Nearly every online marketer is concerned with transparency and brand safety.  These seem to be the biggest roadblocks to brands embracing display.  A quick search on AdExchanger reveals a dozen articles on both subjects in the last few months.  We’ve all talked about it yet there still seems to be more misconceptions than fact surrounding what is possible when it comes to real brand safety.

In April of 2010 Winterberry Group LLC reported that “transparency and brand safety have conspired to inhibit billions in potential online display ad spending.”  Brand safety continues to undermine a broad investment in display advertising by brand advertisers despite the obvious efficiencies being delivered by today’s ad networks.  This is because truly preemptive brand safety is difficult so most have settled for simple post impression verification and are therefore not getting a truly brand safe environment for their message.  But the DSP trend changes all that.

With the advent of contextual display advertising, a host of new advertising opportunities opened up for marketers across every industry.  Whether contextualization is done at the site level (i.e. ESPN.com is about sports) or at the page level (i.e. automotive content on CNN.com), the ability to target ads around content—in addition to demographic and behavioral profiles—is pretty powerful stuff.  But with these big opportunities comes an equal amount of risk.  Imagine an automotive ad next to a news story about a horrific car crash or an ad for a vacation package next to tsunami news coverage of the same destination.  We see it in contextual search as well as display.  It’s enough to make any brand marketer anxious about any sort of online advertising.

The good news is that advances in targeting technologies and demand-side platforms (DSP) with real-time bidding (RTB) are creating truly brand safe environments for brand advertiser.  But like transparency, everyone defines brand safety differently.  There is one feature that separates effective brand safety from all the other forms of impression analysis and delivery verification out there today.  It‘s preemptive brand safety.  Is your trusted flavor of brand safety truly preemptive?  Or is the brand safety you rely on from your DSP really just post impression analysis disguised as safe filtering?  This is another great myth in our industry often perpetuated on purpose only because preemptive is difficult to pull off.

Real brand safety, meaning a demonstrably safe environment for brand advertisers to promote their message online, comes from sanitizing the ad space before the ad is served.  Of course pre-impression analysis for relevance, performance and safety is very difficult to do so it’s the least prevalent form out there.  Evaluating billions of impressions a day for relevance takes a robust platform and deep integration with all of the real-time bid aggregation points.  But it pays huge dividends in both safety and efficiency by guaranteeing quality impressions and eliminating the need for pass-backs.  With truly preemptive brand safety you only buy what is safe with no waste in the equation.

But when pages are not categorized for meaning in advance, ads get served next to inappropriate content and the advertiser simply gets a report indicating that it has happened again.  That isn’t preemptive brand safety, that’s reactive reporting.  It’s the equivalent of closing the gate after the cows get out (and wander down the middle of the highway disrupting traffic and endangering your brand investment).

Everyone knew it would take third-party providers to deliver real brand safety.  Back in March, Jonathan Margulies at Winterberry Group acknowledged as much right here on AdExchanger.  It isn’t in the publisher’s best interest to exclude their own content from advertising—for a publisher the more bidders the better.  The interest in such technology falls on the demand side but most advertisers lack the engineering infrastructure to build such a solution.  In the display advertising landscape, providing preemptive brand safety is the responsibility of the demand-side platforms.  Various platforms provide this ability in varying degrees – from zero to truly preemptive.  Ask the question.  Make sure you are getting the truly preemptive kind that boosts efficiency and return on spend.

Reprinted from AdExchnager.com Data-Driven Thinking column, a media community blog containing fresh ideas on the digital revolution in media. Follow LucidMedia on Twitter @lucidmediaVIP and follow AdExchanger.com @adexchanger.com.

A recent study released by the IAB from Bain & Company titled Digital Pricing Research has ignited a virtual flurry of newsworthy commentary from prominent publications like iMedia and MediaPost about the place–and need–of networks in the advertising value chain. Declining prices, channel conflict, and devalued brands are the mantra of the network bashing fervor. It wasn’t so long ago that the pundits were predicting the end of the ad network model as we know it. The latest headliner from Adotas called Can 314 Ad Networks Really Thrive is an especially insightful look at the ongoing ad network saga.  The reality is that we’re just getting started.   Look at the statistics from the eMarketer study that said more than 90% of advertisers surveyed plan to use ad networks on their media plans and 75% said they planned to increase their spend to networks.  And look at what factors differentiate the plethora of ad networks according to the agency study; you might be surprised to find that price is dead last.  The major differentiators they cited were quality of inventory (28%), targeting (27%), and transparency (11%) then service, optimization, reporting, reach and finally price.  It’s true, there are a lot of ad networks out there, and there’s a good reason too.  Niche audiences and strategic reach.  And it’s not always one or the other either; it’s frequently both at the same time.   Advertisers need massive reach to get their message across and hit their key performance indicators.  It’s just a numbers game at the end of the day.  And not only do they need the big numbers, they also need to diversify and specialize and focus depending on what they’re selling or who they’re branding to.  One day they may need laser like focus on a specific audience segment and the next they may need numbers on a Biblical scale.  Usually they need both at the same time: the “massive niche” if there is such a thing.  Actually, there is!  That’s where we come in.  Our “super-clustered” network brings both audience breadth and contextual depth for performance and branding campaigns delivering focused reach all in one campaign.   And until there are more of these hybrid networks out there that are big enough to satisfy the advertiser’s thirst for reach and at the same time give them the finely tuned depth they require for performance, we will have a plethora of specialized networks serving the market.   And it’s not just the advertisers driving the explosion of ad networks either; it’s also the publishers. On any given day the average publisher has less than a 50% fill rate. Sure they have their own sales force out there beating the bushes for ad dollars but content is exploding all around them at a logarithmic rate and there is no way they can sell it all. Frequently they don’t even know exactly what they have to sell. Sure the high quality impressions get sold first, and for a premium price, but that’s the tip of the iceberg and there’s always the massive remnant pool of impressions hanging around unsold. So they have to farm it out to the networks and the number of networks grows again. That’s where the channel conflict that everyone tattoos ad networks with comes from but it’s self inflicted and on purpose. Channel conflict only happens when things are selling. Call it too much selling if you want. Maximum yield trumps any sales force’s headache any day of the week. And maximize they do. Just look at PubMatic’s brilliantly elegant default optimization service. Their ad network optimization automates the reselling of unused inventory back to the network enabling publishers to instantly redirect to the highest paying impression every time allowing them to approach an almost theoretical 100% fill rate. What a wonderful world advertisers and publishers are living in today all care of the ad network explosion. In fact, without the hundreds of ad networks out there delivering the reach, segmenting the audiences, and backfilling the impressions, we wouldn’t have nearly as robust an advertising industry as we enjoy today. The networks might have started out as a necessary evil but they are far from that now. They have matured into the strategic reach partner of any successful advertiser or publisher out there, and they are working–hard.

I just got back from the MediaPost OMMA Publish New York event and must report on one very significant outcome.   There was an outstanding panel discussion called Publishers and Ad Networks: Can This Marriage Be Saved which I was fortunate enough to attend.   The talk was on the value and life expectancy of the ad network model moderated by the very direct but never dull Wenda Harris Millard, President of Martha Stewart Living Omnimedia and author of the now infamous “pork bellies” quote.    On the panel was the similarly outspoken and sometimes abrasive Jim Spanfeller, President and CEO of Forbes.com most recently in the spotlight for his preemptive eulogy for ad networks.   Forget the dripping irony that Wenda sits on the board of an ad network and Jim recently launched his own ad network, these two are regularly in the news with pessimistic views of the ad network’s value to interactive advertising and big brands going online today.  The round table discussion also included Jarvis Coffin, Co-founder and CEO of Burst Media who was incredibly refreshing and sensible, and Matthew Boyd, SVP at ValueClick Media, who together defended the ad network point of view exceedingly well.  It quickly became apparent that ad networks have a solid and necessary place in the value chain within display advertising as Jarvis and Matt, along with Ed Montes, EVP at Havas Digital, talked through the evolution of the ad network and the necessary reach, optimization and aggregation role they fill today.   We saw that the original “horizontal” network model has undergone a Darwinian evolution using things like vertical-ization, contextual technology and behavioral targeting to reach point where they are not just “sopping up” the remnant inventory, but rather adding new and unforeseen value to the big brands that need to be online with their messaging.   I could not resist the Q&A period at the end to put the question to Wenda as directly as possible in hopes of some final punctuation to the argument that has raged in the industry for weeks.  But she took the question with great aplomb and very candidly responded that –based on what was said on the panel–ad networks are more “pearls than pork bellies” to brand value and display advertising.  So I think we can now finally put this issue to bed and get on with driving the industry forward.  Thanks have to go to MediaPost and OMMA for the great venue and the “chutzpa” to host the topic and get the names up on stage.  It was a very satisfying event.

It’s no surprise that the context in which an ad runs has a significant impact on its effectiveness. And that’s especially true for brand advertisers who are going online with their campaigns. I’ve looked at this phenomenon from plenty of different angles around here with posts like Advertisers Say Contextual Offers Best ROI and Contextual Targeting Yields Highest Return for Brand Advertisers. And now to finally put the question to bed, a new study by OTX Research has been written up by Jonathan Lemonnier in AdAge that seems to prove the point conclusively. They found brand recognition could be increased 19% just by running ads in context. By in context, or contextual targeting, they mean running ads on non-endemic content that happens to be relevant to the products or services being advertised. That is not the same as site targeting which is running ads on sites that are inherently relevant due to the overall genre they serve. This is significant to note because it opens doors for advertisers to dramatically increase their reach. No longer must advertisers find vertical endemic sites specific to their industry, they can now reach out to a much larger general audience to find specific content pages talking about related topics and actually drive greater brand recognition. This can even mean dipping back into the remnant pool for high performing inventory if you have the means of categorizing pages to determine true meaning. And that is sort of the rub here; it is not easy to discern meaning from the chaotic web and especially the unmanageable long tail. And when you step outside the well-lit confines of the premium world you immediately face the challenge of brand safety. How do you guarantee your brand will run alongside appropriate content when you are out there in the wild-wild-west of the web’s seedy underside? Real meaning and real preemptive brand safety take granularity. I am always surprised when I look at the contextual targeting solutions on the market today and realize how shallowly they categorize. Everyone seems to offer 20 or 30 top level channels and that makes good sense on the surface. That accurately reflects how advertisers generally see the world. But a high level channel only puts you in the ball park and does not significantly drive up targeting-based effectiveness. Knowing that a page is about Automotive is a good start but that will not significantly impact clicks considering this level of broad channel-ization has already become a commodity within our industry. It is when you can discern manufactures of cars from types of minivans that you really begin to be relevant to the end user. And when you can determine a page is not only Automotive in nature, and not only related to subcompacts, but also that it is about Toyotas, Hybrids and Fuel Efficient Alternatives, and Hydrogen technologies, then you have deep targeting that has a significant impact on relevance and performance. But sadly there are very few systems that can go beyond a rudimentary understanding to determining the page was really about hydrogen fuel cells on environmentally friendly vehicles. Without this level of granularity you are stuck running an Automotive ad or if you are lucky maybe a subcompact ad and settling for average click rates. The real performance comes in when you know the true meaning is environmental in nature and you can target eco-friendly ads that will have a significantly higher recall rate. I’ve seen semantic engines with only a few hundred total categories across only two superficial levels into which they must classify all the eligible content from the more than 20 billion web pages out there. To really understand meaning in the sea of available ad space you need a solution that is far more granular. You need at least thousands, if not tens of thousands of categories if you really want to certify that a page is brand safe with the goal of maximizing effectiveness and subsequently increasing ad revenues, ROI and page yield. This is why LucidMedia has more than 14,000 fine-grained subcategories behind our contextual engine. It allows us to determine the true meaning of a page before an ad is served and makes sure the most relevant ad can be shown to the user. In some of our earliest tests this deep categorization yielded a 76% jump in clicks in the average direct response campaign versus the typical run-of-network buy. When you want relevance on the web for advertising you need depth and breadth. Having just breadth only gets you half way there.

Shira Ovide wrote an interesting piece in the Wall Street Journal’s On Technology section about self-service display advertising. I agree that the display side of the business needs to get easier for it to move to the next level but there is one thing in this article with which I am struggling. I am having trouble with them equating search-based text ads to display advertising because while self-service may work for direct response advertising (DR), they are forgetting brand advertising in this equation. In branding, display is inherently tied to the creatives and that does not lend itself to a self-service model.

I know I disagreed with Spanfeller and Millard in the past, and I still do, but I’ll get to that in a moment. Shira’s analogy is like saying branding on the small screen would grow if it had a self service portal because Google did it with AdWords. That never happened in five decades and trillions of dollars went under that bridge. How could the average advertiser create an engaging 30 second video spot? Not on YouTube. Yes, a guy like Dunn who was mentioned in the article, can use Facebook to upload a photograph of his great pants and take clickers to his e-commerce portal page. He’ll likely hit an acceptable performance-based metric as in their example but the big brand crowd will not have the same luxury.

It is different with search and text because search has the key words provided up front for some minimal relevance and almost anyone with a QWERTY keyboard can produce a decent performing monochromatic text ad in 30 minutes. They can even spit out a decent thumbnail image when pressed. But the performance, and subsequent return, of brand advertising hinges on many factors including the engagement of the creative and, when there is a call to action, the quality of the conversion process (as well as the context in which they are served). The best brands in the world go unnoticed with terrible creatives and the best creatives in the world fail miserably with non-intuitive landing pages.

Basically I am saying that engagement and emotion don’t lend themselves to a self-service model like AdWords. And I’m not talking about Punch The Monkey here, I’m talking about creating real brand affinity, brand recall and purchase intent online with display. This gets us back to what Spanfeller and Millard were recently pontificating. While I still disagree with their prophesy that the ad networks are a dying breed and are devaluing the brands they serve, I agree wholeheartedly with them regarding the need for the human creative element in the process. This is why we have the great agencies we do. Maybe if there was a simple and free global library of real-time customizable rich media creatives in all IAB standard ad spot sizes using text-to-speech technology to create automated professional voice-overs then the big brands would dip more than the current toe-in-the-water online but we are a long way from that happening. It never happened with TV and now the big and little screen ads are evolving into even more protected methods like paid placements and integrated endorsements moving further away from a self-service potential. I think self-service is highly applicable to DR but I don’t think it’s the magic elixir that will pull more big brand dollars online. To do that you need context but I’ve already beaten that drum enough for now.

Forbes.com President and CEO Jim Spanfeller recently delivered a preemptive eulogy for ad networks at Needham & Company’s Third Annual Internet and Digital Media Conference. In doing so he has stoked the already searing flames ignited by Martha Stewart Living Omnimedia Media President Wenda Harris Millard and her infamous “pork bellies” quote that algorithm-driven ad networks are devaluing premium brand advertising. Forget the dripping irony that Forbes themselves have recently announced an ad network of their own, albeit a “slightly different” model, when you peel back the spinning layers of this onion the death certificate seems to be lacking any sort of authoritative signature. Everyone seems happy to relegate remnant inventory to the networks and exchanges, as though it were the red-headed stepchild of the display advertising world, but their panties sure get in a wad quickly when it comes to premium advertising going through the same algorithm-driven systems. You need look no further than Forbes own healthy revenue growth, even in a recession economy, to see that the current trends are working and improving on both remnant and premium inventory for the publisher and for the advertiser. Spanfeller prophesied that the ascendancy of online ad networks is probably at its peak as advertisers and publishers struggle to monetize the growing supply of inventory created by social media exploding online. He better be ready to embrace Google’s new Friend Connect social widget! He bravely predicted that the role of ad networks and ad exchanges will diminish over time as publishers take back more control over how their own media is monetized. Yes publisher want and need editorial control but the data seems to indicate that publishers gravitate to what maximizes yield and not what maximizes control. The question is really whether or not algorithm-driven media buying will have a negative long term impact on media buying. I have to give props to Tameka Kee over at MediaPost and my colleague Paul Levine, Vice President of Marketing over at AdBrite, who countered recently with a great deal of pertinent data which seems to prove the doom sayers are missing the mark with their dark prophesies. The roots of this debate spring from the early wild-wild-west days of the ad networks. Back in the early formative years of the ad network’s puberty it was almost an “anything for revenue” attitude and the trust got broken. Just look at how Spanfeller describes how the Forbes ad network is different from us unwashed masses. He depicts his solution as a “clean, well lit place” for monetizing the rapidly expanding long tail. How is that different from the new breed of algorithm-driven ad marketplaces out there? Take a look at the Certified Ad Space within the Rubicon Project. With the stellar reputation of the Rubicon folks, their approach seems like a cleaner and even better lit ad space. Or take a look at how LucidMedia provides brand safety through the most rigorous contextual scrubbing solution on the market creating a guaranteed Ad Safe Environment for premium brand advertisers. And yes, they all are in part algorithm-driven solutions. But these are good for the premium brand advertisers (and good for the publishers) and are not devaluing the brands they serve. Actually the opposite seems to be happening as these brand safe zones buoy up premium CPM rates wherever they occur. What the Spanfellers and Millards of the world are really saying is not that this is bad for the premium brands; they are actually signaling that the new breed of competition is an unwanted threat to their profit margins. The amount of angst created by the wonderful new breed of algorithm-driven advertising spaces on ad networks and exchanges like AdBrite, PubMatic, Rubicon and LucidMedia is a very clear indication of how positive these solutions are for brands going online. If they were truly bad then the Spanfellers and Millards would not care and subsequently draw this kind of attention to them because it would not be threatening their margins. I say that while the ad networks and exchanges must evolve to survive and continue flourishing–Levine’s comment about reach versus engagement for example–they are working just fine and are here to stay for the foreseeable future.

“In a nutshell” is a term to which I gravitate frequently because concision in marketing and public relations, when it comes to messaging, is one of the most critical factors of success. The proverbial sound bite. So here is another “in a nutshell” about how to effectively build brands with online interactive advertising as a significant component of your overall strategy: put it in context. Advertising is all about using emotion to create value and ultimately drive commerce. As Ms. Wenda Harris Millard puts it in the recent AdAge article (which got me thinking about this subject) regarding her now infamous pork bellies comment; “If you think about what does advertising do, it creates desire and eventually causes transactions.” Very well stated. She was providing new insights into what was founding her comment. She points out that ad networks and exchanges who are simply aggregating a critical mass of inventory without considering the context are basically just creating volume which will drive down prices over time. But to me, the future of online brand advertising has it’s cart inherently hitched to value, emotion and desire and that is not found in pure volumes. The only way we can address that aspect of advertising online with the technology that is readily available today, especially when you take into account the now inhuman scale of the web, is through advertising in context. While several types of targeting might work well enough for direct response, when you really want to elicit an emotional reaction for a brand you need to do it in context. When your online advertising approach works in concert with your branding campaign goals instead of working for your campaign, that is when you begin to squeeze the maximum ROI from your online advertising dollar.

MediaPost recently wrote about Marketing Sherpa’s inaugural 2008 Online Advertising Handbook which showed that less than half of their advertisers use online display ads for branding purposes. I was happy to see that advertisers rated the ability to use behavioral and contextual targeting as an important aspect to ROI measurements though. InsightExpress reported that targeting was a key driver in effectiveness and advised advertisers that the context in which an ad is served is just as important as the ad itself. It comes as no surprise to me that context is important and targeting impacts effectiveness and ultimately ROI. What we need now is to take this a step further and understand which types of targeting work best. This is important because I have noticed that vendors in this arena tend to muddy the waters around targeting and in the end confuse the advertiser and their agencies. So I thought I would shed a little light on the difference between behavioral and contextual mentioned in the study. There seems to exist an almost unnecessary tension between the two different methods of targeting in the marketplace. I say unnecessary because when you compare the two, it is important to note that behavioral is to some degree dependent upon a contextual element; it is in part “contextual over time”, but advertisers still see it more as a black-and-white, one-or-the-other, which-one-do-I-choose situation. So it is worth investigating further. And to make matters worse I realized from this year’s Ad:Tech in San Francisco that many networks are claiming to do a mix of both but in reality they do very little contextual. No wonder there is confusion in the marketplace. Just take a look at two quotes from the industry press over the last few years.  “The CPM of behavioral targeting was 24 percent less than the contextual placement, yet it delivered 50.3 percent more imminent purchasers … Therefore the CPM against imminent buyers was 50.6 percent of the CPM of contextual targeting. Behavioral targeting was twice as cost effective.” This is from a case study in TACODA’s Behavioral vs. Contextual targeting research done in 2006. Now here is another more recent quote that seems to almost contradict their findings.  “The last two placements showed about a 19% lift in brand recall over the former, proving that when it comes to online ads, contextual targeting can have an effect. According to Marketing Sherpa data, 40.5% of marketers said contextual targeting delivers good return on investment; behavioral targeting was not far behind at 36.7%.” This is from Jonathan Lemonnier in Advertising Age early 2008. Ok, so is behavioral twice as cost effective as contextual or is contextual more effective? How are advertisers supposed to sort this out? To find the reality you have to compare how the TACODA case study was executed with the actual data in the research paper. If you look at the actual data in the TACODA study cited in the impressive first quote, the actual results are far less remarkable. The study is measuring the right KPI, cost effectiveness based on purchases, but the data in the study does not necessarily support the claim that behavioral was twice as cost effective. The case study assumes a certain CPM paid for both behavioral and contextual impressions and that is where the case study separates from the research data. If you look at just the research paper and don’t take into account the current market value of behavioral and contextual impressions, the subjective part of the claim, the empirical data seems to say that contextual targeting outperforms behavioral especially under a frequency cap!

(Number of looks at an ad)

(Seconds spent looking at an ad)

These images are from the actual study. Researchers measured the number of times the subjects looked at each ad on each page (looks, the first chart) and measured aggregate time spent looking at each ad on each page (seconds, the second chart). In the two charts you notice that contextual targeting outperforms behavioral until you add frequency. Only then does behavioral overtake contextual. In fact, on first exposure, contextual well outperforms behavioral for initial looks and ends up with almost similar look results at the second exposure. The story is even better when you look at seconds spent looking at a particular ad. Contextual well outperforms behavioral in seconds spent looking at an ad on the first through third exposure and it is only as you approach the fourth exposure, which is one reason why we frequency cap, where behavioral clearly begins to outperform contextual. Behavioral definitely has it merits. But this study, used to support a behavioral approach, shows why contextual targeting yields the highest returns for brand campaigns (as in the more recent Advertising Age quote). Contextual ads get eyeballs faster and they keep them looking longer especially under a frequency cap of four. In fact, brand advertisers should consider a frequency cap of three in light of this study. This is why big brand advertisers are doing so well on the LucidMedia Network and why they continue to be a strong focus for us. Contextual gets eyeballs faster and keeps attention longer in today’s world of hyper-short attention spans and overcrowded messages. The moral of the TACODA study should be; “Just don’t over pay for your contextual impressions.” Luckily for advertisers and agencies we are making the deepest contextual solution one of the most economical forms of targeting.