Why Real-Time Bidding Wins

Data Driven Thinking by Zach Coelius of Triggit"Data Driven Thinking" is written by members of the media community and containing fresh ideas on the digital revolution in media.

Today's column is written by Zach Coelius, CEO of Triggit, an online advertising technology company.

Now that Google has bought Invite Media and publicly stated that the purchase was all about real time bidding (RTB), the market seems to have woken up to the fact RTB is not hype and it is here to stay.  The early claims that RTB wasn’t real, then that it was too expensive, then too small, and then too dangerous for publishers have been proven wrong.   Yet, the positive reasons for why RTB is winning have never been well defined.  So here goes.

The most important reason why RTB wins is that it makes publishers more money.   The cause of this is very simple; the more buyers who are active in any market looking at a good for sale, the higher the price will be; such is supply and demand.   In an RTB enabled market, every impression has literally tens or hundreds of thousands of media buyers with their own unique requirements and data looking at each impression and bidding for those that suit them best.  These ad impressions are not interchangeable commodities, but in fact unique events with highly differential values for different types of advertisers.   Thus, it is easy to understand that the more bidders there are looking at each impression, the more likely it is that one of those bidders will be willing to pay a very high price.   Compare this with a publisher’s sales force that might sell to a handful of advertisers or an ad network with maybe a hundred active advertisers at any given point in time, and you can see how RTB results in higher effective prices than other status quo options.

Another cause of the increased prices from using RTB is the proliferation and increased sophistication of cookie-based targeting.  Whether it is a cookie dropped from retargeting, derived from an advertiser’s CRM system or in-market intender cookies purchased from a 3rd party, user data has become incredibly important in display advertising.   The advertisers who buy using cookies are willing to pay five or ten times more to reach these highly targeted users.  Since most users are almost always in the market for some item or service, most impressions usually have at least a few buyers with a cookie on the user who are willing to pay higher prices. Remember that the advertising ecosystem incorporates hundreds of thousands of very diverse marketers with equally diverse target audiences.   Obvious the more of these buyers looking for cookie matches on a publisher’s inventory the better.  Average the effect of this across the highly heterogeneous inventory that a single publisher has and cookie matching with RTB drives prices up.

The third reason why RTB raises prices is that it is significantly more efficient for media buyers.   Using status quo media buying methods, over a third of an advertiser’s funds for display media are wasted sending out RFPs, IOs, trafficking tags, compiling disparate reports, de-duplicating attribution and managing campaigns with manual, tedious, and wasteful processes.   That is money that could and should have gone to publishers as media spend.    As any student of supply chains will tell you, increased efficiency in the chain on the part of the buyers is often passed through to the suppliers as a result of the competition among buyers for the good they need.   While some in the media business claim that an inefficient process is good for media sellers, what they are actually talking about is the competitive dynamic between publishers who have access to the buyers and publishers who do not.    You can see the intra-publisher dynamic playing out in the significant price differential that currently exists between the largest publishers and the long tail.    While increased efficiency will certainly reshape the dynamic among publishers, the reduction of waste is good for publishers as a whole and good for the industry.

Across the board, these three traditional market dynamics have been the drivers that have tripled average media costs on the exchanges since we started tracking them a year and a half ago: more bidders per impression, the highly differential value each advertiser places on a cookie, and increased efficiency. Publishers are already making more money with RTB, and as we on the buy side get better at what we do, and as more advertisers transition to buying in real time, we will only see further price increases.

Yet even with these three market forces causing CPMs on the RTB exchanges to rise, there is another even more important reason why RTB will eventually look at every digital media impression.  This is because RTB is not actually a second channel used to the exclusion of another sales channel.   RTB is additive to any existing channel.  What this means is that a publisher can expose their inventory to RTB bidders while at the same time maintaining all their other channels such as a direct sales force.   When a bidder through RTB submits a bid that is higher than what the publisher can get from their other channels (which is a very common occurrence, with bidding wars for in-market cookies resulting in CPMs over $50), the publisher can take the higher price.   Thus RTB enables a publisher to immediately realize the gains available from showing their inventory to DSPs without them having to abandon the channels that currently work for them.  Moreover, with the use of a floor price, it is trivial for a premium publisher to prevent channel conflict when exposing their inventory in this way.  Quite simply there is no reason why every publisher wouldn’t want to immediately expose inventory through RTB and start gaining additional revenue.

As RTB continues to increase average CPMs on the exchanges, more and more publishers will find it easier, more predictable and more profitable to sell their inventory in this way.   This publisher migration has already begun in earnest as the volume of available inventory on the exchanges is growing exponentially.   It is clear that a virtuous cycle is in full effect.  At the end of the day RTB and exchanges are simply more efficient markets, and looking back at the course of history, it is pretty clear that trying to fight against the market’s drive to efficiency is always a pretty dumb idea.

Follow Zach Coelius (@zachcoelius), Triggit (@triggit) and AdExchanger.com (@adexchanger) on Twitter.


  1. Good post Zach. RTB has a lot of very desirable qualities about it, which is why we have been investing heavily in our own in-house built RTB technology. RTB unfortunately still (1) creates fear, uncertainty and doubt amongst publishers because of the level of transparency provided, and hence the overall proportion of RTB-enabled inventory is still small and frankly (2) just as 10 years later, a lot of people in the industry and around it don't understand retargeting very well, a lot of people don't understand RTB very well at all. For those two reasons we are still investing in traditional tag-based buying and optimization. I'd love to move everything to RTB, it's just going to take some more time.

    I'd be in favor of (a) doing more industry education of what RTB is and how it can help and (b) market participants sharing some pricing and other information that could help make the case for market participants. We'd be happy to work along with other DSPs to help make that information visible to the public in an advertiser/publisher-sensitive manner if there is interest.

  2. Nice Article Zach but i am going to have to disagree somewhat.

    RTB was never made so the publisher could "win". RTB was made so that advertisers could buy media in a more efficient manner.

    The idea of RTB will work in the future but now the amount of advertisers does not scale in a true marketplace. There aren't "Hundreds of thousands of advertisers out there buying in real time". There is only a hand full of DSP's in existence. Lack of demand in a real time bidded environment results in little spend at the minimum price for Publishers.

    Until RTB brings on enough brand dollars to create a competitive environment, Publishers CPM's will continue to have downward pressure.

  3. LookingConfident

    Tyler ... You may have missed the point made by Zach, that for publishers ....

    ".......... with the use of a floor price, it is trivial for a premium publisher to prevent channel conflict when exposing their inventory in this way."

    At any given time (in any auction process), a 'seller' can only get what a (single) buyer is willing to pay at the fall of the hammer. And the more attendees at the auction then (it's logical), the higher the bidding will become.

    Publishers who like to 'play' around with their floor price, may get to "win" that much more, overall. Especially as more advertiser (through DSPs), 'warm' to RTB.

    Zach may care to advise if my above thought is correct or, not. TIA


  4. Matthew Burton

    Do not confuse DSP with with Advertiser. Right now there are around 25 DSPs that are buying inventory through RTB platforms but they represent thousands of advertisers. With more self serve DSPs coming to market this number of advertisers will only grow exponentially.


  5. Steve O'Brien

    Good viewpoint. But I don't think promoting RTB as the best way for publishers to maximize eCPMs is going to win over any new fans among advertisers and agencies. RTB is designed to maximize fees for publishers, just as all other auction systems are designed to benefit the seller (AdWords, eBay, Sotheby's, etc). But auctions don't always produce the best prices for buyers, and they don't always produce the optimal results for sellers, either. Until we start to describe RTB in terms of a win-win for advertisers and publishers (which it is), it'll be a long road to increased adoption.

  6. Mark Pearlstein

    Excellent article. I completely agree publishers will ultimately move every impression to RTB becuase RTB is Revenue Management and not just Inventory Management. And finally, the market dynamics of supply and demand are the best method for determining true value, are fairly universally accepted by all parties, and therefore create a marketplace where everyone believes they are paying and receiving what they deserve. A win-win for all.

  7. @MB and LC

    I think you both agreed with my point that until "advertisers warm up to RTB" and "self serve DSP's come to the market" that there is not enough RTB demand out there to be truly effective for the publisher.

    @LC. Yes you can set floors. But that isnt the point of the RTB auction is it? That is just protection against arbitrage.

    @MB... Isn't every ad network a "DSP" now?

  8. Rob, I whole-heartedly agree that industry education and shared knowledge is key, I would be more then happy to work with you and our brethren on the DSP side to help with these things. For instance a conference around exchanges, DSPs, SSPs and Data very much seems to be in order (hint, hint, John).

    Tyler, as Matthew pointed out, DSPs represent quite a few advertisers at the same time. By combining all the advertisers we represent it is a very large number measured in the tens of thousands. Remember Google Adwords is bidder on the Google exchange and that alone brings a quite a few bidders to the table. To your other point, I think RTB actually was invented for publishers to “win”. To the best of my knowledge Jason Knapp and the Fox guys were the first ones to develop RTB and they are a publisher looking for a higher return. My arguments above outline why I think that higher returns are and will continue to accrue to publishers.

    Steve, I think you are correct that driving up prices for inventory will not make advertisers to happy. But, think that effect is more then balanced by the increase efficiency, targeting, control and better outcomes to be had from using a DSP with RTB. The reason why I wrote this post is that I feel there has been far to much wailing and lamenting among publishers about the “terrible” things being done to them with RTB and that they were looking at this completely backwards. This is very much a win-win in my mind.

  9. Nice article Zach. From a technical perspective, RTB has enabled demand side technologies like us to add value to publisher content without having to go through a complicated and expensive integration. As soon as a publisher makes their inventory accessible through RTB, we’re able to access the page and create new data points that increase advertiser performance and ultimately raise publisher yield (especially for photo-publishers who have had no scalable content classification system in the past). In addition to creating new revenue opportunities for the publisher, the real insight you’ve outlined is how RTB harmonizes with the direct sales channel helping to smooth out their yield curve rather than disrupting it. Well done.

  10. Jason F.

    "Carl Jacobi, a noted 19th-century mathematician, counseled his students to "invert, always invert" when they encountered a particularly vexing problem. I think this is a great way to approach investing. After you compile all the reasons you should buy a stock, invert the question and state the reasons why you should not buy the stock. By doing this, you ensure that your research process is more complete."
    --Charlie Munger

    Same holds true for technology.

  11. I am sorry if this question is really basic. If a publisher gets 50% higher CPM because of RTB, isn't that extra 50% the advertiser is paying coming at the cost of some other impression/publisher ? so the total amount all the publishers make still stays constant with or without RTB

    Or will the move to RTB convince advertisers to increases their online media spend ?

  12. Binnt,
    No, not necessarily. The media value equation looks something like this. quality, volume, and targeting of media bought + the cost of buying media= audience reached and goods sold. In that value chain there is waste. For instance in the old line media buying world or RFPs, IOs, trafficking tag and collating reports a ton of money is wasted on processes and people that add no value. That waste is money that could and should have flowed through to publishers since the value of the media is functionally equal to the end result of goods sold. Another area of waste is media that is targeted to the wrong people that results in no goods sold. If I am selling a product and buy ads that are show to people who will never buy my product that is waste that I need to discount from my total media cost.


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>