GroupM Will Be Watching Where Its Clients' Ads Run, And Where They Should Not

John Montgomery

In the biggest push yet by a major Madison Avenue player to gain some control over the sometimes capricious way that their ads run online, WPP's GroupM unit has cut deals with two of the leading online ad verification services, and will integrate their systems as part of the tool chest that WPP's agencies use to plan, buy and post their online display ads. The deals with AdSafe and DoubleVerify follow an intensive review of the major ad verification services, which included a month-long test involving half a billion impressions of online advertising inventory.

"There's always been the issue of are we getting what we asked for," John Montgomery, COO of GroupM Interaction, one of the largest buyers of online media, tells Online Media Daily. "Just think of the staggering number of impressions we are delivered -- there is no way of knowing that we are getting what we paid for."



Montgomery, who was part of the GroupM team that rewrote WPP's terms and conditions of online advertising buys last year to gain more accountability for its clients' ads and media buys, said the verification systems pick up where the T&Cs left off. (OMD Feb. 2, 2009)

"We have some lengthy language in the T&Cs 2.0 that cover the terms of [ad] adjacencies that require publishers to deliver us what we were promised, but we had no way of dealing with compliance," he said -- noting that while some publishers and networks have their own internal technology for verifying where, when and what GroupM client ads ran next to, the vast majority of online advertising schedules still go largely unchecked.

Montgomery said this is especially a problem with so-called "tier two" inventory -- the kind acquired through online ad networks, exchanges and other intermediaries -- where ads may be parsed to lesser-known sites and destinations delivering the right audience profiles for a client's campaign, but whose content may be questionable at best.

"We started getting calls from clients and we started getting calls from consumers saying, 'Hey, a reputable brand is appearing adjacent to inappropriate content. A legitimate ad is appearing on a porn site," Montgomery said, leading GroupM to pursue a more systematic method for screening and ensuring that its ads ran as scheduled.

While the current generation of online ad verification systems generally post where buys have been placed after the fact, Montgomery said some of the newer applications being developed by AdSafe, DoubleVerify, and others now alert agencies and marketers when their ads are about to be placed adjacent to unsavory content, including porn, hate speech, or content that competes or conflicts with a brand's message.

But there are other problems beyond content, he said, including that the ad schedules delivered other key components of the campaigns' objectives, such as proximity. The magnitude of geographic campaign discrepancies, Montgomery said, was revealed by the test that GroupM conducted of leading ad verification systems during the month of December. That test, covering 500 million online ad impressions for a major GroupM consumer brand, was conducted simultaneously on AdSafe, DoubleVerify, and at least one other undisclosed online ad verification system that GroupM is not yet using.

"We did find that there are some areas of concern," he said. "For instance, impressions falling outside of the geographic area that we specifically bought. We saw some very big variances. Variances as much as 40%."

Montgomery said those margins are significant for several reasons. One, obviously, is that they fall out of the geographic requirements of the insertion order, and may be wasted impressions in terms of a brand's trading area. But he said there could also be some negative aspects associated with a brand losing control of where their ads are being distributed geographically.

"If you think about a pharmaceutical company, there's the potential for a huge amount of risk, because sometimes the brand name doesn't even exist outside of the borders of the U.S."

But the test helped verify the magnitude of geographic dislocation of online advertising buys where a publisher defines its reach based on where its server is located and not necessarily on where its users are coming from.

"Sometimes the definition of a U.S. campaign is whether someone clicked on a site," he explained. "If someone from Russia or China clicked on a U.S., that's defined by them as a U.S. impression."

Montgomery said GroupM will use the verification systems to audit and post the delivery of its online advertising schedules, but also hoped it would help GroupM agencies work more closely with publishers to help prevent such lapses and discrepancies in the future.

He said GroupM ultimately opted to go with two services for now, because both AdSafe and DoubleVerify had certain advantages and weaknesses that make a tool test approach a more logical way for GroupM agencies to go than picking a sole provider.

"I think it's healthy governance for us to work with two suppliers," he said, adding: "It's always good from a performance and pricing point of view to have two companies rather than rely on one."

3 comments about "GroupM Will Be Watching Where Its Clients' Ads Run, And Where They Should Not ".
Check to receive email when comments are posted.
  1. Mike Einstein from the Brothers Einstein, March 17, 2010 at 1 p.m.

    Sounds like a lot of billable makework for the agency. Which, of course, is the whole point.

  2. David Shor from Prove, March 20, 2010 at 7:09 p.m.


    Yes, that's true but even more interesting is the fact that the real benefit is knowing which audience segment within the traffic streams are working for them. Tools like TrackSimple and Legolas Media do similar work (on the audience side), and Adometry does a good job of DoubleVerify-like activity without all the hype.

  3. John Grono from GAP Research, April 12, 2010 at 3 a.m.

    Don't tell me that for the past decade clients have been paying for online ads that weren't served to the geo-target or were served on inappropriate sites!?! Who would have thunk it ...

Next story loading loading..