Online advertising is an interesting modern parallel, a commodity of nearly limitless availability, generating (occasionally) vast revenues for clever manipulators of access and pricing.

It is not due to the genius of any single player (well, there is Google, more on that below) but a combination of fragmented access and the diabolical conceit of inventory based pricing (read: CPM).

The fact is, there is just a ridiculous amount of online ad inventory, like, choose your own expletive based measurement unit – mine is fuckton. There are a fuckton of impressions available. And, if in some magical way, you could get unfiltered access, there are just so many, the price would approach zero.

The curious thing is that the producers of online advertising inventory, almost without exception, base their businesses on the creation of more not less inventory. The poster child being HuffPo. When CPM is the currency of the realm, who can blame them.

When advertisers begin to demand audience not imps, well, most players are not prepared.

The ad tech companies get it and are attempting to perfect their respective blackboxes in order to extract value from this ocean of imps. But publishers and ad networks are a different story. Their business models are based on selling impression volume.

And of course, there is Google. With the move into network display and rumors of a data-exchange, they are on a trajectory to become online’s De Beers.

What these ad tech companies all seem to have in common? They often some version of a better calculator bolted to some serious tech gee whizzery. They all promise more accurate metrics and some sort of optimization, usually price or audience based or both!

(Curious that an industry supposed to be innately quantifiable can support so many companies who essentially count things.)

A charitable take: There is a distinctly 1950s Popular Mechanics-style optimism to the industry, accelerated by the wide-eyed boosterism of blogs like adexchanger and iMediaconnection.  The prediction, a friction free utopia. When the technologies get perfected, they tell us, online advertising will run itself, a perpetual motion machine of algorithmic audience targeting, real-time-bidding and performance optimization.

Overly optimistic but basically harmless.

A less charitable take: The industry is rife with hucksterizing, black box peddlers. There are persistent rumors that what is sold as algorithmic is occasionally coders optimizing manually. But this hearsay aside, the too-good-to-be-true sales pitches have more than just an echo of film flam. This scene from from the The Grifters captures it perfectly:

In January 2010, for the first time in 23 years, Pepsi cancelled its annual Super Bowl advertising and joined the social media marketing hysteria by launching their much-hyped “Pepsi Refresh Project“. In Super Bowl ads from 1999 to 2009, Pepsi spent over $142 million to encourage consumers to drink the Pepsi brand. The company was the second-biggest spender on the event behind Anheuser-Busch InBev. So abandoning its commitment to legacy advertising in favor of a hip, new and expensive social media program was considered a bold move, a real game-changer.

Pepsi endorsed one particular internet channel and it was social. Users would be ”engaged”, people would flock to ”like” Pepsi over Coke, and twitterati were recruited to tout the virtues of Pepsico products by hashtagging the shit out of #DietPepsiRocks. It all sounded so nice and progressive, like putting manure in your gastank.

Classic social networks are like echo chambers populated by self-proclaimed experts in this field who happen to be, as Peter Shankman recently pointed out, often young idealists or deluded zealots without any business sense. But Pepsi saw the light. No more TV in 2010 for their flagship brand in the fight for soft-drink supremacy against archrival Coca Cola. A massive $20m dollars we allocated to promote this pure-play social campaign, boost sales and grow market share in the US. Millions more were deployed to support the initiative. So Twitter’s servers were lubed up. Zuckerman was überconfident. Social media was pregnant with anticipation and promise and the ad industry held its breath. Pepsi was geared up to make a splash.

But did the gamble pay off? No, it did not. For the first time ever, Pepsi-Cola’s trademark brand dropped from its perennial second place ranking among American soft drinks to number three. Actual campaign data now shows what we few, proud, zig-when-they-zag internet marketeers foresaw a year ago: social media is a dawg yo. Behold the numbers. The Refresh Project achieved remarkable success and did pretty much everything a social media program is expected to: Some 80 million votes were registered, a whopping 5.5 million “likes” on the Pepsi Facebook page. More than 120,000 twitter followers. The only thing it failed to do was sell Pepsi. Say what? Yes, in other words, it was an impressive performance of emply clanging noise, a carnival of cymbals celebrating the charlatan king of social media. The company benched advertising for its beverages in 2010 as it pursued the “Pepsi Refresh Project”, but since that time, its flagship brand has been surpassed by Diet Coke in terms of share of the U.S. soft-drink market. Pepsi junked online social media and quietly returned its beverages to the 2011 Super Bowl telecast. Massimo d’Amore, chief executive of PepsiCo Beverages Americas confirmed the fiasco and called for a change in their marketing strategy: “We need television to make the big, bold statement…”

So Pepsi paid $20m dollars to find out that social media is like porn—a fun yet senseless distraction from real work. Not to mention a poor substitute for proper marketing. Lessons learned? Coke is it.

Related articles

• Pepsi Takes Active Role in ‘X Factor’ (nytimes.com)
• Pepsi: There’s Madness To The Madness (adcontrarian.blogspot.com)
• Pepsi Ads Return to TV (adweek.com)

Doesn’t the growing third party data segment mean that online publishers have given up on audience?

In the old days, you bought ads in GQ because you assumed readers of the magazine fit the segment you wanted. GQ got paid a premium because they aggregated that audience.

But the trend in online is to use third party data to slice out audiences from different pools of traffic. Information gathered about users’ web habits, search history, their gender etc. can be used to make assumptions about the kind of ads to serve, carving out a segment independent of the publishers audience.

“Thanks New York Times but I will go find my own users on your website.”

If advertisers feel they need these data suppliers to truly aggregate the audience they want, this begs the question: what value do publishers bring beyond raw ad impressions?

Of course the assumption is that all the data and attendant blackbox targeting actually works.

But assuming it does perform as advertised – no one should be buying premium.

If publishers were doing their job, there should not be a serious market for third party data.

(Our suspicion didn’t begin there but Augustine Fou had a great post a few years ago comparing behavioral targeting to belief in Santa.)

We recently moved even closer to being convinced BT is mostly a con when we learned, in a chance airport meeting, that a well known company requires premium ad buys in its network to include a certain percentage of behaviorally targeted traffic.

There could hardly be better evidence that behavioral targeting’s effectiveness is questionable than having companies bundle it with premium buys.

(Film studios pull a similar move with movie rental companies. Want to buy Inception and True Grit? No problem. You just need to take Revenge of Crap II: The Crappening and ten other titles that are only making it to DVD because we can make you buy them.)

Theoretically, behavioral targeting makes sense, its segmentation 2.0 -
aggregate some data points, makes some assumptions about the user. But is always just that – assumptions based on historical data and there is simply too much noise in the channel.

Following the airport meeting, we have a new theory. Behavioral targeting is being used as an online media philosophers stone. The media alchemists are using it to turn remnant traffic lead into premium traffic gold. Take your unsalable inventory, wave the BT magic wand over it and  – presto – premium traffic.

(Wall Street has their own version of this. It is called a Collateralized Debt Obligation)

Alchemy’s success in perfecting transmutation is well known. There is still a lot of lead out there.

A variant of a familiar aphorism is apt when it comes to remnant traffic: A hog in a silk waistcoat is still a hog.

“We need to talk to you. We need to understand your expertise, your hopes … We need to hear your aspirations, your needs. You need to hear our limits, our red lines.”

We do? Web entrepreneurs and digital company founders asked themselves in bewilderment. We’re all about breaking rules, pushing bounderies, distrupting stuff. “Creative destruction is what we do”, said one participant.

The ensuing twitterstream was as interesting and amusing as Sean Parker’s bling-alicious after-party at swanky Paris night club Matignon. Synthesio made a useful infographic about the topics & tone of the even.

It was a fascinating two days in sun-drenched Paris. We were privileged to rub shoulders with the BBC (billionaires’ boys club), senior government officials and media tycoons such as Murdoch, Zennström and Zuckerman whose empires collide online and off with increasing frequency.

Unfortunately, as  professor Larry Lessig pointed out, the French president’s desire to rein in the web is hopelessly misguided. Google’s exec chairman agreed by saying that policymakers should tread lightly and avoid “stupid” rules. Stupid being defined here, presumably, as boneheaded policies and official panic. The battle lines are being drawn.

Conflicting visions of the Internet – notably about how regulated it should be – are pitting companies like Amazon and Google against governments when the subject is privacy and copyright online.

The impact on web advertising will be enormous. How enormous? Really really big. Too big to quantify just yet. (As we publish this, it is May 26 +3)

But until governments understand what the Interweb is really about, they should refrain from regulating it. Wittgenstein springs to mind:

”Wovon man nicht sprechen kann, darüber muss man schweigen.”

Freely translated: Dude, if you dont get it, just STFU.

Bonus Video: Sean Parker demands Duck Confit!

Pick a random ad network site and look for these tabs:

Publishers
Advertisers

Click the publisher tab and you read some version of:

We turbo-charge your eCPM! Premium advertisers paying premium prices… blah blah blah.

Click on the advertisers tab and you read some version of:

We will save you money! Efficiency. Optimization. Targeting. Or maybe the coded version of same: “We deliver ROI”.

In Kubrick’s outstanding contribution to the Vietnam War genre, Full Metal Jacket, Matthew Modine‘s character, Private Joker, is dressed down for having a peace symbol on his body armor and the words “Born To Kill” written on his helmet. “What’s that supposed to be, some kinda sick joke”? a colonel asks him.

Honestly, how can you represent both interests simultaneously?

The big holding company media agencies diversified so they could serve competing clients. So it is interesting this  curious nexus of publisher and advertiser is not just tolerated, it is industry standard.

It leaves us asking the same question the colonel asks Private Joker, “Whose side are you on son”?

Well, it is a hardball world, maybe we all have to keep our heads till this ad network craze blows over.

Television, as your parents probably let you know with irritating frequency, is a passive medium. The remote, sealing the deal when it removed TVs one interactive feature, changing channels. And that was in the mid-1950s.

Television is also, almost exclusively, an entertainment medium. Sure there are 24 hour ‘news’ channels but these beep and flash like the Vegas Strip and primarily deal with trivialities, titillation and murder.

The Web is something entirely different. But maybe because we view it through a screen, we conceive of it as being in the same as TV. The web can be like TV, Hulu, for example. It can be like print, NYT, Le Monde, or The Guardian, for example. It can also be like radio, Pandora, LastFM etc. But ‘like’ is the operative word. Because even when it is imitating those things, the Web is still not those things.

On the Web, Television’s interruption model is mostly useless. On the Web, people are directed, seeking. They do not want to be ‘interrupted’. Plus, as Clay Shirky correctly points out, the web is simply more cognitively demanding then TV.

Marry goal-oriented to high-involvement and getting our attention in the so-called ‘attention economy‘ is way harder on the Web than on TV.

This is what drives a hackneyed lament like “The CTRs are falling. The CTRs are falling.”

It is also driving Web video. But when TV is the model, you get a 20 second ad before a 30 second video.
But the Web is a great place to advertise. It is just that it is a baby and media and advertisers are like confused first-time parents. They had a dog before. It was easy enough to care for so they thought – “Hey, lets have a kid!” But a child is vastly more difficult to care for than a canine. (And yes, in this clumsy analogy, TV is the dog.)

We can’t keep porting old media concepts and expectations to the Web. It is not a single channel. And it is not experienced like TV.

Given that Sweden (and just after them, Denmark) tops The Economist’s list of digital ready economies, the report is an amazing glimpse into the future. Too bad it is only available in Swedish
(UPDATE: Our Swedish is even worse than we thought. The report is available in English as a PDF right here.)

A graph from the chapter titled: Growing up with the internet  jumped out at us. It compares the percentage of children age 3-7 who use the internet periodically in 2009 & 2010.

Source: Svenskarna och Internet 2010

Another interesting graph that illustrates a subject close to our GlibHippo heart – Frequency of online gaming by age group. (We think this category also includes watching video but our Swedish is not that good and the description is not that clear)

1. Grey = Occaisional
2. Light Red = 1-3 times a week
3. Dark Red = Daily

Check out The Economist for a summary of in English. But browse this report using Google translate for a more in depth picture of what tomorrow will look like wherever you live.

Adage reports, CBS Corp. Chief Research Officer David Poltrack, saying:

A growing amount of data that matches audience measurement with purchase information shows that using demographics to target commercials is “essentially invalid,” he said, “resulting in a misallocation of television advertising investments.”

Finally. Now will someone tell the media agencies?

And when it comes to online, we would like to add — our 100% objective anecdotal evidence is that this rejection of legacy segmentation should be even more far reaching.

As we have said before and will continue to point out – classic segmentation on the Web is largely a waste of time. All old-school segmentation in the age of the internet is overly simplistic. It makes advertisers and agencies feel comfortable but it has nothing to do with consumer media habits.

Of course, you craft messages for a target audience but in a world of 1000s of niches and fragmenting media habits, segmentation like your grand father Don Draper knew it is over.

If we had a euro for every time we have were asked for specific verticals independent from a campaign’s goal… well, you know we’d be making it rain €€€. A campaign for a well-know console driving game title, for example, that could only be run on automobile sites. Airlines that only want to advertise on ‘travel’ sites etc. Segmentation as we knew it is a dead heuristic.

Of course, performance advertisers have known for a long time that their customers are everywhere. Maybe now brands will get a clue. And it will be crazy if we have a legacy TV network to thank for the change.

Related articles

• CBS Says Age and Sex Don’t Matter (marketingpilgrim.com)
• CBS Wants To Kill Demographics, Marketers Should Help (rohitbhargava.com)
• Jack Myers: Legacy Media’s Slice of the Social Commerce Economy (huffingtonpost.com)