"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 Josh Anton, founder and CEO at X-Mode.
The location data industry is poised for vast changes over the next year and decade. Increasing privacy regulation and better understanding of data quality are indicative of the shift to come – one that’s eerily similar to what happened during the subprime mortgage crisis of 2008.
Almost every industry player is going to feel the pinch; some will go under, and many others will spend years trying to get their feet back under them. Those who saw the crisis on the horizon and began pivoting early will not only be best positioned to rebound, but also gain a competitive advantage while others are wringing their hands and wondering what happened.
Which camp would you like to be in?
The looming crisis
Today, companies that build audiences or handle location-based media or measurement have built their models around a combination of three location sources:
Always-on SDK-based location data from GPS, beacons and internet of things. This data is high quality, with 50 or more data points collected per day.
Location aggregators that feature mostly ad-based SDK data, with some data from direct publishers. This data is moderate to low quality, with five to 20 data points collected per day.
Inaccurate, privacy-problematic bidstream data, which comes from running ads on installed mobile apps. This data is of the lowest quality.
Though no formal data quality ratings exist today, I see the combination of these three sources that location companies build their products off of as the data equivalent of a subprime mortgage environment. In subprime mortgages, the location sources were simply replaced by different types of bonds that varied dramatically in quality. When the housing market went bust in 2008, the low-quality bonds essentially disintegrated, leaving too few high-quality bonds to go around. (The movie “The Big Short” does a great job of likening this collapse to a game of Jenga.)
The location data industry is now on the cusp of a major shakeup. Combining great location data with less-than-desirable data from location aggregators and the bidstream has led to a proliferation of location-targeted marketing campaigns that fail to deliver. This in and of itself is unsustainable.
However, the catalyst for the forthcoming data crisis will not be poor quality, but new privacy legislation. In 2020, the California Consumer Privacy Act (CCPA) and other new regulations will threaten the very existence of many of today’s location data players that don’t own the direct source of their location data and, therefore, cannot enforce privacy permissions at the same scale they collect data today.
Garbage in, garbage out
When the data crisis hits, much of the finger-pointing will rightfully be directed at bidstream data. This latitude and longitude data is typically owned by the publisher and can include basic information about the ad unit, publisher, IP address and other elements.
On the privacy side, bidstream is problematic as a foundation for the location-based data market. As a result, many entities that buy and use bidstream data don’t have permission from the end user. Under Europe’s GDPR, that’s already a problem. And with the CCPA and other US regulations on the horizon, the use of this data will become even more problematic.
Furthermore, in terms of accuracy, bidstream data isn’t reliable for measurement or audiences. The horizontal accuracy coming from bidstream data typically doesn’t have sufficient quality to pinpoint a user’s location well enough for proper ad targeting. Also, the location information is only based on the short window of time in which a person is viewing an ad. That person could be on a train, walking home or standing in one place for a long time, but that important behavioral information wouldn’t be apparent based on bidstream data. And because bidstream data typically lacks speed, bearing or altitude information, it’s hard to determine with confidence that someone was at a specific venue.
Quality problems aside, when new privacy legislation comes into effect, bidstream sources of location data will be invalidated, given the inability to sell that data without consent. Meanwhile, mid-tier location aggregators will struggle to gain opt-in consent to collect and sell location data. Their data supplies will be dramatically reduced, and their prices will need to be raised accordingly.
The “A bonds” of the industry will survive, but they will have to grow significantly to accommodate the new demand in the industry, and that will take time.
A necessary recalibration
As a result of these market forces, including the loss of super-cheap bidstream data sources upon which this industry was built, prices for always-on SDK location data will rise dramatically in 2020 and beyond, cutting margins for demand-side vendors.
It will take time to educate the market on the need for higher quality – and thus, more expensive – location data in the wake of privacy legislation, particularly given that more marketers don’t understand the poor quality of existing sources.
Ultimately, privacy legislation will help to legitimize and clean up the location data industry, just as the subprime mortgage crisis led to a widespread recalibration of mortgage lending practices. The transition, however, will not be easy.