By Jeff White
In a recent post, I outlined how consumer commitment to a particular intent or interest can have a big impact on the reliability and robustness of targeted audience segments. Advertisers benefit from knowing whether a consumer segment represents “diehard fans” or “casual observers.” Today, I am going to break down how we derive that commitment level from event attendance, and the steps we take to ensure that you can trust our data.
Data Fidelity is the Foundation of Trust
How do we know for sure that our data is reliable? When you deal in data that includes location information, there are occasional spurious signals that spoof location and render the data unreliable. Other signals seemingly place a consumer at a location, but don’t necessarily confirm a visit to that location. To build reliable consumer audience segments, based in part on location visits, we need to get these two elements right. Then, to associate consumer devices with event attendances we need two additional elements.
You have a couple of choices as a data vendor. You can assume that most location signals are legitimate and keep them all, or you can throw out anything that doesn’t fit a verifiable pattern. Gravy has chosen the latter approach. It’s a more difficult proposition and poses the risk of discarding some valid data, but it means that our customers can trust the results. Similarly, if you want to add context around event attendance, you need to verify that the details of the event are accurate. By ensuring our location signals and event details are valid, we also ensure that our downstream event attendances and behavioral consumer segments are valid.
The Seven Elements of “Verified”
At Gravy, we often use the term “verified.” We use this for both verified location (i.e. we are sure a device was at a location) and verified attendance (i.e. we are sure a device was at a place at the time a specific event occurred). There are three data elements we use to verify that a device has visited a location, and an additional three data elements we use to verify the details of an event. We then cross-reference these data points to verify that an event attendance occurred.
How to Ensure Location Accuracy
You can see from the diagram above that location accuracy is essential to verifying our data. Once we’ve captured a unique device identifier, we need to assure that the device was at a particular location or venue. This is a common issue faced by companies collecting location data, and there are two key challenges involved.
First, we need to verify that the signal location is reliable. We want to be confident that the geolocation coordinates reflect the actual device location, and that any conflicting information is reconciled or discarded. For example, what do you do when you see a device in two disparate locations over a short time period? Or, what do you do when you see a large number of signals at a location in the woods near a cell tower – at all times day? We have algorithms in place to identify these anomalies and either discard the data or verify its fidelity.
The second aspect of location information is spatial relationship fidelity in the real world. We need to have a good understanding of the physical world to know if someone was standing on the street or in the adjacent restaurant or concert hall. Many location providers have pin or grid-based systems that assign device locations to general areas. An entire area may be classified as a shopping or entertainment district, providing no visibility into consumer engagement. There is no ability to differentiate between which shop or theater was attended, or tell if the consumer was simply walking down the street. In contrast, Gravy uses high precision polygonal geolocation mapping that lets us interpret a geolocation signal as an actual place, and not just an area.
How to Ensure Event and Attendance Accuracy
Event accuracy represents a different type of challenge. We start by collecting event data from a large number of sources – all using different data formats – and normalizing it. Then we programmatically address duplicate event listings and conflicting event details. We also match each event time with its known location, or place. Finally, we associate the street address for that place with the polygonal area representing the boundaries of its location.
This event data then comes together with verified location signals to verify an event attendance. Gravy’s industry leading AdmitOneTM local event attendance solution enables us to verify actual event attendance at specific locations. Gravy uses these data points to verify commitment by consumers to particular intents and interests in their real lives. This data provides more robust insight into consumer segments and is the basis for Gravy’s TruLifeTM Audiences. Marketers can use these audience segments for consumer insights or to reach highly-engaged consumers through targeted advertising campaigns.
Verifying What People Do Is Complicated, but Worth It
There is a lot going on behind the scenes when you consider what’s involved in verifying an event attendance. In fact, we’ve only just hit on the highlights. There are many more nuances that come together to form our complete solution. However, it’s worth the trouble to better understand who consumers are in their real lives, as it enables us to predict their interests more effectively and segment them more accurately.
At Gravy, we believe that what people do is the best indicator of who they are. People don’t attend events by accident: They do so intentionally, expending extra time, effort and money. This intentionality and sacrifice are the best indicators of consumer commitment to something important to them. For this reason, committed consumers should be an important part of every marketer’s customer targeting. Gravy can help you get started.