is the last thing you see before you fall asleep, and the first when you wake up, and serves as your portal to everyone you know and all the knowledge of the world. As a communication device the smartphone has only gotten richer: first sharing audio only, then text, then photos, now full motion video in real time.
This “always on” aspect of the smartphone equates to data — lots of data. In the mobile application ecosystem, and in the real-time programmatic advertising architecture that surrounds it, phones provide bits of information about the user (albeit pseudonymously, following industry-standard privacy practices) in order to improve the value of ad placement. These billions of pieces of data can be measured, matched, and combined through algorithms to form a persona that best matches requirements set by the advertiser. Buying and placing an ad based on this persona happens in a fraction of a second, from the time you open an application on your phone to when the little banner appears along the bottom.
One of the most common pieces of data that gets passed along in this daisy chain of information is the device’s location. GPS and location data power many of the most useful features of the phone from driving directions and maps to making recommendations for the closest restaurant or gas station. The device’s location is incredibly useful to make sure that the ideal persona and the right device gets an ad or offer only when that consumer is within a reasonable distance to the advertiser’s location.
But smartphone location data can also serve as the vital link from digital advertising to real world attribution. While the process is complex, the concept is incredibly simple and can be captured in a single question: for each consumer to whom we sent an ad, did that consumer subsequently visit the store?
Who can answer that question? Location-based mobile platforms are best positioned at the crossroads of these data streams, and those with the best data infrastructure and scale can provide it today. At YP Mobile Labs, we call it a Store Visit Report and we use this data with our advertisers to measure the effect of a mobile advertising spend on real conversions and in-store revenue.
The reporting starts off by defining a very clear geographic boundary for each store. We are then able to measure any observations of consumers within those boundaries as a visit based on the location data from their smartphone. Others approach this by different means: by surveying a panel, by tracking background location from a specific app, by relying on beacons, etc. While each method has its own strengths and weaknesses, we prefer the scale available in using billions of points of GPS data.
For each of those observations, we can comb our campaign data to answer a few key questions: “Was this visit by a user who was part of our target audience?”, “Did this user receive an ad impression, then subsequently go to the store?”, “If so, did the user click an ad before going to the store?”
The answers to these questions have proven to be incredibly interesting and quite valuable for our customers. They can help to capture and quantify the value of a marketing campaign by correlating the marketing received by individuals on their mobile device to their subsequent offline behavior. The value of the findings are perhaps most powerful when this attribution can be tied back to known in-store metrics, like average conversion rate and average purchase size, to be able to calculate a real return on investment for your mobile marketing spend. Being able to show a line, or to at least draw a clear arc, from mobile marketing spend to cash register receipts is a powerful story to tell.
The next logical step, of course, after capturing in-store lift due to mobile marketing, is to work for continuous improvement in subsequent campaigns. In adopting a disciplined approach to monitor and optimize towards better performance, advertisers will find that the standard variables apply: from tweaks to creative and messaging, to the value of the offer, to the types of media units and landing page enhancements — all are areas of opportunity.
We’ve all talked about the promise of big data, as opportunities abound for advertising in our always on, mobile state. The next big leap as an industry is perhaps making amends for the over promises of the initial online advertising revolution that left us with many clicks, more data, but perhaps not more meaning. The next step in attribution turns out to be mobile, and quite literally, it was under foot the entire time.
*Jason Uechi is Director of Engineering and Data Science at YP Mobile Labs.