As Programmatic Continues Meteoric Rise, Location Accuracy Continues To Drop
Thinknear’s Q1 report on the state of location data shows that accuracy has slowly declined over the last three quarters.
The ability to accurately serve a location-based ad to a consumer hasn’t improved since Q1 of 2014, according to a recent report by hyperlocal ad network Thinknear.
While mobile advertising undoubtedly became more sophisticated over the past year, location data accuracy is still a problem for marketers, inhibiting effective targeting.
In the fourth edition of its Location Score Index — which is based on “sampled and analyzed data from over a billion ad impressions and location accuracy tests on more than 500,000 consumer ad experiences” — Thinknear posits that as the programmatic space continues to grow quickly, the quality of location data is subsequently declining, returning to levels last seen a year ago.
It is well documented that as location ads have risen in popularity, targeting accuracy has suffered. While the industry has been flooded with location data, only some of it is of high quality; as we wrote in May 2014, a host factors can lower data quality.
For example, in cities with many tall buildings like New York, interference with cell phone location signals is common. More problematic are apps that ask for user registration and a zip code when a person downloads the app; that zip code is then centered around the user’s home or work, and thus there is no way to find that person as they go about their day.
In order to meet the growing demand for impressions with location data, many publishers have substituted less-accurate data gleaned from the aforementioned zip code downloads and the like, rather than relying on real-time data derived from GPS and wi-fi positioning.
It will take a true commitment from mobile advertisers and data providers alike to reverse the trend, but by identifying the issue and relying solely on first-party geo-data from the most reliable sources, accuracy stands a chance to climb in the coming quarters. In the mean time, below are key statistics from the report detailing the state of location data accuracy at present:
- Hyperlocal Accuracy: 37 percent of total location data was accurate to within 100 meters of a user’s true, real-time location. This figure is the same as Q4 2014.
- Local Accuracy: Nine percent of location data was accurate between 100 and 1000 meters. This is down one percent from Q4 2014.
- Inventory Accuracy:By these figures, it is clear that 48 percent of data was accurate by 1000 meters or less, and the remaining 52 percent — the majority — of inventory was off by more than 1000 meters. 10 percent of the total data was inaccurate even by 100,000 meters — or more than 60 miles