Ad buyers often point to digital media’s richer datasets to explain why more of their dollars are going online. But a new study is casting doubt about just how accurate the data used by media buyers really is. The analysis by Truthset in partnership with The Coalition for Innovative Media Measurement shows only half (51%) of the digital targeting and audience measurement data is accurate.
Truthset examined 15 different data providers, analyzing the linkages between 3.9 billion anonymous email records and 90% of known U.S. postal addresses. Truthset then validated the information against the U.S. Census Bureau, Pew Research Center, and a variety of other independent, self-reported, and declared data from more than 20 million records. The findings were presented at the CIMM Summit, the annual gathering of media researchers.
The study was designed to explore the opportunities for marketers to future-proof their use of privacy-safe identity data and to improve accuracy and reduce marketing waste at a time when cookies and mobile ad identifiers are being phased out. Truthset – which was founded in 2019 by industry veterans from Nielsen, Salesforce, LiveRamp, and Procter & Gamble – says the results shine a light on the accuracy of the data records that marketers increasingly rely on for advertising and customer engagement. The report’s authors say the results also underline both the challenges and opportunities present in today’s data and identity marketplace.
“At a time when advertising and marketing budgets are shifting, this study proves that more accurate data can help drive results – financially and otherwise,” said Truthset CEO Scott McKinley, a former Nielsen executive.
Truthset did not identify the winners and losers in its analysis. It noted, however, that even those data providers who did not score very high did have some accurate information, so none were labeled complete misses by the researchers. One firm that had an accuracy rate of just 44% reportedly had a wide range of success, with a quarter of its data as good as 80% to 90% accurate. But because email and postal addresses are often the foundation of other demographic identification used for targeting, the error rate is reason for media researchers to be concerned.
“Big data is going to play an important role in the future of the advertising ecosystem, but right now, the industry is focused solely on scale,” McKinley said. “For the first time ever, we have a common tool that can be used by the industry to assess the accuracy of this data and to provide a path forward to deliver better audience targeting, increased CPMs for publishers, as well as more accurate measurement that can help elevate consumer experiences.”
The analysis says inaccurate data costs marketers billions of dollars each year. Truthset found that removing the inaccurate datasets and utilizing accurate audience data could significantly increase on-target percentages and return on investment. For example, the report shares an analysis of a typical CPG campaign, demonstrating how a CPG advertiser could improve its returns from $1.08 to $1.54 per dollar invested, simply by improving the quality of the identity data and the match rates used to support the campaign.
One of the biggest stumbling blocks for data providers is keeping up to date, with those that had the worst performance usually relying on older information. The analysis also came to the conclusion that the size of the dataset was less a factor in determining its accuracy.
“Consumers change their email addresses pretty regularly, and we need robust practices for ensuring that identity data is accurate and up to date,” said Jon Watts, Managing Director at CIMM.
One of the potential fixes identified for the database accuracy shortcoming would be to link metadata to email and postal addresses, which researcher say would help remove outdated information from a database.