Beyond The Hype: Why Real-Time Bidding On Ad Exchanges Is Challenged

Pascal Bensoussan of Aggregate Knowledge“Data Driven Thinking” is a column written by members of the media community and containing fresh ideas on the digital revolution in media.

Today’s column is written by Pascal Bensoussan, VP of Products at Aggregate Knowledge, a buy-side optimization platform.

As the “AdWords-ish-ness” of DoubleClick’s AdX 2.0 interface makes obvious, the goal of today’s exchanges is to bring greater scale and efficiency to the display space. There have been several reactions to this by the advertiser:

  • Reaction A: “Great! Media buyers’ ability to cherry pick impressions and effectively price each one will usher in a new era of performance-based marketing.”
  • Reaction B: “Rats! These exchanges will force a consolidation of 90% of inventory into the hands of same 3+ Goliaths.”
  • Reaction C: “So what?”

Don’t be quick to dismiss perspectives B and C as the view of the naysayers and uninformed. A look at the differences between the display and search marketplaces yields some important insights supporting these views.

Remember, search grew wildly from the widespread adoption of small advertisers who could go create a text ad themselves and fund their campaign on their credit card. Display is the opposite: due to the higher costs of entry, display is a medium-to-large advertiser/agency game. Even with the emergence of self-serve interfaces offering pre-packaged creative templates, the generation of non-amateur-looking display creative with interactive rich media components is beyond the capabilities of the vast majority of small businesses. Layer on the complexity of managing audience data and impression-level bids required for participation in an ad exchange, and most media buyers are brought to their knees.

The point is that exchanges won’t have the massive advertiser base that search leverages to maintain competitive eCPMs for publishers. The one exception is a few highly competitive categories such as retargeting, in which advertisers and ad networks will be fighting over the same cookies. AdWords boasts well over one million advertisers. The largest display networks have, what, maybe a thousand?

Fragmented advertisers, all trying to test various pockets of inventory at different times, will certainly create irrational bidding patterns that will challenge even the best real-time bidders. To make the pricing dynamics a bit more complex, publishers will maintain artificial price floors through reserve prices. On the one hand, those reserve prices will force enough transparency from publishers to get advertisers to pay. On the other hand, they will create deeper channel conflicts if the same inventory is sold directly. For most categories, this leaves all but the dregs of display inventory out of the exchanges for the foreseeable future.

For many marketers, display campaigns over ad exchanges will appear to under-perform compared to search campaigns over content networks, unless their branding value is actually captured. The Direct Response marketers, most experienced with performance-driven, auction-based marketplaces, will be using this new source of inventory, as the second place to spend their search budgets. In addition, brand marketers will be challenged to move their advertising budgets to ad exchanges because of brand safety concerns and because most exchanges do not trade video inventory. Even if standards for internet video advertising were to emerge, publishers would still be unwilling to risk their highly premium inventory on the exchange. Further, premium advertisers would not buy long tail video inventory without guarantees on content quality and brand safety controls. Those challenges are not insurmountable, but they will certainly slow down the rate at which marketers adopt the exchanges.

These reasons are why many agencies and even a few data providers we, at Aggregate Knowledge, have spoken with recently are concerned about (or rather unconcerned with) the nascent exchange marketplaces.

Back to the real world: why will Exchanges succeed?

Google. Google’s determination to bring all unsold and preemptible impressions from DFP publishers into the exchange will certainly create the critical mass of inventory necessary to bring large advertisers and large agencies to the game. Though, in the short term, Google will likely need to subsidize AdSense publishers with a higher cut of revenue to take lower yielding display ads.

Networks. This just in: Networks are not dead. They will continue to provide the targeting, scale and consistent delivery that large advertisers demand across many sources of inventory. Most ad networks are already leveraging ad exchanges to backfill IOs after depleting their direct publisher inventory. With nascent real-time exchanges, the most technology-oriented ad networks will be able to successfully arbitrage exchange inventory for their clients.

Arbitragers. Just as in search, the big winners in the Ad Exchange world will be the real-time arbitragers. These are the players – including ad networks, agencies, holding companies, large advertisers with in-house media buying capabilities, and small-timers, some with their own bidder, some that run on others 3rd party bidders – who can chew up the complexity of exchange inventory to spit out consistency and predictability for advertisers – and make a killing in the process. That been said, most pure play arbitragers will need other revenue streams to sustain their business as arbitrage opportunities get increasingly difficult. This is exactly what happened in the SEM space.

Demand Platforms. Those next-generation buy-side optimization agencies are the heroes of the exchange era – the ones that are going to deliver on the efficiency, targeting, and scale promises of the exchanges. More than just bidding for exchange inventory, those technology savvy optimizers will leverage a high performance, data-driven, and transparent infrastructure to execute campaigns efficiently and at scale. They will smash down the optimization silos, with data infused at every level to deliver the best possible experience to consumers, cost-effectively. To succeed, demand platforms will be relying on four core technology components:

  • Robust data management infrastructure to collect, normalize, and manage the user, site, and campaign data under one roof.
  • Easy-to-use interface to manage biddable audience segments and campaigns with precision, control, and speed.
  • Real-time decisioning engine able to leverage all the data available to bid for and match the best impressions with the best campaign, the best creative, and the best content, algorithmically and in split second.
  • High performance reporting infrastructure delivering timely status and actionable insights on the cost and performance of the campaigns, the audience reach, and the media purchased.

While I confess not being totally impartial, I remain convinced that without those four core technology components, real-time bidders are doomed to become a commodity. It will be fascinating to see how it all plays out.

Follow (@adexchanger) on Twitter.

Enjoying this content?

Sign up to be an AdExchanger Member today and get unlimited access to articles like this, plus proprietary data and research, conference discounts, on-demand access to event content, and more!

Join Today!


  1. Matthew Goldstein

    Pascal, great post.

    I totally agree with you that the ad networks are not dead — though I do see some consolidation happening. As you said, those ad networks with a competitive technology advantage will succeed in their ability to use data to arbitrage the abundance of exchange inventory.

    My big question is what do the publishers end up doing as the ecosystem quickly changes? Will they continue to send all of their inventory to the ad networks or will the evolve with a new model?

  2. Matthew, the inflection point all revolves around ecpm. As bidders on the exchanges get more efficient and better at reducing waste we can and will be able to pay more for the good impressions that we want. That will raise average ecpm for publishers who put their inventory on the exchanges. Once that ecpm grows to the point that it is higher then what ad networks can offer you will see a ton of publishers making their inventory available for bidding on the exchanges rather then giving it ad networks. You will see a huge tipping point for the exchanges when that happens.

  3. Excellent post, Pascal. Very good points about the coming evolution of the ad world that will still take a little while. Euphemism of the day: “the “AdWords-ish-ness” of DoubleClick’s AdX 2.0 interface”.

    Zach – I’m also very interested in seeing how the publisher participation angle plays out. As long as the exchanges can keep enough mindshare while they reach for the network effects Pascal mentions, waiting for larger numbers of advertisers, and exchanges can reduce the barriers to trial and to “heartbeat” – that is, to keep the light on and keep a set of tags live in whatever ad system they have – I’m hopeful we’ll get to that exchange-driven future.

    The danger (for exchanges and networks too) is that the transactional costs of acquiring publishers, and having them plug into a system that is not quite ready yet don’t sour them on the experience and make this transition take longer than it should. Some of these “switching cost” delays are going to be a big part of this transition but hopefully all the players in the space can minimize these.