Over the past year, GroupM media agency Essence has reduced its investment in third-party data segments for certain clients by more than half.
While the decline has been most prominent in Europe, it began well before GDPR took effect and is totally independent of the regulation, said Ryan Storrar, Essence’s head of media activation in EMEA.
“Ultimately, we use data segments that we think will provide the greatest return,” he said, “and we have alternatives which are increasingly effective.”
Those alternatives are available through platforms like Google, Facebook and Amazon, which offer segments based on their rich user data baked natively into their platforms – and often at a lower cost than third-party providers.
“These segments prove to be effective both in terms of scale and cost efficiency,” Storrar said. “They have a very good deterministic data, a spotless match rate and high scale.”
When transacting with premium publishers, Essence has also found it’s more effective and efficient to target using their contextual data instead of third-party segments.
“If you’re looking to add scale to your campaigns, premium inventory providers can be more effective than going down a modeled data route,” Storrar said.
While third-party data is still widely used in media buying, Essence began reducing its reliance on the tactic as part of its philosophy to test the ROI of every tactic used on a given media plan. Essence found that most third-party data providers model their segments to the point where they’re watered down and ineffective.
“Some third-party lists have been useful to certain campaigns,” he said, “but generally, we’ve found question marks over the accuracy and scale of third-party segments and the ROI we get by buying them.”
Essence was able to reduce its reliance on third-party data in part because, with Google as its largest client, it’s a heavy user of Google Marketing Platform (formerly DoubleClick). If clients don’t want to buy through Google, Essence will only use third-party segments available natively in DSPs after testing their accuracy and ability to yield ROI.
“Outside of DBM or in the absence of first-party data, third-party data remains a go-to tactic,” Storrar said. “But we’d measure the impact of third-party data segments relative to what we can do with contextual targeting.”
Measuring the accuracy of third-party data is difficult, especially in Europe, where the market is more fragmented than in the US. But if marketers are paying for third-party data, they should at least understand how effective it is, Storrar said.
“When data has a cost, it’s worth measuring that cost relative to value,” he said.
Third-party data still has a role in programmatic, and it powers much of the ad tech ecosystem. But there’s also junk out there that marketers should be aware of and make sure they’re not overpaying for, Storrar said.
“I'd encourage vendors and marketers to place greater focus on scrutinizing the accuracy and effectiveness of third-party data segments,” he said. “The industry would benefit from a clean-up.”