The Exchanges' Redesign of the Web

"the executioner" opinion expressed below is written by Sam Temes, Associate, Platform Management, Ad Exchanges, at Razorfish.

The ExecutionerIn the last few months, those of us working in the world of ad exchanges have been bombarded by articles and discussions about issues that will directly affect our careers: privacy regulation, data sources, emerging business models and so on. In the same time period, I've seen virtually no discussion about how exchanges may positively affect Internet experiences for everyday users like most of my friends and family.

Several years ago, ad networks and content ads turned blogs into businesses and gave large-scale publishers an outlet for excess inventory. Designers were encouraged to create as many impressions as possible. As a result, we all became used to seeing hundreds, if not thousands, of online ads every day and learned to ignore or block 99% of them. Now that the network model is disappearing, the user experience to which we're accustomed will disappear as well. As advertisers and their agencies become fluent in the value of publisher assets (see The Value of the Ad Cow), publishers will be driven to redesign for only the most valuable assets, potentially making the Internet a more quiet and peaceful place.

User behavior is already showing that some ads really do work, while others don't (or at least not nearly as well). In looking at the cost of impressions by size, large skyscrapers are proving to be far more expensive than smaller logos simply because the average person (i.e. my 60 year-old mother or banking friend reading their favorite blog) is more likely to click and ultimately make a purchase from a large ad than a small one. The higher value in the exchange translates to more money for the publishers, raising the question of whether or not scale is still King. There is room for a new type of publisher who doesn't split articles on to 4 pages with 8 different ads to generate 32 times the impressions, but instead, provides less cluttered content with one large, top dollar ad. Envision one ad per page (or even fewer) that delivers the right message, in the right place, at the right time. I'd click, and I'm pretty sure most of my friends would too.

Ads with data behind them are showing similar lifts in performance over basic anonymous impressions, and the data is necessary to truly achieve the above – it's just trickier to explain to our friends. As publishers hone in on their most valuable data, they are being incentivized to maximize it and will likely change the model by which we pay for our free content. The New York Times provides their content without a registration because they could traditionally produce enough ad volume to stay in business. Soon, we (as exchange buyers) won't pay for the ads that our friends and family ignore, forcing the Times to find a new way to charge visitors for content. They will likely generate revenue by requiring us (as visitors) to login and maybe even answer brief surveys that provide the data behind the most profitable ads. With open source tools, small blogs can afford to do the exact same, putting content behind a gate that anyone can open by answering a few questions.

As we build out more tools to measure the true value of display, we are likely to find there isn't enough of the best stuff and far too much of the worst. Publishers will be pushed to create more of the good inventory and less of the bad, leading to a fundamental shift in how we experience the Internet. I'm looking forward to seeing fewer ads that are more relevant, even if it means answering a few questions and logging into a website to get my free content.

Follow Sam Temes (@SamTemes), Razorfish (@Razorfish) and AdExchanger.com (@adexchanger) on Twitter.

5 Comments

  1. The network model is not going away, a few numbers has been thrown around, but there are roughly 300 ad networks, and the numbers growing. I do agree that there will be some consolidation, maybe Google will emerge as the victorious champ soon enough, with their deep pockets. Exchanges will be just another tool ad networks will leverage on. Exchanges will be also be another tool publishers use directly to monetize their remnant inventory. There are many who has done so already by passing the networks. And those publisher that are just producing mass impressions just for the sake of producing it will be relegated to run low tier cost per sales/lead campaigns.

    Reply
  2. Jim Spanfeller

    This has long been the theory behind ad networks, exchanges and lately data echanges. The issue is of course, that in practice this theory does not work. Both from sites where registration is or at one point was required as well as in large BT practices that use past web behavoir and implied understanding from IP addresses.

    The idea that ad networks and now exchanges have pushed money into blogs is correct but at the end of the day the level of funds directed at any specific blog is not nearly enough to fund a full time salary plus web hosting costs.

    The core notion behind ad targeting is fundementally flawed for anything but direct response or put another way demand fulfillment. By definition demand creation is a much more messy occupation but is also a much more effective way to change mass perceptions.

    Reply
  3. Matt Greitzer

    @Jim Spanfeller, sounds like you are saying targeting doesn't work. Interesting position. Google already proved the notion behind ad targeting is extremely effective, and extremely lucrative. Extensions of this, like Yahoo’s search retargeting product, are some of the most effective advertising you can buy, anywhere. The problem isn't that ad targeting doesn't work, it's that it hasn't been done well due to lack of sophistication on the buy-side, and lack of capabilities on the sell side. And while you are correct that DR advertisers will be first to the party, there is huge demand for this from branded advertisers as well - some are even building their own tech platforms to manage this.

    As for this comment, "at the end of the day the level of funds directed at any specific blog is not nearly enough to fund a full time salary plus web hosting costs," I'll let Ad Exchanger field that one.

    Reply
  4. Jim Spanfeller

    @Matt, What I am saying is that classic BT does not work. Retargeting does indeed work but it also has a fair degree of ...well...creepiness to it. And used unehtically can disrupt the trust levels that must exsist between buyers and sellers to have an efficient market. For example I am aware of several instants where an ad network has worked with a client to get potential customers into their purchase funnel...and then through pixeling of their clients web site has turned around and sold targeting to the clients competitors of the folks who abandoned the purchase funnel somewhere before complition. So fundementally they used the efforts and spending of one client for the benefits of another competitive client and to add salt to the wound my guess is that the initial client had no knowledge of this let alone any compensation.

    Reply
    • While I understand your opinion regarding the retargeting, I don't fully agree. Retargeting is only creepy and mischievous when ad networks make it so. Part of what exchanges are doing is "opening the kimono," so to speak, so retargeting is more transparent and networks aren't able to get away with the less than honest practices that have become common place.

      None-the-less, the core of what I was getting at is that the experience for users is changing because the business model of how media is sold is changing. I think that premium inventory may stay premium and be sold in the same manner as it has been for a long time. What is interesting to see is how long tail inventory creation changes as direct response advertisers become more savvy in their buying.

      Appreciate the feedback and engagement from both you and Matt!

      -Sam

      Reply

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>