“Data-Driven Thinking" is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Eli Portnoy, general manager at Thinknear.
If you’ve ever been served an ad for a Los Angeles restaurant while visiting the San Diego Zoo, you’ve seen how imprecise location-targeted ads can be. As a consumer, this sort of misstep can be annoying, but as an ad buyer, it’s an infuriating waste of money.
Like anything else, digital ad inventory is subject to the laws of supply and demand. As demand for location-specific ads has increased over the past year, so too has the value and, consequently, the price of the inventory. This has led to a sharp increase in the supply of location-enabled ads, or so it would seem.
The location data that is sold to advertisers, typically at a premium over standard inventory, are often inaccurate at best — and dishonest at worst. Many publishers simply approximate user location based on a number of imprecise factors, some of which having nothing at all to do with physical location. Very few actually deliver what they promise.
Latitude/Longitude Does Not Equal Precision
A common answer offered by ad networks when discussing the accuracy of their location targeting is that they only use lat/long location data. Lat/long refers to coordinates on a map, usually implying a single square meter. Unfortunately, the fact that a lat/long refers to a specific point on a map doesn’t mean the intended user is actually there.
Lat/long is the convention by which publishers, exchanges and ad networks pass location information between them. Unfortunately that means that regardless of the source of the location data or how imprecise it is, the ad network will ultimately be assigned a lat/long. So whether the publisher thinks a person is within a 10-mile or a 5-meter radius, the lat/long will imply that person is in a very specific and precise single meter that could be off by more than 10 miles.
Just Say No To Centroids
Centroids stand apart from the inaccurate measures I mentioned above because there is a modicum of location data that is applied to the practice. As the name implies, centroids are highly inaccurate location readings that translate lat/long coordinates in the center of a geographical area, such as a city or ZIP code. The problem is that these center points are generally miles away from where the user actually is located. The result: an expensive scenario in which entirely too many people are served ads that do not actually apply to them.
Centroids operate under the assumption that anyone near this geographic center is interested in the same thing. They ignore the nuances of neighborhoods as well as the physical locations of landmarks, shops and attractions. While they might eliminate the problem of serving an ad to someone hundreds or thousands of miles away, it doesn’t mean the ads they do serve will be in any way targeted.
Question The Results
It’s easy for media buyers to take campaign reports at face value without questioning the results. But for location campaigns, buyers should really push for insights and location data that matter to clients. If an agency is just looking at CTR, it’s letting its vendor off easy. Look at how performance, engagement and conversions vary by location or how specific audiences engage with ads by location. That kind of performance can only be measured and shared when the data is accurate.
There’s no doubt that accurate location targeting will increase ad engagement. Those engagement figures will continue to rise. But unless the industry works to ensure that targeting and measurement are both exact, media buyers still won’t get what they’ve paid for.