Study: Mobile’s Data Problem Is Getting Worse

my_locationAs machines take over the advertising industry, the behavioral data that publishers pass on to marketers has evolved into something like a currency. The value of advertising inventory, once determined by subjective estimations of audience, is now calculated in an instant based on the demand for various bits of information that publishers can provide to marketers about the consumer behind a device.

In the mobile advertising industry, where data is harder to come by, location data has emerged as a pricey asset. But a new report suggests that location data on major exchanges, which often can cause inventory to sell for three to four times as much, has decreased in quality over the past three months as publishers move more inventory to programmatic environments.

The report, released by Thinknear, the mobile advertising subsidiary of mapping giant Telenav, found that of the 68% of mobile ad impressions which include a latitude-longitude, only 37% of those coordinates are accurate to 100 meters (roughly the size of a football field). That’s down nearly 10 percentage points from the fall. Meanwhile, the number of mobile ad requests that served inaccurate location data to targets between 100 and 100,000 meters increased by 12% over the same period.

The study finds that only a minority of ad impressions with location data are accurate to less than 100 meters.

The report is the third in a series of studies the firm conducted in conjunction with Location Score, a technology which Thinknear built to predict the accuracy of location data included with ad requests.  The product, and many of the insights from the report, stem from a model the company built over the past year by buying millions of mobile ad impressions and asking those users to share their location separately. The company then compared the location information the publisher included in the ad request with the user’s real location, and identified the discrepancy.

Loren Hillberg, general manager at Thinknear, says the rise of inaccurate data is likely due to a spike in traffic from mobile websites where determining a user’s location can be more difficult than it often is in the confines of a mobile application.

LDS14_logo_cascding“Over the last 12 months, the available inventory for potential location-aware advertising has probably doubled or tripled, and yet during that time the quality of that data hasn’t changed dramatically,” said Hillberg. “We attribute some of the drop to the overall growth from the mobile web, where location is a bit more challenging to collect from a developer’s perspective.”

Hillberg, however, dismissed earlier suggestions by the company that the rash of inaccurate data was a product of systemic fraud in the mobile advertising industry. Some in the mobile advertising technology industry have worried that mobile publishers were intentionally using less precise location information, such as a zip code, as a stand-in for more precise, and valuable, latitude and longitude data.

“The overall data suggest that there is not obviously systemic fraudulent behavior, although there are specific cases where you see publishers trying to mask one set of data as another,” said Hillberg. “When you look at the question [of fraud], I think it’s much more about developers being ignorant of what they can do to improve the data they pass along.”

For Thinknear and other mobile advertising firms focused on location, the data quality issue offers both a threat and opportunity. On the surface, a rash of poor data reduces the effectiveness of location-based targeting, undermining the value proposition they offer marketers. The study found that inventory with vetted location data outperformed run-of-network data by nearly two times in certain non-click conversion metrics

At the same time, the data issue creates a need for specialization within the mobile advertising industry — something that should help solidify the role of more focused companies like Thinknear. The rise of programmatic exchanges displaced the early mobile-local advertising networks that build businesses aggregating publishers who offered location-enabled inventory. Without control of supply, mobile-local specialists were forced to compete solely as demand-side platforms assisting with targeting, attribution and other capabilities.

Steven Jacobs is Street Fight’s deputy editor.

Find out more about how to address this key problem in local targeting at the Local Data Summit in Denver on March 5th. Get key insights and strategies you can bring back to your company and use. Don’t miss out! Click here for more info.

  1. Rick Waghorn
    February 4, 2015

    All very interesting for those of us not smart enough to have a programmatic sales interface to throw at this marketplace.

    The rise and rise of programmatic is interesting in the sense of where it all leads.

    If I was a big ad tech house – or the VC backers – I wouldn’t go reading J Tainter any time soon.

    He might suggest that for every extra layer of complexity that programmatic adds – muscling in on local/location – so they take one step nearer the point of systematic collapse.

    The cost of hiring in the best mathematical minds to drive programmatic on is another example of a tribute too far in the mind of a Tainter.

    Hope springs, however.

    For when complex societies collapse humans tend to start again simply. From the bottom up. And locally.

  2. March 12, 2015

    Surprised there’s no mention of the decrease in amount of location data available based on the change in iOS8 to require distinct background location vs foreground location permission requests.

Leave a Reply

Your email address will not be published. Required fields are marked *

Next Post

How Bad Local Data Leads to Bad Calls — And Frustrated Customers