Commentary
RTB Is Growing Up
- by Aaron Doades , Op-Ed Contributor, September 27, 2012
It was recently reported by Parks Associates that real-time bidding (RTB) accounts for 12% of total digital display advertising. While 12% may seem like a small percentage when you consider how much airtime the industry devotes to RTB, that 12% actually represents a roughly $800 million business.
RTB is quickly moving beyond its infancy and into its adolescent
stage of life. The first demand-side platform (DSP) was introduced in the early 2000s and it took less than one decade for DSPs to change the way that media had been bought and sold before. We’ve already come a
long way since the inception of using data to reach audiences, and moved beyond initial challenges such as scale, standardization, and infrastructure. As a result, we now have search data, site-level
data, offline and online data flowing through our auction-based systems. Each of these positive steps and signs of growth help to push the industry; however, as with any stage of life, new challenges
set in.
advertisement
advertisement
Audience is still the name of the game for RTB, as emphasis is placed on buying the right user at the right time. If we know that a user is in the market for a new television, what does it matter if he/she is on a website that seems completely unrelated to electronics? The user has demonstrated either direct or indirect purchase intent for a television, so let’s show them an ad for one.
But while this thinking is grounded in a fundamentally sound theory, it leaves out one major piece of the equation: the location of the ad. I’m not talking about contextual placement. After all, RTB is about audiences -- not about contextual relevance. When I say “location,” I’m referring to the page that the actual ad is shown on. While it’s great that we have the ability to reach the right user at the right time, what’s the point of it all if the purchased ad is never seen by the user?
Conversely, if the placement is designed to make users accidentally click on ads, then the user may immediately navigate away from the site completely and campaign metrics simply don’t even hold true. The fact of the matter is that as technologies have become more sophisticated, marketers have noticed more data points surrounding what’s actually going on with their ad campaigns. As a result, ad placement and ad measurement have become pressing issues in RTB.
What the digital industry really needs is increased quality inventory and attribution models that take viewability into account. The adolescent stage of RTB is going to focus more heavily on inventory reliability and the ability to measure and monitor ad performance. Was the ad above the fold or below the fold? Which creative was displayed, and what inventory channel was responsible for the sale? With insight like these, publishers will have the opportunity to increase the value of their ad inventory, specifically for those sold on auction market. And, marketers will settle concerns about brand safety.
My prediction is that all inventory will eventually become biddable, except for custom executions. Creative is no longer flat in the data-driven world, as rich media and online video help to bring more branding dollars to the auction market. And as a result, RTB’s 12% will likely grow. It’s no secret that dollars drive innovation and demand. The opportunity in RTB is simply too big for publishers and for marketers not to execute against. The more dollars that flow into the data-driven market, the more publishers will get on board and move inventory to the exchanges -- which will in turn increase inventory quality.
This next stage will open up new doors for publishers to get in on the audience market. RTB is growing up.
Sorry, context does matter. I am fascinated by RTB and see great applications, but if the premise is only audience matters, and context (subject matter) or perceived quality/trustworthiness of the environment in which the ad appears is discounted, you have an issue that I think advertisers are beginning to understand. I was at a conference on this subject a couple of years ago where the buzz was that RTB would be 50%+ market share by 2013. Why isn't that going to happen? I think it's this overly simplistic data/audience-driven paradigm, which is going to tend to keep this market on the DR end of the spectrum. And it doesn't have to be that way...we just have to stop thinking about ad impressions as commodities, and put audience data in proper context (pun intended).
Hi Ron - Thanks for the comment. It seems like your response addresses two points: Audience is not the only thing that matters, and context/environment matter greatly.
Let me start by saying I agree with both of your arguments. Audience has been the main focus of RTB until this point, but I believe it's been a bit of an incomplete approach. The general philosophy of RTB continues to evolve and I think that the next phase is a much greater emphasis on inventory quality.
Having said that, I believe that context (through the lens of the page's content) only matters if it's being used to gauge purchase intent or brand preference of the audience for that page. The reason that a lot of contextual buys are so successful is that the page's content is being used as a proxy for the user's purchase intent or preference. If this insight about the user is best gleaned from the context of the page, then context matters. If an advertiser is able to say that someone is in market for a new car through the use of 3rd party data, in a manner that has a higher degree of certainty than the current page's subject context, then context is less crucial (in that particular case).
Regarding quality/trustworthiness of the environment in which the ad is appearing, I believe this is of utmost importance. One of the points I tried to emphasize in the above post is the need to look at the actual quality of the placement regarding safety, viewability, etc. I don't consider this to be "context" in the way that the word is generally used, so we may be saying the same thing but with a different understanding of the word "context".
Thanks Aaron. Sounds promising and grist for continued growth for sure. I would still argue though the premise that every impression can/should be tied to a transaction (full funnel perspective), but really that would be splitting hairs in the sense that what you propose is a significant advancement. I'm using context in two senses...the content/relevance of the page in which the ad is placed, and basically the brand/brand equity of the content in which it is situated. Placement would be still another variable. I contend for example that an investment products/services ad, with equal levels of audience targeting, is (far) more valuable on WSJ.com or Economist.com than it is on most blogs no matter how well-showcased, even most financial blogs for that matter, or say a syndicated finance page on the website of a pennysaver newspaper. I realize that die-hard data afficianados think this is b.s. - but remind them that they're 12%, not the 50% they said they'd be. People paying the bills do understand this, frustrating as it may be to many readers. To really flourish, you'd have to be able to assign values/weights to all these types of variables...some of which are highly nuanced and, yes, even subjective (gasp!)...in the exchange. If you can't, publishers will have to take steps to ensure that their true premium inventory is never available via these outlets.
A great discussion. My concern with DSPs(well one of my concerns) is the commoditisation of the industry and content. At its most base the DSP will opt for the lowest CPM (given other thresholds and parameters). When one party has content that is not getting the ad revenue wanted (that may even be contractually guaranteed), lowering prices can put that content back on the CPM radar of the DSP. Sellers can end up in a tit-for-tat price war. We end up commoditising everything including the content. So, if I was a content producer or owner I would only be offering up 'remnant' inventory to DSPs, and taking whatever crumbs fall off the table to support sales efforts to attract the value that premium content deserves, lest that content also be consigned to the downward value spiral of CPM led DSPs,
At the risk of over-simplifying, RTB targets the user at the right place and time. Because it is still early days, we should rightly debate the state of current practice, but for sure it beats the thumb-in-the-air approach of traditional planned buys, which is simply not scaleable at the transaction volume the industry demands. In the not too distant future (as in the next 12-18 months) the combination of RTB, social intelligence and search will begin to turn display into a deadly accurate tool for seamless communication-- context be damned! As for parsing the domain-level performance of RTB and overlaps with CPA, CPV, CPC-- we're doing that.
"The reason search excels in marketing is that it's all about intent, and what's even better, it's about identified intent neatly labeled by the search query. In the history of marketing, it's never been easier than this to intercept a motivated buyer."
The future (or, a more measurable means of gauging both advertiser and publisher's ROI) will be all based around the targeting of users and audiences & in Real Time-On Search Engine SERPs, in the 1st instance.
That day - when 'search meets display' is not too far away, I feel.
Preferably with any/all such targeting of impressions (across the one, independent & neutral global marketplace), being only of the viewable type.
http://seekingalpha.com/instablog/36191-lookingconfident/1117881-x01-x01-x01-x01-x01-x01dnt-and-sharing-pii-time-to-discuss-a-separation-of-church-ad-technology-and-state