Nailing Down a Moving Target
"Last year it was search-enabled marketing, before that, customer relationship management, and before that it was Internet-enabled sales. BT is the next thing," says Dave Morgan, CEO at Tacoda Systems, a provider of behavioral targeting services.
The discipline of BT figures out who customers are by how they behave online. By building a better profile, marketing departments can build more effective ad campaigns. And marketers can improve their pitches to consumers no matter where they go on the Web. Publishers squeeze more dollars out of more inventory, and consumers get relevant ads for products and services they may actually want.
Goodbye, scattershot advertising. Hello, cost-effective, tailored pitches. But as attractive as it sounds, BT hasn't defied the laws of Internet gravity just yet.
Media buyers worry that the method is more Web publisher-oriented than marketer-focused, and that it's experiencing growing pains. In addition, privacy concerns, both corporate and personal, threaten to slow BT's growth.
No one is sure how quickly behavioral marketing will go mainstream. But Greg Smith, executive vice president of planning and analysis at Carat Fusion, says he thinks the new advertising has big-time potential. "Trends begin as fads," he says. "Then everybody gets the fever."
BT mathematically develops customer leads by tracking Internet usage with bits of software, usually cookies, hidden in Web browsers.
These cookies tell the story of where Web surfers go and what they see and do. By harnessing statistical mathematics, a marketer can build a remarkably specific profile of what users want and what they might buy. Marketers can infer all sorts of things: what cars people might drive, shoes they prefer to wear, and places they would like to visit.
For example, Ford could be promoting its 2006 F-150 with big buys on Cars.com, AOL, or Yahoo. But by using BT, Ford can suss out a potential F-150 buyer and market to him at other sites he might visit. For trucks, let's say that's uhaul.com. Ford then buys on uhaul.com, promotes the truck, and, if things go well, the behaviorally targeted buyer clicks through to the Ford spot, becomes a decent prospect, and everybody gets paid. "I see BT as ultimately oriented to awareness campaigns and relevant reach," says Alan Schanzer, managing partner, MEC Interaction.
The magic of BT is that Ford can market at a great discount--whispered figures put the savings at about 30 percent over prime sites--while uhaul.com gets more cash for its inventory (rumored to be about four times basic run-of-network sales). And consumers get an ad better targeted to their interests.
"Everyone is winning. Consumers get a more relevant ad, marketers get more targeted reach that is cost-effective, and publishers get more money for their inventory," says Omar Tawakol, senior vice president at Revenue Science, a provider of BT services.
The idea is catching on. Revenue Science counts The Wall Street Journal Online, ESPN.com, FT.com, and Reuters among its clients. And Tacoda recently announced adding the 3,000th Web site to its network. The Tacoda network, which includes AccuWeather.com and Gawker.com, now reaches 190 million unique users.
Yahoo is rolling out a behavioral targeting system called Consumer Direct. AOL is reportedly launching a similar service. And San Francisco-based Jumpstart Automotive Media will debut a network for the auto market later this spring.
Microsoft Corp. is also in the game with its new adCenter Incubation Lab (adLab), a 50-person research lab aimed at developing next-generation advertising models. David Jakubowski, general manager of search strategy at adLabs, declined to comment specifically on Microsoft's behavioral strategy, but he did say, "We believe software innovation will drive the future of ad innovation."
Despite the hype and the hope that BT can be the silver bullet for advertisers, the practice faces some major issues that may limit its growth in the short term and beyond.
BT's problems start with its past. Many advertisers argue that it was designed and built by Web publishers, for Web publishers. Indeed, the funders of the early BT startups were large publishers who were losing too much inventory to inexpensive run-of-the-mill ad networks. These content providers were focused on finding a way to turn low-performing ads into fresh dollars.
"Back in 2001, [the publishers] were the only people who would write checks," says Tacoda's Morgan. "They needed to make more money."
BT works, by all accounts, but it can also promise more than it delivers. Buyers often complain that the behavioral ad companies do little more than wave the magic BT wand to recast a publisher's unsold inventory as "targeted," then sell the same banner at four times the markup. "It's sort of a square peg for a round hole," says Amy Auerbach, vice president at Media Contacts, the interactive arm of Havas' MPG.
One buyer, who requested anonymity, drew an analogy from the earliest days of direct marketing. Thirty years ago, he says, buyers could find out from subscription data who was getting home delivery of, say, The Wall Street Journal, The Financial Times, and a local classified newspaper such as The PennySaver. Often it was the same person, but PennySaver could never charge the same ad fees as The Wall Street Journal, even though it was being read by the same subscriber. "It's the same for behavioral. Even though we are seeing some measurably higher results," the buyer says, "It comes to a point very quickly where this inventory is crap."
Can't Get on the Scale
But publishers vehemently disagree with the notion that they are peddling dead stock.
"That 'missing value' argument misses the point," says Jack Smith, vice president of product strategy for 24/7 Real Media. "If [BT] did not work and the return on investment were not there, marketers would not pay for it."
BT also bumps up against problems of scale and diminishing returns. The costs add up quickly: fees include publishers' ad servers, analytic system costs, targeting system costs, buying fees, ad network fees, agency fees, and creative costs. The expenses can turn even the most basic behavioral campaign into a complicated and expensive endeavor.
"From the advertiser's side, results can be very mixed," MPG's Auerbach notes. "Return on investment does not increase with cost, so it does not directly translate into performance."
Auerbach maintains that despite BT's advanced mathematics, the process remains riddled with trial and error, which means that money can easily get wasted. Many feel that the only way advertisers can protect themselves is by targeting across very large audiences.
"Unless there are lots of consumers, lots of data, and lots of different types of offerings, it's hard to get the components to work together," Tacoda's Morgan concludes. "And when we get them, we'd better not waste any LowerMyBills [types of] ads on them."
Some startups are trying to create mass audiences without the larger sites by joining lots of smaller Web properties. For example, Jumpstart Automotive Media will debut an auto BT network this spring, taking a smaller-publisher banner ad inventory and its cookie information and matching that data to potential customers. "We are building what amounts to a private label network of partnerships looking for scale," says Joe Kyriakoza, vice president of product development at Jumpstart.
Skeptical buyers still wonder.
"Great idea," notes Robert Davidman, CEO at EarthQuake Media, a cross-advertising optimization company. "But how much more value will I get over a simple, run-of-network buy? It's so easy to get stuck with bad inventory."
Who Let the Ideas Out?
Then there's privacy. Skeptical advertisers say that BT vendors are placing their pixels all over the place, and that soon enough questions will arise about what kind of data from those pixels can be shared.
Remember all those companies that needed to be hired to mount a BT campaign: the publisher, the publisher's ad server, etc.? Each company is probably placing a cookie somewhere near the customer. These cookies amount to a huge trail of digital breadcrumbs leading the client, and the client's competitors, back to the customer.
"If Chrysler runs advertising on 30 sites for $100,000, those 30 sites have an obligation not to use the money spent against [Chrysler]," says Bill Harvey, CEO of Next Century Media, a Gardiner, New York-based market research firm. "But sometimes that happens."
Harvey maintains that the industry needs firm rules: full disclosure, no personally identifiable information without permission, and no data from one advertiser used to aid a competitor.
Privacy issues are a burden. Ad executives privately complain about the privacy addenda and non-compete clauses they're forced to sign just to start the business process. And even those on the BT side of the industry acknowledge there's a problem.
Tacoda's Morgan says there will be tension when advertisers realize behavioral marketing companies are making millions by selling marketer data to someone else.
You Call This Creative?
Probably the toughest question facing BT is the creative issue. Can the message keep up with the technology? As good as the Web is at finding customers, advertisers still aren't sure what to say to those customers once they find them.
"This is what the industry really sucks at," Carat Fusion's Smith says. "Are we stuck with the same dumb elevator pitch? Where is the new creative that can convert the targeted lead?"
Smith thinks the online industry has enough trouble with the basics, much less designing and customizing ads on the fly to serve millions of disparate users who are behaving in wacky and immeasurable ways. "I know it would be better if I had custom creative for behavioral spots, but that's tough to get these days," MPG's Auberbach adds.
The issues now facing BT now are those presented by a world where the Internet is limited to personal computers. But as gadgets proliferate--cell phones, personal digital assistants, wireless Ethernet devices, advanced mobile entertainment--BT will face even bigger hurdles.
"The silver bullets don't come that often and that quickly. Google and search were one," Morgan says. "I do not expect BT to be like search, and to expect so is ridiculous. It's a great product for what it does."