Published Thursday, May 30, 2019
Location data creates a golden opportunity to reach consumers with hyperspecific marketing messages. Knowing where someone is located speaks volumes about what they need, what they want, and where they will go next. If marketers can effectively harness location data, they can operate at the intersection of digital advertising and physical geography.
But research shows that up to two-thirds of this data contains inaccuracies, even among data sets that have been filtered for errors and inconsistencies. This means that over 60 percent of data points are a mile or more off — leading to wildly off-target campaigns.
The success or failure of location data marketing depends on the quality of the underlying data. Marketers must vet vendors carefully to learn who can provide the most accurate data available. Otherwise, they risk sending advertising off into the ether.
Curating data that is both error-free and actionable is a process that should include these four steps:
Many location-based insights are developed using a relatively small sample size. A subset of users opts into having their location data collected. From that, marketers extrapolate behaviors for much broader categories of consumers, which amounts to a guessing game. In practice, the only way to know where consumers go is to track their devices directly on the widest scale possible. In the best cases, location data is collected from more than a billion devices multiple times a day. Hard data, unlike “developed” data, reveals the complex and often contradictory ways people move through the consumer landscape. That way, advertisers can reach them where they actually go instead of where they assume they will be.
It’s a basic scientific principle that the more diversified the data set, the more reliable the resulting analysis. Location data is available from multiple sources, and basing campaigns on just one (or several) sources gives marketers an incomplete picture. Drawing on diversified data sources uncovers more insights overall. More significantly, it allows data to be cross-checked for validity. If two or more sources report a signal at the same coordinate, it’s probably accurate. Therefore, so are the marketing insights that data informs.
We only need to look at ourselves for anecdotal evidence that consumer behaviors are unpredictable and ever-changing. Examining historical location data reveals the cycles and patterns of how people travel, helping marketers predict where they will go next. Looking into several months of historical data reveals some things, but going back a year reveals even more. A full 12 months of data accounts for seasonal trends, weather patterns, annual events, and all other factors that affect buying patterns. This strategy is in line with the previous two: A deeper data set delivers better location data accuracy.
As the stats above indicate, efforts to filter out inaccurate location data are often ineffective. Filters are proprietary products, and some vendors have vastly more development capabilities than others. A stringent development process begins by discarding at least half of all reported locations. The remaining half is subjected to exacting quality control measures based on years of iterative improvements. During that process, data is sent through multiple layers of filtering to remove all kinds of inaccuracies. Effective proprietary filters deliver marketers the most complete, current, and correct insights available. Armed with that information, it’s relatively easy for marketers to maximize return on advertising spend.
Location data can either be an incredible asset or a complete distraction. Publishers are trying to tilt things in their favor by flooding the ecosystem with bad location data. If marketers are going to leverage this resource, they need to be proactive and innovative about cleaning it up.
To learn more about how Valassis Digital achieves location accuracy with better prediction and stronger results, check out this report