Peeking Inside The Black Box Of Demand-Side Platforms And Exchanges

Ad Agents“Ad Agents” is a column written by the agency-side of the digital media community.

Adam Cahill is SVP, Director of Digital Media at Hill Holliday, a full-service communications agency.

Last week I gave the presentation embedded below on DSPs and exchanges at ad:tech in NY.

The premise of the presentation was that information is hard to come by in the category, because:

  1. Publishers obviously don’t want to talk about how much inventory makes its way into the secondary market
  2. Google has become the most important exchange, and they aren’t discussing the progress Adx is making
  3. Agencies are making announcements that make it sound like they have more going on than is probably the case

To try to get toward some real data I spoke to a handful of very smart, connected people who know what’s going on, and where things are headed including Turn’s Bill Demas, Zach Coelius of Triggit, DataXu’s Mike Baker, Tony Katsur of MediaMath, Sarah Fay from x+1 and Iggy Fanlo of Adrite.

I made four observations based on those discussions, as well as my own experience.

  • This is a really big deal – DSPs already account for 10% of US display spend, and RTB impressions represent about 35% of total US impressions. And all of this has happened essentially in a year’s time, without much interest from the client-side.
  • It’s not that complicated – If you’re an agency or a brand that wants to wade into this directly, you can get the vast majority of the value of the entire ecosystem by partnering with one of the six DSPs that is built for the future (ie, RTB). I didn’t include Appnexus on the list, because my understanding is that I as an agency can’t really call them up and have them become my DSP. If that’s wrong, apologies to Appnexus, a company that is always mentioned as an important player in RTB.
  • It’s not really about technology – The big change that’s going on is that what it means to plan and buy a big portion of digital advertising is changing. We’re moving from a world where the value the buyers create existed in the pre-launch phase of a campaign timeline to one where the value exists after the launch of the campaign. That means we need to re-train people, and bring new types of people into the industry.
  • Predictions – I closed with some short term and long term predictions.In the short term the composition of exchange inventory will expand beyond display, with video and mobile becoming easily accessible. Also, the quality of inventory will improve as premium publishers realize the economics are more favorable in the exchanges than they are when they allow networks to sell their excess inventory. In the longer term, I shared some of the specific predictions I heard about how big the exchange space gets in the next few years. For what it’s worth, the assumption I’m working with and building toward is the one that about half of display will be bought through a platform in the next two or three years.

Again, the whole point of the presentation was that it’s hard to get data, and this was my attempt to get some specifics on the table. If you have other data or a different point of view, would love to hear it.

Follow Hill Holliday (@hillholliday) and (@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. Adam, much appreciated for making your presentation online via AdExchanger.

    What’s really inside those blackboxes though?

  2. Awesome work Adam. It is really helpful to see an honest view of everything that is happening from the agency perspective.

  3. April Krueger

    As the manager of several exchanges, it’s nice to see some predictions on where exchanges are headed. From 2009 to 2010, I’ve seen the natural flow of media buying shift between exchanges on a daily basis, but often wonder if other companies are seeing the shift as well. I’m very pleased to hear your prediction is positive, in that, “the composition of exchange inventory will expand beyond display” and we’ll be playing with exchanges for years to come (hopefully)!

  4. There was zero discussion of what exchanges are doing to CPM’s. Exchanges are great for advertisers — but horrible for publishers.

    Its going to push all publishers into rebellion and go behind low-priced paywalls. No?

  5. Chris,
    There has been lots of discussion on the pages of adexchanger about why well designed exchanges actually increase the prices that publishers are getting for their inventory. For instance see .

    Having a scalable, transparent, effective and efficient way to buy media is good for publishers because it enables them to sell that which they produce much more effectively then the status quo.


  6. Hi Chris:
    A solid, piece and point of view. Appreciate the thought leadership and work on sizing the market. Hope you will continue to update the preso and make available via

    Agree that sophisticated platforms are not on the radar of the majority of CMO’s. But predict they will be as a driver of relevancy onsite and offsite.


  7. Hi Adam,

    Nice work on getting some facts and figures out of the most talked about but opaque spaces.

    The one thing I do think is slightly misleading from my perspective is the talk of the RTB ‘market’ and the suggestion that this is a new channel altogether. The trading of display media in real time is a huge shift in infrastructure but ultimately it’s still the same display units that have always been bought and sold and the same marketing goals. So whilst RTB may not be on the agenda of many CMO’s, we shouldn’t find this troubling as I doubt the issue of display media is something that will ever be not be.

    As an industry I think it’s up to us to adapt to these huge infrastructure shifts but at the same time see them in the broader marketing context.