A new study has found that a remarkably high percentage of traffic across Index Exchange’s network comes from real humans — a real feather in the cap of an enterprise that has been a leader in the fight against ad fraud.
After several months measuring a representative sample of the hundreds of billions of monthly impressions that are auctioned by Index Exchange, Moat has concluded that a remarkable 98% of Index’s inventory is genuine human traffic. That’s compared to an industry average of around 14% bot traffic across all networks and exchanges.
The statistics come as the result of an initiative by Index to onboard the market’s leading verification providers in order to allow buyers to independently assess the quality of the inventory on its exchange on an ongoing basis.
The idea, Sullivan said, is that clients that use an independent provider for viewability and fraud detection will be able to evaluate Index’s inventory using the same methodology they use to measure their own campaigns.
“Ever since [Index] was founded, we’ve said we’re only going to represent high-quality sources of media,” said vice-president of partner success Steve Sullivan. “As a result, today when we turn on services like Moat, we find that most of the traffic we deal with is human.”
In addition to Moat, Index Exchange will also onboard Integral Ad Science, White Ops, Pixalate and ComScore to provide independent measurement of non-human (NHT) or invalid traffic (IVT), as well as viewability on the exchange. Index says it will publicly release scores from the other invited providers over the coming months.
While that initial score is high, Index didn’t perform quite as well on viewability. Around one third (35%) of its impressions were viewable, falling slightly short of the programmatic marketplace average of around 40%, per IAS’s Q2 marketplace quality report. It’s likely that Index’s low rate of nonhuman traffic had a negative impact on its viewability scores, since non-human impressions often report as viewable or unmeasurable.
Moat’s viewability measurement is favoured by tech platforms like Turn and publishers like Bell Media and Facebook. (It’s also likely used by a lot of agencies, but they’re more reticent to announce vendor partnerships.)
Recently it became the first measurement provider to be accredited for viewability in mobile by the Media Rating Council.
Suppliers face increasing scrutiny
Sullivan, who prior to joining Index worked with ad security firm White Ops, said Moat’s measurement of 2% invalid traffic serves as validation of Index’s reputation for being a zero-tolerance environment for fraud.
“After all these years, human traffic is suddenly quite important, and everyone is scrambling to change the way they operate so they can make sure they’re only selling real human views,” he said. “We haven’t had to do anything different. We’ve been uniquely prepared for this particular shift because of our focus over the years on quality. We’ve found that when you focus on quality media, you end up automatically minimizing the amount of non-human traffic that you’re exposed to.”
Index’s success in eliminating fraud has stemmed as much from its principles as from its technology, he explains. Index vets every seller before allowing them onto its exchange, and has a zero-tolerance policy for suppliers caught auctioning bad inventory.
Despite having drawn the full attention of advertisers and industry trade organizations, ad fraud remains as much a problem for ad buyers as it was a year ago. Major exchanges like AppNexus continue to struggle with the perception that they’ve turned a blind eye and let fraud run rampant in their inventory.
However, the industry may have a solution in the works. The American IAB’s Trustworthy Accountability Group (TAG) — created last fall with the express goal of wiping out fraud — has announced a major new initiative to examine and accredit legitimate programmatic sellers, and stamp them with a “Verified by TAG” certification to let buyers know they aren’t selling fraudulent ads.
TAG is also working with a number of exchanges to develop a Payment ID system that will track payments across networks and resellers to identify the originator of a transaction. By cutting through the layers of anonymity in the programmatic environment, TAG says it will make it much harder for fraudsters to disguise themselves as legitimate sellers.
Index is one of the platforms helping TAG develop the new Payment ID, along with AOL, AppNexus, The Rubicon Project and Google.
“The TAG Registry and upcoming Payment ID system will act like a ‘two-factor authentication’ for the digital ad supply chain,” said TAG CEO Mike Zaneis in an IAB release. “Through the TAG Registry, buyers will be able to ensure that they are working with trusted parties at every step of their campaigns, while the Payment ID system will ensure that payments only go to legitimate players, choking off the cash to criminals.”
A multi-currency world
Although verification providers sometimes offer average scores for exchanges based on impressions their buyer clients have acquired there, it’s unusual for an exchange to pro-actively invite verifiers to measure its inventory.
Index CEO Andrew Casale said his company hopes to lead the market in integrating all the viewability and IVT vendors that agencies now rely on to determine the value of the impressions they buy. “To the big holdco that has signed on with Moat, they now plan from a Moat lens, transact from a Moat lens and evaluate all their exchanges and vendors from a Moat lens,” he said. “If you go to a different holdco, it might not be Moat. It might be IAS, or some other vendor. To us, that was the advent of ad tech moving into a multi-currency world.”
Whereas conventional media markets tend to use a single currency (like Nielsen or Numeris ratings) to determine the value of media, digital media has as many currencies as there are measurement providers. That means there’s a real potential for problematic discrepancies. If, for example, the buyer and seller use different vendors to measure how much of a campaign was viewable, that can create significant friction.
Casale said Index’s plan to smooth out this friction is to bring on all of the major providers, so that it can always determine the value of its inventory using whatever currency the buyer prefers. “If we have a holdco that uses Moat, then when we’re transacting we can make sure it’s always based on Moat. No discrepancies, no friction,” he said. “We don’t know if long-term, does one currency win and take it all over? But for the time being there’s multiple and we want to support them all.”
As a byproduct of the onboarding process, Index was able to compare the viewability and IVT scores that different providers measured for the same sample of inventory. According to Casale, they confirmed that the discrepancy problem is still very real — he said that while most of the providers had similar scores, a handful had “polar opposite” results. “We don’t know what to do with that knowledge at this stage. We may put it into a study and put out research, but that’s still to be determined,” he said.
