GroupM's Strategy To Abandon RTB Is Safe, But Is It Sound?

There’s been significant discussion around GroupM’s decision to forego open ad exchanges, opting instead for deal IDs and private marketplace deals with “premium” publishers. 

A couple of key reasons/concerns for going the direct route include viewability and the presence of bots. These two separate types of plaques are having a negative impact on the ad exchange business -- no question about that -- however, there are solutions in the market to mitigate the percentage of “bad” campaign impressions in open ad exchanges. As many know, the IAB has already gone through the painstaking process of defining viewability as well as approving auditors via the 3MS initiative.

In the past few weeks, multiple bot and fraud schemes have been uncovered. In that vein, companies such as WhiteOps, Doubleverify and C3 Metrics continue to build businesses designed to combat fraud and/or identify which media partners which are adding value. Yes these ancillary tools might take additional resources to implement into a client’s respective ad tech stack, but I’d wager a number of clients (particularly direct-response) would be comfortable with these incremental ad tech costs in lieu of paying 2x or 3x the cost of media for a private exchange or ID deals which rely on potentially arbitrary, (dare I say “old school”) negotiating techniques to determine value on a given block of inventory.



Further, there are DSP and SSP players in market which support RTB in private marketplace and deal ID campaign scenarios. The fact that GroupM appears to have foregone any/all RTB activities is a bit of a head scratcher unless they are getting extremely preferred pricing from publishers on behalf of their clients (at which point wouldn’t these publishers be better served allocating inventory to the open auctions?). The opposite is supposedly occurring, where publishers are getting a premium compared to what they would on the open market -- a win-win situation supposedly, but I’m trying to figure out for whom. 

At a minimum, there is operational efficiency gained via automation through private marketplace and deal IDs. Going a step further, perhaps GroupM is able to determine if they want a particular impression from a block of inventory (i.e. accept 90% of what a publisher passes through a deal ID). My next question if I were a client would be: "Is each impression being individually valued and priced accordingly?" If not, this assumes all impressions are of equal value, not unique “snowflakes.” Many an RTB pundit would argue that geo, audience, device, time of day and a multitude of other variables should be evaluated when purchasing an impression -- even if it’s sourced from a premium publisher.

If GroupM is indeed using some form of RTB behind the scenes in their private marketplace deals then the argument for setting up direct deals becomes much more compelling. However, if fixed-rate private deals are flowing based on pre-negotiated terms, then I’d argue there is room for further client ROI.

Additionally, the private exchange methodology is also an interesting way to dissuade clients from the “in-house” movement which has been en vogue the past few quarters. A brand is more likely to try handling alliances with 10-20 ad exchanges versus 100 or perhaps 1,000 publisher relationships (the latter set-up may even sound like the beloved ad network model).

Understandably some clients and agencies may continue to wait for the white knights mentioned above to further clean up the open ad exchange ecosystem. However, in the interim if there are opportunities to leverage RTB technology and its efficiencies in private exchange or even deal ID settings -- why wouldn’t you?

10 comments about "GroupM's Strategy To Abandon RTB Is Safe, But Is It Sound?".
Check to receive email when comments are posted.
  1. Johnny Advertising from, July 10, 2014 at 2:14 p.m.

    GroupM is not abandoning open exchanges, they are merely shifting all their long tail weight into RMG (which they own...and sucks.)

  2. Seth Ulinski from TBR, July 10, 2014 at 3:57 p.m.

    Johnny -- check the link to the interview above, GroupM is indeed abandoning ALL open ad exchanges and RTB.

  3. John Grono from GAP Research, July 10, 2014 at 11:11 p.m.

    How come people are only getting concerned about bots now. They have been a fact for a decade or two. The smarter operators have been taking them into account for ages.

  4. Seth Ulinski from TBR, July 11, 2014 at 8:15 a.m.

    John- thanks for the feedback. Agree that bots have been an ongoing nuisance. In addition to the co's mentioned above -- could you share details on which operators (i.e. SSPs, DSPs, Pubs, agencies, other ad tech vendors) have been combating bots and how?

  5. John Grono from GAP Research, July 11, 2014 at 9:49 a.m.

    Primarily 'black lists'. Factoring in 'overall' discounts (e.g. 25% of traffic is non-human so discount your ratecard or your PIs.) into T&Cs. Nothing scientific - more pragmatism.

  6. Seth Ulinski from TBR, July 11, 2014 at 10:17 a.m.

    Thanks for the follow-up, John. Having sold a significant amount of digital media (premium pub, adnet, DSP) I never had a client ask for a discount based on an anticipated/estimated bot % of traffic...that 25% figure might be reasonable but there would need to be evidence/data (likely 3rd party post campaign report) to back it up. I'm curious to see what WhiteOps and others will continue to develop from a preventative standpoint.

  7. John Grono from GAP Research, July 11, 2014 at 6:12 p.m.

    Seth, that is if you buy is based on PIs. We focus on audience delivery here using a hybrid audience measurement system. In essence PIs are the 'element' used to construct a campaign to deliver an audience for the client. It's analogous to buying TV GRPs to deliver audience reach. The client is more interested in the reach.

  8. Shaun Pena from Consultant, July 13, 2014 at 1:51 a.m.

    Seems to me that even if there was a high amount of waste in RTB, if the overall performance or value of the campaign was better with RTB, why would you throw the baby out with the bath water? They must have some pretty compelling evidence internally showing direct outperforms exchange, even with all the shnazy programatic data.

  9. Anni Paul from BoscoSystems, July 13, 2014 at 8:14 p.m.

    Are there nuisances and waste? Of course. But give it time. Mobile RTB in particular is where we're going to see a more efficient and refined process in the future, I think...

  10. Seth Ulinski from TBR, July 14, 2014 at 8:48 a.m.

    Shaun -- great point and certainly will be interesting to see some testimonials / case studies. In the meantime it's a bit of a mystery given the efficiencies demonstrated via RTB to date (even with the "waste"). When I worked on the ad network side, we just couldn't compete with DSPs using manual buys. The adnets themselves have all but disappeared or morphed into SSPs/DSPs.

Next story loading loading..