Dentsu Aegis Network has spent the past year furiously integrating with publishers so advertisers can target consumers using PII-based identifiers.
These identifiers let Dentsu clients target more precisely, consolidate reach and frequency management around a single ID and optimize marketing spend. Advertisers get the benefits of walled gardens without the wall.
The program, known as Publisher Addressable Marketplace (PAM) by publishers and M1 by advertisers, had three beta clients during Q1 of 2017 and now has 30 advertisers on board.
Meanwhile, publishers participating in PAM have tripled from last year and now include big names like Pandora, AOL, Time Inc., Conde Nast, CBS Interactive and Weather. That growth means advertisers can now buy across more publishers than they could just one year ago.
Dentsu expects several hundred million of its clients’ media dollars will flow through the platform this year, attracted by the promise of efficiency.
“We are seeing 50% lift or more across the board,” said John Lee, president of M1. “We see underexposure and underreach at shocking levels, and overfrequency of certain individuals at disturbing levels, when we compare the old world to the new M1 world.”
But the program is still getting off the ground, in part because it’s tricky bringing on publishers by promising advertiser spend. Merkle first announced M1 in September 2015 and PAM in June 2016, before Dentsu Aegis Network acquired Merkle and decided to continue the program and put its advertisers’ buying power behind it.
The challenge: M1 needed to create a scaled marketplace with enough supply, and high enough match rates, to offer an attractive alternative to walled gardens.
Publishers would only join if the dollars committed matched the technical heavy lifting required to enable the PII-based matching and buying. And because the program uses PII, Dentsu Aegis and the publishers it brought aboard needed to navigate privacy and data security issues.
While Dentsu Aegis Network may have a slight lead in building out a proprietary buying and identity platform, it isn’t the only agency building out proprietary pipes to win clients and prove value.
Other agencies have the same idea. GroupM’s mPlatform started making noise in 2016. IPG Mediabrands’ AMP uses an Acxiom-created universal ID for its buying platform. Publicis’ Spine focuses on a Publicis-generated ID.
The big differentiator for M1 and PAM is proprietary tech integrations with publishers.
While previous agency networks and trading desks struck deals with publishers to arbitrage inventory, Dentsu Aegis Network said M1 does not mark up inventory. Nor do its clients pay a tech tax, since the setup allows Dentsu Aegis clients to cut out ad tech. Instead, they get access to clean, vetted supply where they can find real people.
Here’s how Dentsu Aegis Network is getting the program off the ground.
Solving For Scale
“Is there scale? Absolutely,” Lee said. “That wasn’t the case a year ago.”
PAM publishers created a deduplicated reach of 225 million US consumers from matching hashed email addresses or login information to the M1 identifier. This process created a connection similar to Facebook Custom Audiences. To improve match rates and reach further, the program will also employ “tier-two” matches, such as cookies, that resolve back to an M1 ID.
For instance, Pandora matches with the M1 ID using hashed email addresses but recently added scale by linking device IDs.
“It took us a while to get the match to a point we were happy with,” said Dave Smith, SVP of monetization and yield at Pandora. But now those match rates fall within industry averages. By the end of 2017, 70-75% of Pandora’s audience matched with the agency’s data sets, he said.
From there, the strength of the match depends on the strength of the Dentsu Aegis client’s data. Dentsu Aegis said the match rate with Pandora now averages upward of 90%.
To do the matching, Dentsu Aegis and publishers in the program often employ a creative hodgepodge of technology, which could include outside tech vendors, header bidding or custom integrations. Dentsu Aegis and the publishers involved wouldn’t share specifics of how they sync up the M1 ID with a publisher’s inventory.
Attracting Publishers Amid Fears Of Privacy And Data Leakage
Afraid that Dentsu Aegis will target a publisher’s audience elsewhere?
“All our contracts and tech [are] designed so we cannot use data on other publishers that’s not for the benefit of the publisher,” said Charlie Weiss, GM and SVP of publisher solutions at Dentsu Aegis Network. “We would never reprofile a user and run [that audience] in another environment, because it would dissolve the trust we had with them.”
But concerns about data leakage and privacy initially slowed publisher adoption. “It took a long time from the legal side to get this done,” said Jeremy Hlavacek, head of global automated monetization at Watson Advertising, which oversees Weather Co. assets.
And Dentsu Aegis insisted the program is fully GDPR compliant.
Despite these fears, PAM publishers can access the buying power of Dentsu Aegis Network clients – just as WPP’s Xaxis appealed to publishers that accessed the buying power of GroupM advertisers.
“[This program] is a next-generation PMP infrastructure,” noted Matt Prohaska, co-founder and principal of Prohaska Consulting. “And I’ll bet their average deal size is larger than $82, which is the stereotypical frustration of publishers.”
Also, publishers like valuing their inventory not just by content quality, but data quality, said Weiss. That’s a stark contrast from when he started at the end of 2016.
“They are viewing [data] as a way to differentiate their inventory and see it as a growth lever,” he said.
That shift has made it easier to onboard publishers. In 2018, PAM hopes to onboard connected TV and OTT inventory.
Chopping Off The Long Tail
Meanwhile, buyers using M1 can access a network with the type of reach and frequency tracking available within walled gardens, but without the brand safety concerns.
“We are out on the bleeding edge of the trend that already exists, which is clients shying away from the long tail and the opaqueness in understanding true audience frequency,” Lee said. “M1 and PAM are really just the realization of that trend.”
And PAM publishers can scoop up budgets from advertisers fleeing the long tail.
“This setup plays to our strengths in terms of being a large, scalable publisher, and having trusted and safe environments,” said Judith Hammerman, SVP of data commercialization and programmatic sales solutions for Time Inc., which joined PAM late last year.
And, the setup cuts out ad tech middlemen with their extra fees.
Perhaps most importantly for Dentsu Aegis, an agency-based identifier to access unique inventory differentiates the agency amid fierce competition and downward pressure on budgets.
According to Lee, M1 is “at the heart of the data story that Dentsu Aegis Network is telling,” making it responsible for “some very big successes” in 2017 agency pitches and reviews.