History of the Ad Exchange Landscape Part IV: The Ad Network Is Dead, Long Live the Ad Exchange

Part IV: The Ad Network Is Dead, Long Live the Ad Exchange

I’ve written a few times about Ad Exchanges on this site, as they are one of the more exciting areas to think about in digital advertising right now.  Ad Exchanges such as Google Exchange, Right Media Exchange, or Microsoft Exchange are an elegant solution to this mess of redirects and inefficient monetization for the industry at large and offer the revolutionary opportunity to do true audience targeting.  Via an ad exchange, sellers can auction their inventory to the highest bidder through a single redirect, and buyers can evaluate and bid on that inventory impression by impression using rich pools of their own or 3rd party data.  This rich targeting gives ad exchanges a big advantage over ad networks in terms of transparency and targeting capabilities and since ad exchanges aggregate the inventory from thousands of publishers in the same place, exchanges also offer exponentially more reach than any ad network ever could.  Combine that with the fact that the exchange is also highly transparent in terms of pricing and doesn’t mask a markup on the media like an ad network, and you can see why both advertisers and publishers are pretty excited about the ad exchange opportunity.

Of course, if your business is optimizing ad network spend, the exchange is actually a threat, since the exchange aims to do exactly the same thing as a network optimizer, but just on a larger scale and therefore, with greater efficiency.  But unlike many of the ad networks out there, the network optimizers weren’t just an office of sales people, they had real technology, teams of developers, and knew how to build product.  So instead of trying to push publishers away from the ad exchanges, they’ve for the most part embraced ad exchanges and real-time-buying (RTB) systems, and have worked to aggressively assimilate demand from the ad exchanges into their auction and optimization algorithms.  RTB is certainly a growth area for digital advertising, but since ad networks still provide about 70 – 80% of the demand dollars out there most publishers can’t walk away from that source of revenue just yet.

So where does digital advertising go from here?  Well, there are a few other issues lurking in the background where Supply Side Platforms will play a key role.  The first issue is data leakage, where advertisers seek to data-mine publisher audiences though dropping their own cookies out of ad buys, or even using javascript tags to scrape the page content the ad serves on as well.  In my experience, publishers drastically underestimate the risk of data leakage to their business, mostly because there aren’t many tools in place to help them manage the problem.  Currently there are only blunt-instrument approaches, but the SSPs are hard at work building new tools to expose the problem to publishers and help them take on this issue head first.  Also on the horizon is the concept of audience futures, or guaranteed RTB, which combines the targeting benefits of RTB with the placement and priority guarantees of a premium ad buy, and by the way, can bypass the ad exchange altogether.  Data Management Platforms like Demdex and RedAril have already built a model for audience profiling systems, but it’s such a natural place for the Supply Side Platforms to go because it provides a new premium sales product to publishers and uses most of the same infrastructure as the ad exchange monetization plumbing they already provide.

Hard to say exactly where this will all land, but it should be interesting to find out!

Read the prior sections of this series:

Part I: Rise of the Ad Networks

Part II: Network Fragmentation and the Ad Ops Problem

Part III: Network Optimizers to the Rescue (?)


  1. Loved this 4 part series. Very insightful into the evolution of the ever changing landscape. So, how do the Exchanges themselves actually make money if their is price transparency? Do you have a write-up on revenue models between all the players in an exchange environment?

  2. Hi Zachary,

    The exchanges make money by charging a small fee on each transaction – like a credit card merchant. I don’t have a revenue model written up on this site, but almost everyone works on a ‘cost-plus’ model, meaning they charge a % fee on the media cost that runs through their system. The DSPs charge the advertisers a % fee, the SSPs charge the publishers a % fee, and the ad exchanges charge the DSPs a % fee. Those fees are usually 10 – 20%, depending on the company and client volume.

    The idea was to make the pricing in an RTB environment very transparent, which contrasts to the opaque model that ad networks use, where they charge advertisers a fixed price and usually don’t say what the media actually costs. So Ad Networks that happen to buy through exchanges will usually work on an arbitrage model, buying the media for one price, and charge the client another based on a variable markup. Finally, the data brokers and data exchanges. These companies have the most complex model out there, and might charge on a per cookie basis, usually called a ‘per stamp fee’, a fixed fee for exclusive use over a set time frame, or even use a separate auction to set the stamp fee. All that said, data fees typically back into a $1 – 3 CPM fee, incremental to the media cost.


  3. This series had been great reading.

    So what you’re saying is that as a publisher, since 70%-80% of the market is still owned by networks, I cannot yet effectively rely solely on an exchange? I still need to daisy-chain (or let a PubMatic type of service do that on my behalf) and in the best case throw that exchange somewhere in the chain?

  4. It might not be quite that much, but yes, ad networks still control a large share of budget, so if you are focused on driving the most amount of revenue through this channel, you can’t ignore ad networks.

    It’s an interesting time though because the lines are really blurring between ad networks and RTB from a demand standpoint. Advertisers initially started to use networks for cheap reach, but that’s evolved over time. Today, the most successful ad networks have in house technology, and are actually performance engines that target and profile audiences. Many of those networks are actually out buying on an RTB platform for the same reasons advertisers are – it’s just more efficient from and optimization and operations perspective. In a few years, I would imagine that most all remnant advertising will be done through an RTB channel, even if an ad network is the buyer. In fact, if you work with an RTB platform that accepts network demand, it would be a smart thing to only take a flat bid from those networks to minimize the opportunity for them to arbitrage the media.

    So at the end of the day, the ideal setup, from a revenue standpoint, would be to work with a system like a PubMatic, Rubicon, AdMeld, or other solution that can auction your inventory to the highest bidder on an impression basis, regardless to whether or not they are a network, DSP, or something else. Even if the network cannot transact in RTB, the system can consider them as a static bidder in the auction. You may get the most yield out of an RTB channel for the first and second impression per user, but then see more money from a combination of networks after that. That’s not really something you can replicate on your own with a daisy-chain, but you could probably get close by putting a frequency cap on your RTB tags, and then running your network daisy chain below that. Not a great way to do it, but it would work.

    The main problem you have there is that frequency is a really key driver of bids for one segment of bidders in RTB (advertisers prospecting for new users), and virtually irrelevant on another segment (advertisers remarketing to users they’ve seen before). So if you look at a histogram of eCPM by frequency, you see two clusters – one at the front where there’s lots of competition but low bids and not much price elasticity, and another way at the tail where there’s not much competition, but there’s tremendous price elasticity, and high bids. You’ll miss out on that tail section without an SSP in place.

    Hope that helps –


  5. Can you explain why guaranteed RTB can bypass the ad exchange altogether? Where can I find more sources about guaranteed RTB? Thanks

  6. Hi Wes,

    Guaranteed RTB can bypass the ad exchange because it doesn’t need the ad exchange to create inventory liquidity – that is, it doesn’t need the exchange to provide the ad impressions to bid on because it knows exactly where those impressions will come from – the specific publisher with which they have a guaranteed deal.

    This can be a difficult topic to read about, because there isn’t really a market to speak of right now, at least not in the way you are thinking about it. I would search AdExchanger and Digiday for ‘Private Marketplaces’ or ‘Private Exchanges’ for more context.


  7. Ben,

    Thanks a lot for the informative articles. I am a little confused as to the working of the DSP’s and the SSP’s. Where do these fit in the supply chain from advertisers to publishers and what roles do they play? How are their roles different from an ad exchange? Also, a few more details about the Data Management Platforms will be helpful.


  8. Thank you for such a great series. Very informative and usefull.
    But can you describe the work of ad optimizers in more details?
    For example, we have a publisher who wants to get more from his inventory, we have several ad networks publisher can work with and an ad optimizer.
    Do the publisher need to sign up for all ad networks if he wants to work with ad optimizer?
    As I understand ad optimizer analyzes statistics for publisher on each ad network, but ad optimizer can get this statistic only if he gets publisher’s credentials from account on every ad network.
    Or maybe they have another workflow?
    I’m confused a little bit. Could you explain?

  9. Thanks Anton –

    You could structure your relationship with an optimizer in a few ways – some publishers like to create and manage network relationships themselves, or may have existing network relationships from before an optimizer integration. Other publishers like to just have their optimizer manage all their indirect demand relationships, whether it be with a DSP or a more traditional network. I think in most cases publishers tend to have a couple network relationships already, but then transition those to their optimizer, which allow sources many more network relationship’s on the publisher’s behalf. That said, it’s not unusual for publishers to keep up with those legacy network relationships.

    The only thing you need to really understand is that the optimizer has to understand what the effective yield is from each network. If you give the optimizer direct access to the network’s reporting, they can get the performance figures on their own, of if you don’t want to hand over your credentials, you’ll have to provide regular updates – most optimizers have an interface where you can update this figure yourself. If you choose to provide the updated stats yourself, you’ll just have to understand that the optimizer won’t be able to do as good a job. A lot of networks actually have reporting APIs that allow them to pull updates every hour or even more often, and there’s no possible way you can keep up with that manually.

    Hope that all makes sense –


  10. Hi Agam,

    In terms of your questions around data management platforms, you’re in luck, as I wrote a four-part series that explains DMPs in quite a bit of detail. You can find the links to those stories below:

    As to your question around DSPs and SSPs – DSPs sit on the buy side of the transaction, and enable advertisers to create, target, and optimize their media buys over ad exchanges. SSPs on the other hand service the sell side of the transaction, and enable publishers to sell their inventory on ad exchanges, and optimize between lots of sources of demand. These are biased systems, meaning they are designed to serve either the buy side or the sell side, and offer proprietary technology and algorithms to accomplish that goal. Ad exchanges on the other hand, are agnostic platforms, meaning they don’t serve buyers or sellers, but simply provide an open marketplace. So the ad exchange just hosts the auction, but the DSP is like the buyer’s expert, guiding them to the best inventory for their purposes, telling them when and how to buy. Likewise, the SSP is like the seller’s expert, saying where and how to sell their inventory, how to price it, who to sell it to, what rules to put in place, etc. Then, considering everyone’s positioning and rules, the exchange provides a market for the two parties to transact.

    Hope that helps!


Leave a Reply

Your email address will not be published. Required fields are marked *