To RTB or Not RTB, That is the Question

This is an unbaked opinion piece.  Please comment to keep the conversation going as this will certainly provoke some commentary.

I’ve been participating in many conversations (conferences, panels, articles, journals, meetings, etc) about Real Time Bidding and it’s potential affect on digital media.  There are very few true real-time-bidding inventory sources today and only a few Real-Time Ad Platforms (i.e. AdNexus).    Without revealing too much information from our day to day in the office, we see that less than 25% of our inventory is coming through Real-Time Bidding.  According to conversations with our peers, this is fairly consistent with the industry.

The big question is not whether RTB is coming (or here), but rather if you (as a marketer) need to harness it.

Demand side platforms (DSP) build/buy/partner with RTB-enabled platforms
If the big agency holding companies have their way (note: I’m part of MDC Partners), then the majority of standardized display will be running through these platforms.  If these platforms and capabilities are RTB enabled and the processes and procedures are in place to run in Real-Time, then there can be a very large demand (liquidity) for RTB enabled inventory.  This means that the Supply Side Platforms need to on-board enough inventory to be sold in Real-Time.  I include data (& audience) within the “inventory” label.  1st, 2nd, and 3rd party data (& audience) needs to be RTB enabled as well.

You need both the Demand Side and Supply Side to be real-time enabled to make this happen.

This theoretically seems easy, but how many “big-name” publishers are real-time-transaction enabled?  The majority of RTB inventory is long-tail today… and this needs to change to really attract the significant dollars to this space.  The Supply Side will either aggregate at large exchanges or use Supply Side technology to make this happen.

If you do not buy in real-time, can you exist in the future? Simple:  yes.

I know people in the marketing space (including myself) are trying to draw parallels to the financial markets and I’ll continue to do so here.  There are many financial brokerages that have access to transact very quickly and they do.  But other brokerages and buyers are able to transact (with significant dollars) in near-real-time (or less than near-real-time) and still be in solid financial positions and the ability to move markets.

Speed does not guarantee success in markets: finance or marketing.  The art of orchestrating the sciences is where the men are going to be separated from the boys. You can give me access to AdNexus tomorrow morning and just because it’s fully RTB enabled, does not mean I’ll be any better than buying site direct or thru a non-RTB network.

Unless the algorithms and the “view” of real-time inventory is exactly the same, no two RTB platforms will perform exactly the same.

Have I convinced you that you do not need real-time bidding to be successful today?

Now the contrary:  The faster an RTB platform can operate (in milliseconds today), the more inventory it has access to.  This is a big deal.  If you are building models to predict the future, then you want to see as many impressions as possible to have access to pick from.  Some say buying clout does not matter but does it?  A question to ponder.

The net/net:  RTB is a buzz word today and many people are talking about this space.  I believe that where there is smoke there is fire, thus, there is something to take note about RTB.  With the opportunity to value individual impressions and data with the ability to go the last mile and action it, potentially ahead of competitors, RTB does have an edge to those who can use it.

I believe that the long term winners in this category will not be the scientists, but rather the artists who know how to apply the processes, procedures, and strategies for their clients.

Please comment to keep the conversation flowing.  Would love to hear your perspective.

  • dannyflamberg

    The debate turns on the value of real-time buying. What do I win or lose in the moment? Are their incredible opportunities or amazing efficiencies that compel me to act in real time? Can any network convince me that at a particular moment in time they can extend my reach, frequency or persuasiveness? If not, its a moot point.

  • http://www.optimizeandprophesize.com/ jonathanmendez

    I’m going to take your orchestration analogy and run with it.

    Lots of people are proficient at reading charts and using instruments but of course a conductor is needed to make beautiful music. I think we will find the Leonard Bernstein of RTB eventually but remember – conductors are working with the music written by others. The real value is the music itself. That’s what people come to listen to. In our world the publishers are the Mahlers & Sondheims and as long as they’re living they will have the say who will conduct and orchestrate what they compose. I don’t think anyone is ready for Carnegie Hall yet but it sure is fun having great seats for the rehearsals.

    • http://www.darrenherman.com dherman76

      I’d argue that people went to go see Leonard Bernstein

      • http://www.optimizeandprophesize.com/ jonathanmendez

        fair point. maybe the real issue is why are there so few Leonard Bernsteins.

        • http://www.darrenherman.com dherman76

          Because it’s an art.

  • dannyflamberg

    The debate turns on the value of real-time buying. What do I win or lose in the moment? Are their incredible opportunities or amazing efficiencies that compel me to act in real time? Can any network convince me that at a particular moment in time they can extend my reach, frequency or persuasiveness? If not, its a moot point.

  • http://www.optimizeandprophesize.com/ jonathanmendez

    I'm going to take your orchestration analogy and run with it.

    Lots of people are proficient at reading charts and using instruments but of course a conductor is needed to make beautiful music. I think we will find the Leonard Bernstein of RTB eventually but remember – conductors are working with the music written by others. The real value is the music itself. That's what people come to listen to. In our world the publishers are the Mahlers & Sondheims and as long as they're living they will have the say who will conduct and orchestrate what they compose. I don't think anyone is ready for Carnegie Hall yet but it sure is fun having great seats for the rehearsals.

  • http://www.darrenherman.com dherman76

    I'd argue that people went to go see Leonard Bernstein

  • http://www.optimizeandprophesize.com/ jonathanmendez

    fair point. maybe the real issue is why are there so few Leonard Bernsteins.

  • http://www.darrenherman.com dherman76

    Because it's an art.

  • bokelley

    Darren,
    You make some great points here. From my perspective, RTB is a major infrastructure upgrade. What’s exciting to me is how many possibilities this unlocks.

    We all exist in an Internet world where we don’t ask the question “can I get there from here”. Can you imagine if there were three or four private networks (AOL, Compuserve, Prodigy), each of which were closed to each other – you couldn’t even email somebody on a different network? 20 years ago, that was the state of the world. Think of all the innovation that we’ve seen on THE internet. RTB turns a bunch of disparate display networks (AOL again, MSN, Yahoo) into THE display network.

    What will evolve is a mix of art and science (just like the web) but will end up looking very different than what we see today. You’re right – just handing somebody a plug into the Internet won’t let them do a whole lot, and nor will RTB magically create value. The tools, systems, and most importantly, the people who operate on top of the ecosystem are the fire creating your proverbial smoke. Our mission at AppNexus (thanks for the plug!) is the tools and the infrastructure, and we’re thrilled to power some of the smartest, most creative people in the business.

    Brian

    • http://www.darrenherman.com dherman76

      Happy to plug solid companies.

  • bokelley

    Darren,
    You make some great points here. From my perspective, RTB is a major infrastructure upgrade. What's exciting to me is how many possibilities this unlocks.

    We all exist in an Internet world where we don't ask the question “can I get there from here”. Can you imagine if there were three or four private networks (AOL, Compuserve, Prodigy), each of which were closed to each other – you couldn't even email somebody on a different network? 20 years ago, that was the state of the world. Think of all the innovation that we've seen on THE internet. RTB turns a bunch of disparate display networks (AOL again, MSN, Yahoo) into THE display network.

    What will evolve is a mix of art and science (just like the web) but will end up looking very different than what we see today. You're right – just handing somebody a plug into the Internet won't let them do a whole lot, and nor will RTB magically create value. The tools, systems, and most importantly, the people who operate on top of the ecosystem are the fire creating your proverbial smoke. Our mission at AppNexus (thanks for the plug!) is the tools and the infrastructure, and we're thrilled to power some of the smartest, most creative people in the business.

    Brian

  • http://neuvc.com/ Jerry Neumann

    The promise of exchanges is the deaveraging of both audience and price. And the promise of RTB is doing that in a way that permits third party optimization.

    If you’re deaveraging in a non-RTB environment, then you need to write rules that sit inside the exchange platform. This means you can’t optimize very effectively (or at all, if the exchange insists on being a black-box, as they all are to a great extent.) In a RTB environment, you can–theoretically, at least–see the context, see the audience and bid: everything is in your (or your DSP’s) control.

    I agree that if you’re Tide and your audience is everyone who wears clothes, then deaveraging is not very important. You’re marketing to the average. But if your product is more targeted or if your target audience is changeable (and who isn’t?) then you need to dynamically optimize. And if you are going to be effective at dynamically optimizing, then you need to be bidding in real-time.

    On your last point: advertising has always been an art. Sort of like cave-painting. Some of the cave painters are geniuses, but their tools are poor and their canvases rough. The cave painters are about to go through their very painful introduction to technology. Many of them won’t survive, but the ones who do will make an art we can’t even envision today.

    • http://www.darrenherman.com dherman76

      Jerry, thanks for stopping by!

  • jneumann

    The promise of exchanges is the deaveraging of both audience and price. And the promise of RTB is doing that in a way that permits third party optimization.

    If you're deaveraging in a non-RTB environment, then you need to write rules that sit inside the exchange platform. This means you can't optimize very effectively (or at all, if the exchange insists on being a black-box, as they all are to a great extent.) In a RTB environment, you can–theoretically, at least–see the context, see the audience and bid: everything is in your (or your DSP's) control.

    I agree that if you're Tide and your audience is everyone who wears clothes, then deaveraging is not very important. You're marketing to the average. But if your product is more targeted or if your target audience is changeable (and who isn't?) then you need to dynamically optimize. And if you are going to be effective at dynamically optimizing, then you need to be bidding in real-time.

    On your last point: advertising has always been an art. Sort of like cave-painting. Some of the cave painters are geniuses, but their tools are poor and their canvases rough. The cave painters are about to go through their very painful introduction to technology. Many of them won't survive, but the ones who do will make an art we can't even envision today.

  • http://www.darrenherman.com dherman76

    Happy to plug solid companies.

  • http://www.darrenherman.com dherman76

    Jerry, thanks for stopping by!

  • jeremy

    I think RTB is exciting because it opens up a new “type” of buying/selling that hasn’t been available previously.

    Following your financial analogies (because I think they are a really good analog to the ad market), there will always be “Warren Buffets” who look to buy assets with a longer term perspective in mind. However, in recent years, the rise of technology in trading platforms has allowed places like Goldman Sachs to make lots of money with “high-frequency trading” where milliseconds and fractions of a cent make all the difference. Now I don’t think RTB will get to the high frequency trading levels right off the bat, but I do think you might see RTB create a day trader type attitude towards media buying/selling. (Note: I am not endorsing or critiquing the pros and cons of a day trader/high frequency approach vs. a value investing approach. You can make or lose lots of money with each. I’m just saying that I think its a good analogy to what is happening in the marketplace.)

    For me, the bottom line is that RTB is a new way to buy/sell media that didn’t really exist before and will thus create opportunities that can be exploited by smart people.

    • dmcjoshua

      To clear up a common misconception, there is no analogy that can be accurately drawn between RTB and flash trading (aka high frequency trading). Flash trading was incredibly profitable for firms like Goldman because they could see all other incoming bids before casting their own, effectively seeing where the market was headed before placing their order — this is a form of frontrunning. The profit was in the asymmetrical information, not the speed of the trading.

      No RTB system today allows bidders to see the other bids, in fact you only get to see your own clearing price after the fact. There is no inherent value to bidding in real time, except in that it allows a calculation of value down to the per-user level that is effectively impossible with cookie segmenting.

      • http://www.darrenherman.com dherman76

        Josh, you beat me to the comment around flash trading. Spot on.

      • jeremy

        Fair point re:frontrunning. That was the wrong analogy. The overall point I was trying to articulate was that when technology enables faster transactions, I think that creates market opportunities.

  • senithomas

    The RTB debate has always been one that has been driven by the Demand Side of the equation. To your point unless the publishers get on board you cannot have an efficient RTB based inventory platform.

    Today there is no economic incentive for publishers to upgrade their systems. If anything it will drive down prices of premium inventory as it will provide more efficiencies in buying targeted inventory as all the algorithms will self optimize toward the cheap long-tail inventory that most of the RTB systems are currently selling. Especially, since the RTB platform often allow buyers to resolve the URL string allowing you to cherry pick specific publishers.

    A marketplace cannot survive if there is no supply and the quite frankly the current system is unsustainable for premium publishers. The difference between the NYSE and display exchanges is that each stock has an inherent value that is then effected by supply and demand. Even the commodity exchanges have different grades to exhibit the value of the asset.

    RTB is a nice concept, but there are still many issues that need to be resolved before we all chase after the next shiny new toy.

  • jeremy

    I think RTB is exciting because it opens up a new “type” of buying/selling that hasn't been available previously.

    Following your financial analogies (because I think they are a really good analog to the ad market), there will always be “Warren Buffets” who look to buy assets with a longer term perspective in mind. However, in recent years, the rise of technology in trading platforms has allowed places like Goldman Sachs to make lots of money with “high-frequency trading” where milliseconds and fractions of a cent make all the difference. Now I don't think RTB will get to the high frequency trading levels right off the bat, but I do think you might see RTB create a day trader type attitude towards media buying/selling. (Note: I am not endorsing or critiquing the pros and cons of a day trader/high frequency approach vs. a value investing approach. You can make or lose lots of money with each. I'm just saying that I think its a good analogy to what is happening in the marketplace.)

    For me, the bottom line is that RTB is a new way to buy/sell media that didn't really exist before and will thus create opportunities that can be exploited by smart people.

  • senithomas

    The RTB debate has always been one that has been driven by the Demand Side of the equation. To your point unless the publishers get on board you cannot have an efficient RTB based inventory platform.

    Today there is no economic incentive for publishers to upgrade their systems. If anything it will drive down prices of premium inventory as it will provide more efficiencies in buying targeted inventory as all the algorithms will self optimize toward the cheap long-tail inventory that most of the RTB systems are currently selling. Especially, since the RTB platform often allow buyers to resolve the URL string allowing you to cherry pick specific publishers.

    A marketplace cannot survive if there is no supply and the quite frankly the current system is unsustainable for premium publishers. The difference between the NYSE and display exchanges is that each stock has an inherent value that is then effected by supply and demand. Even the commodity exchanges have different grades to exhibit the value of the asset.

    RTB is a nice concept, but there are still many issues that need to be resolved before we all chase after the next shiny new toy.

  • Jed Nahum

    Darren:

    Interesting thoughts. You ask “If you do not buy in real time, can you exist in the future?” An interesting and related question to me has been whether, if you don’t sell your inventory in real time enabled fashion, you can exist in the future. As you point out, the demand side has quickly warmed to RTB and we’ve seen budgets move out of non-RTB buys and into real time. Nothing motivates a publisher sales force so quickly as when they see dollars move away from them.

    Ironically for those sales forces, I believe RTB is the last nail in the coffin of inventory commoditization. Unlike Wenda, I’m not freaked out by commoditization – to me it simply means that price becomes discoverable, not that price declines to zero. Setting that aside, the power of the demand side means that there is pressure on the sell side to sell RTB. For many pubs, the first taste has been good: there’s pent up demand for RTB since not enough inventory is available in that format (“demand liquidity,” as you say). So yield is nice. However, there are elements of prisoner’s dilemma in the move to RTB. Publishers who move first may get good yield returns initially, but after the market moves, the pent up demand dries up. Once everything is RTB, all inventory becomes fungible at some price, and maybe overall yield is actually less than it was pre RTB.

    All this worry about price, however, will be mooted if the incredible control and transparency of RTB enables offline marketing dollars to come online. That seems to me to be the ultimate in demand liquidity.

    • http://www.darrenherman.com dherman76

      The sell side (and previously on the demand side) has benefited from
      rate card opaqueness. RTB removes this and to your point, could
      surely put the final nail in the coffin for many pubs.

      If we believe an efficient market can exist, then it’s not if, but
      when this happens.

    • senithomas

      I believe this analysis is a bit short sighted.

      The economics simply don’t play out on the publisher side and even the additional shift of media dollars does little to solve this problem.

      For example a premium publisher with 100,000 impressions per month (to simplify) makes $15 CPM on his direct sales and $2 CPM on exchange based media, which also has more folks in the value chain so to be optimistic lets say on the exchanges the take home is a $1 CPM. Now say that the pub has a 40% fill rate so their total revenue is $600 through direct sales and $60 for all the rest.

      Where is the economic incentive to switch?

      You will never get the Forbes or NYTimes caliber publisher to switch over no matter how much pressure there is. They are better off decreasing their fill rate than selling on secondary markets that are RTB enabled, as RTB buying systems allow you to resolve URL as do verification services like AdSafe and DoubleVerify, which lead to heavy discounting of their impressions.

      Also dont forget that as advertisers we NEED publishers, we can only bend them over so many times and as the offline components of these pubs decrease in revenue there is more and more pressure for the online components to bring home the bacon. We need to find a compromise.

      In addition, this is all fine and dandy for DR/performance marketers, but when you look at the branding opportunity, which is the largest chunk of cash that could be shifted online the RTB solution does nothing to help the buyers and advertisers discern the environments that they are buying in from a brand impact perspective. Sure AdSafe can tell you there are no dead babies on the page, but they can’t tell you what the inherent quality of the page it.

      Finally, when you refer to making the price discoverable it is 100% based on supply and demand of a commodity, but we all know that not every impression is created equal and without some method of tiering the inventory based on valuation variables an efficient marketplace cannot exist. The power balance is so out of wack in the current ecosystem that I would hardly call this marketplace ‘efficient’. Sure buyers are happy, but content providers are dying. The low price also hides much of the over hype of data targeting, which usually isn’t worth the price of the data outside of highly intent driven verticals vs. brute force exchange bid management.

      I’m sure this isn’t a popular opinion, but all I’m saying is that we need to throw the supply side a bone and they will help the buyers move more branding cash online cause the agencies are the ones that benefit the most from shifting budgets as commission rates for digital planning/buying are an order of magnitude more than traditional.

  • dmcjoshua

    To clear up a common misconception, there is no analogy that can be accurately drawn between RTB and flash trading (aka high frequency trading). Flash trading was incredibly profitable for firms like Goldman because they could see all other incoming bids before casting their own, effectively seeing where the market was headed before placing their order — this is a form of frontrunning. The profit was in the asymmetrical information, not the speed of the trading.

    No RTB system today allows bidders to see the other bids, in fact you only get to see your own clearing price after the fact. There is no inherent value to bidding in real time, except in that it allows a calculation of value down to the per-user level that is effectively impossible with cookie segmenting.

  • Jed Nahum

    Darren:

    Interesting thoughts. You ask “If you do not buy in real time, can you exist in the future?” An interesting and related question to me has been whether, if you don’t sell your inventory in real time enabled fashion, you can exist in the future. As you point out, the demand side has quickly warmed to RTB and we’ve seen budgets move out of non-RTB buys and into real time. Nothing motivates a publisher sales force so quickly as when they see dollars move away from them.

    Ironically for those sales forces, I believe RTB is the last nail in the coffin of inventory commoditization. Unlike Wenda, I’m not freaked out by commoditization – to me it simply means that price becomes discoverable, not that price declines to zero. Setting that aside, the power of the demand side means that there is pressure on the sell side to sell RTB. For many pubs, the first taste has been good: there’s pent up demand for RTB since not enough inventory is available in that format (“demand liquidity,” as you say). So yield is nice. However, there are elements of prisoner’s dilemma in the move to RTB. Publishers who move first may get good yield returns initially, but after the market moves, the pent up demand dries up. Once everything is RTB, all inventory becomes fungible at some price, and maybe overall yield is actually less than it was pre RTB.

    All this worry about price, however, will be mooted if the incredible control and transparency of RTB enables offline marketing dollars to come online. That seems to me to be the ultimate in demand liquidity.

  • http://www.darrenherman.com dherman76

    Josh, you beat me to the comment around flash trading. Spot on.

  • http://www.darrenherman.com dherman76

    The sell side (and previously on the demand side) has benefited from
    rate card opaqueness. RTB removes this and to your point, could
    surely put the final nail in the coffin for many pubs.

    If we believe an efficient market can exist, then it's not if, but
    when this happens.

  • jeremy

    Fair point re:frontrunning. That was the wrong analogy. The overall point I was trying to articulate was that when technology enables faster transactions, I think that creates market opportunities.

  • senithomas

    I believe this analysis is a bit short sighted.

    The economics simply don't play out on the publisher side and even the additional shift of media dollars does little to solve this problem.

    For example a premium publisher with 100,000 impressions per month (to simplify) makes $15 CPM on his direct sales and $2 CPM on exchange based media, which also has more folks in the value chain so to be optimistic lets say on the exchanges the take home is a $1 CPM. Now say that the pub has a 40% fill rate so their total revenue is $600 through direct sales and $60 for all the rest.

    Where is the economic incentive to switch?

    You will never get the Forbes or NYTimes caliber publisher to switch over no matter how much pressure there is. They are better off decreasing their fill rate than selling on secondary markets that are RTB enabled, as RTB buying systems allow you to resolve URL as do verification services like AdSafe and DoubleVerify, which lead to heavy discounting of their impressions.

    Also dont forget that as advertisers we NEED publishers, we can only bend them over so many times and as the offline components of these pubs decrease in revenue there is more and more pressure for the online components to bring home the bacon. We need to find a compromise.

    In addition, this is all fine and dandy for DR/performance marketers, but when you look at the branding opportunity, which is the largest chunk of cash that could be shifted online the RTB solution does nothing to help the buyers and advertisers discern the environments that they are buying in from a brand impact perspective. Sure AdSafe can tell you there are no dead babies on the page, but they can't tell you what the inherent quality of the page it.

    Finally, when you refer to making the price discoverable it is 100% based on supply and demand of a commodity, but we all know that not every impression is created equal and without some method of tiering the inventory based on valuation variables an efficient marketplace cannot exist. The power balance is so out of wack in the current ecosystem that I would hardly call this marketplace 'efficient'. Sure buyers are happy, but content providers are dying. The low price also hides much of the over hype of data targeting, which usually isn't worth the price of the data outside of highly intent driven verticals vs. brute force exchange bid management.

    I'm sure this isn't a popular opinion, but all I'm saying is that we need to throw the supply side a bone and they will help the buyers move more branding cash online cause the agencies are the ones that benefit the most from shifting budgets as commission rates for digital planning/buying are an order of magnitude more than traditional.

  • http://www.MyOpenKimono.com paulbenjou

    Darren, Your POV is well taken. For now I tend to temper the full time use of RTB with non-RTB buying and direct site placements.
    Love the last bit in your post “….long term winners in this category will not be the scientists, but rather the artists who know how to apply the processes, procedures, and strategies for their clients.”

  • http://www.MyOpenKimono.com paulbenjou

    Darren, Your POV is well taken. For now I tend to temper the full time use of RTB with non-RTB buying and direct site placements.
    Love the last bit in your post “….long term winners in this category will not be the scientists, but rather the artists who know how to apply the processes, procedures, and strategies for their clients.”

  • Looking Confident

    What a great article and discussion comments being made. Love it!!

    < A marketplace cannot survive if there is no supply and the quite frankly the current system is unsustainable for premium publishers.>

    Well, “if your build it – they will come”, no? And if you build it in such a way that you create a level playing field (for both the buy and sell side) then “buying clout” becomes irrelevent on an impression or, 'click by click' basis …Supply and demand becomes the focus and creates the rewards …The same kind of rewards that will almost ensure that most”Big End of Town” publishers will come to play in some shape or, form. I think it's all very much “early days” just yet.

    http://seekingalpha.com/instablog/36191-looking