Are banner ads really dead?

The most effective ad campaigns have clear goals. Surprise, advertising on the Web is no different. Five crucial keys to banner ads that worked.

The sky might as well be falling. Or is that your bottom line?

You paid a fortune to plaster banner ads across the Web — and no one clicked. Flipping ahead a few pages into your company's future, you see that Chapter 11 comes next. And the sites selling ad space are at the same point in the book.

Since 1997, the percentage of people who click on Internet banner ads — known as the click-through rate — has sunk from a little more than 1 percent to an anemic 0.4 percent. For buyers and sellers of online ads, many of whom were hammered in last April's market correction, it would seem the end is near.

Yet among the world's financial fortune tellers, online advertising looks healthy and full of promise. Jupiter Research projects that online ad revenues will reach $16.5 billion worldwide by 2005, up from $3.5 billion in 1999. Investment banking firm Lehman Brothers offers an even more optimistic assessment: $32.3 billion by 2005.

How then to explain the grim news from research firm AdZone Interactive, that online advertising declined 7.6 percent from July to August of last year? Consider another Jupiter prediction: Consumers will pay less attention to online ads in 2005, when they will be presented with 950 ad impressions per day, up from 440 in 1999. And what about the disappearing stock prices of online ad companies DoubleClick, Engage, and 24/7 Media? Worried yet?

Since the banner ad's debut on HotWired in October 1994, it has been a disappointment. For cash-starved dot-coms that have fallen by the wayside, it appears that online advertising never delivered. Of course, blaming the banner is easier than taking responsibility for a dismal business plan, myopic marketing, or lavish spending. The Web was supposed to launch 1,000 companies, financed by ad revenues and e-commerce.

But even Yahoo, the granddaddy of ad-dependent sites, saw its stock plummet by 80 percent in the first 10 months of 2000. It's clear that ad-supported content sites have suffered. Consider the fate of highly regarded crime news site APBnews.com, which laid off all its employees in June when the company failed to secure further funding. Or of Salon.com, which laid off 25 workers last year — about one-fifth of its workforce. With its stock languishing at around 50 cents per share, down from its $10-and-change high, Salon's future — and that of many ad-dependent sites — looks bleak.

Though Stephen Reed, senior vice president of sales at Salon.com, dismisses the notion that banners have failed and insists that online ad sales comprise a good portion of his company's revenue, he concedes that "click-throughs have been declining, so everybody's got to look at ways to make banners more compelling."

Bruce Judson, faculty fellow at the Yale School of Management and the author of HyperWars (Scribner, 1999) and Net Marketing (Wolff New Media, 1996), maintains that advertising alone isn't enough to sustain a business. "I have seen hundreds of business plans and I have never seen one where a site that creates original content is profitable based on advertising alone."

For sites selling both ad space and products, the numbers still don't add up. Even Yahoo, a successful company by any measure despite the humbling of its stock price, appears skittish about its future ad revenues. In its November 2000 quarterly report, the company warned, "We rely heavily on revenues derived from Internet advertising, which may prove to be an ineffective means of advertising for our clients." This despite a 105 percent increase in Net ad revenue between the first nine months of 1999 and the same period in 2000.

Though such caution is standard practice for corporate filings, a possible risk for Yahoo represents a grave threat to smaller companies. Could it be that after years of declining health, the banner has finally kicked the bucket?

Banner years

Surprisingly, online advertising experts say no. Leslie Laredo, president of the Laredo Group, a research, training, and consulting firm dedicated to Internet-based advertising, says she doesn't believe for a second that the banner is dead. "When people ask, 'Is the banner ad dead,' what are they talking about? Is it effective in delivering a message? Yes. There've been studies done by companies like Engage, ASI, Dynamic Logic, and others showing that viewing banner ads increases brand awareness. Is that measurable through clicks? No, it is not. Is the banner ad dead as a direct marketing tool? No, it just needs to be understood as a direct marketing tool, which I don't believe it is right now. The other issue is that most of the banner creative is terrible."

David Zinman, senior vice president of product marketing at Engage Media Services, says he hears people ask if the banner is dead at every conference he attends. "Banner advertising as defined by the people that ask the question is probably not very healthy, because [they're talking about] a static banner where you don't target it, you don't track it, and you're just waiting for people to click on it. Nobody should be doing that kind of advertising because it is not very effective."

He says banner ads can be done well, especially with rich media, and that it has real advantages over traditional ad media like print and TV. "You can do advanced targeting to reach exactly the people you want to reach," he says. "You can track exactly how many sales you get as opposed to how many clicks you get on it. You can determine which ads are working based on conversion instead of just letting an ad sit up there that doesn't generate any results for you. All of those things make the Web the best medium for advertising today." Why? "Because it's very inexpensive, and it is extremely accountable." The most effective ad campaigns have clear goals. Surprise, advertising on the Web is no different.

Effective advertising in any medium requires clear goals. "You want to be thinking about what you are trying to do with your ad," says author Judson. "Are you trying to create brand awareness? Are you trying to sell something? Are you trying to bring people to your site?"

It may sound obvious, but the distinction between ad campaigns designed to increase brand awareness and those designed for direct marketing is crucial. As Charlie Buchwalter, vice president of media research at Ad Relevance explains, click-through rates may be shrinking, but that doesn't mean banner ads don't work. While some in the ad industry continue to believe that traditional media offer better vehicles for branding — television, for example, is said to be more emotionally engaging — more companies are looking to the Internet to refine the perception of their brands.

David Steinberger, vice president of sales at Shockwave.com, points to Compaq's sponsorship of Tim Burton's animated Web show Stainboy as an example of effective online branding. In the week after its launch, he says the show — playing beneath Compaq's crimson banner — received more than 1 million page views. "The click-through rate isn't great, but everyone who's been on our site has seen Compaq multiple times, and that's what [Compaq is] looking for," Steinberger says.

Ranked 35th in the world for traffic during August 2000, according to Jupiter Media Metrix, Shockwave.com hardly represents the typical Web site. But the site's huge user base — 7.4 million unique visitors a month at last count — drawn by a mix of free online games, music, and animation, suggests that the Web may soon rival cable television as a mass advertising medium.

That kind of audience has enabled Shockwave.com to sign big-name sponsors as underwriters of its content. Companies like Altoids, Hewlett-Packard, IBM, Nike, Sony, and Visa have purchased sponsorships, some costing more than $50,000 for a six-week period. Jack Daniel's currently pays to have its logo on a virtual pool table in the Real Pool game.

According to Chris Kantrowitz, CEO of The Groove Alliance (maker of the 3D gaming technology behind Shockwave's Real Pool game), the Jack Daniel's logo is viewed 45 million minutes a month. "That's like buying [an ad] on an episode of Friends," he says. Turns out, click-through rates "may not be as important as advertisers thought. When you're watching a commercial on TV, there are no click-throughs. It is just a branding exercise."

Says Steinberger, "The lesson here is that branding on the Web is a very, very viable opportunity, just not through standard means. For the longest time, banners were really the only option out there and everyone was saying the Web was about direct response [and that] branding on the Web doesn't work. I would like to see someone raise the argument that playing Real Pool with the Jack Daniel's logo on the table is not effective branding." Following customers around with cookies may not be enough to get results. You also need to anticipate their next moves.

While advertising networks offer many different ways to serve ads based on past behavior, another approach - and one less likely to get you in trouble with privacy advocates - is to deliver ads that complement adjacent content. Merchandising Avenue CEO Phil Trubey believes that his company's service, which uses human merchandisers alongside real-time optimization technology to display and promote key products in context, delivers superior results. "Only 0.4 percent of people who see a banner ad actually click on it. The click-through rates we're finding [at company client SheNetworks] vary week by week from 6 percent to 12 percent."

Trubey contends that targeting ads based on past behavior doesn't necessarily reflect current interests. For example, someone who just purchased a house might suddenly start visiting landscaping-related sites. But ads delivered based on previously visited Web sites wouldn't necessarily include landscaping products or services.

Perhaps unsurprisingly given the nature of his company, Trubey disagrees with those who contend that Internet advertising is best for branding. "The future of banner advertising is direct marketing," he says. "You need to put more intelligence in the system because the creative — the actual display of the banner ad — just isn't compelling by itself."

Another way to put more intelligence into the system is to work with ad campaign management companies like Net Perceptions, E.piphany, or BroadVision. Steve Van Tassel, vice president of commerce solutions at Net Perceptions, stresses that companies need to take advantage of the information they gather in the marketing process. "Be sure to capture the information and then close the loop," he says. "Set it up in a way in which you learn from the experience." For Web marketers, it's not how many people click, it's what they do after they click that counts.

For Steve Myers, media director at Miller/Huber Relationship Marketing, there's more to marketing than clicks. "I couldn't care less how many people click on my banner," he says. "I want to know what they do after they arrive at my client's site. Are they filling out the registration information? Are they purchasing? Are they coming back again? That is the information we're really, really interested in. I'd be happy to get 0.1 percent click rates all day long as long as I know I'm generating an efficient sale or an efficient registration or multiple page views for my clients."

Such online metrics are the forte of advertising networks like Double Click, Engage, 24/7 Media, and dozens of similar companies, each with its own specialties and methods. These companies contend that they help businesses know their customers, deliver personally tailored ads, and measure the ads' effectiveness. "The Web will die without targeted advertising," says Engage's Zinman. "Without that ability, the Web becomes like a direct-mail model. You kill a bunch of trees, hire a bunch of postmen, and hope that somebody opens the envelope."

Zinman cites a service his firm launched last fall called Engage Echo. "The idea is to turn banner advertising into a retention medium, instead of just serving ads to an audience that you purchase based on the sites they are going to, or based on a profile that we develop." The service works by tracking visitors to your site who match a certain criteria, such as having made a purchase, with a cookie (a file deposited on Web browsers' hard disks that records — usually anonymously — where visitors go and what they do online). Engage then remarkets to these customers by serving them targeted ads the next time they visit your site.

Before the Net, this sort of highly focused marketing was nearly impossible. "Advertising has always been by its very nature reaching people who don't necessarily know anything about your product," Zinman says. "[But] once you acquire customers, you're not done. You have to retain them."

Advertisers' failure to insist on ad tracking has tarnished the banner's reputation, says Laredo. "People don't know how to buy [banners], don't know how to analyze, don't track post-click. This whole banner-is-dead issue is so overblown because we're not even close to really being good at this business. It is basically a four-, maybe five-year-old business at the most, and the truth is that only in the last two years have people really started to understand it."

Albert Lopez, CEO of advertising network AdFlight, stakes his business on post-click tracking. His company offers a way for advertisers large and small to automatically optimize ad content, dynamically replacing underperforming ads (banners or rich media) with more effective ones as a campaign progresses, rather than after the fact. Like Myers and others in the online advertising industry, Lopez argues that an ad's conversion rate — how many people take the action desired by the advertiser, whether it's providing registration information or buying a product — is far more important than its click-through rate. "If you look at banners that get the best click-through rates," he says, "a lot of those clicks come through trickery and they don't convert. Advertisers want acquisition, not visitors." Because Web advertising can be measured so precisely, many buyers of online ads, dissatisfied with low click-through rates, now demand CPA (cost per action) deals rather than those based on CPM (cost per thousand impressions).

It's a trend that's likely to accelerate. According to an August 2000 Jupiter Media Metrix report, 30 percent of online advertising will be performance-based in 2005, up from 20 percent today. The report suggests that marketers focus on "quantifiable mediums such as direct mail, direct response TV, and the Internet." It also advises publishers to embrace multiple pricing models.

Forrester Research, meanwhile, estimates that 26 percent of the $5.4 billion spent for online ads in 2000 was performance-based, a figure that will grow to 50 percent of the $17.2 billion projected to be spent in 2003. Which ever set of figures proves more accurate, it's clear that media buyers like the idea of guaranteed results and that ad sellers will have to adapt.

The banner ad's problems stem, in some part, from the fact that Web advertising's effectiveness can be so easily measured. As Tom Hyland and Pete Petrusky of PricewaterhouseCoopers put it in a recent report, "Ironically, it is the Internet's strength for tracking and capturing information that has hampered its growth, as this medium is being held to a much higher standard for delivering results to marketers."

One site that has embraced online accountability and done a significant amount of CPA business is Jackpot.com. Launched in May 2000, the site offers free games and gives consumers a chance to win $1 million every time they play. Six months later, more than 3 million players had signed up and the company became profitable. Giving money away turns out to be an extremely inexpensive way to acquire new customers: Jackpot.com spends less than one dollar per new user (compare that with the estimated $158 spent per customer acquired by now-defunct e-tailer Pets.com). But the lure of free cash also makes people more tolerant of advertising, according to Jackpot.com president and CEO Keith Coen.

"Our product is very effective for the simple reason that people are interacting with the advertiser-sponsors," Coen says, "and they can't help but notice the messages. For example, with our slot machines, when [site visitors] click to spin the wheels, they are actually hoping that Capital One's credit card lines up three times in a row so that they win either cash or Lucky Loot, which is our loyalty incentive program." His site also offers secondary prizes like discount coupons and other offers. He notes that because the offers on his company's site are benefits driven, which in turn makes consumers more willing to act on offers, Jackpot.com has done well selling CPA advertising.

Like Merchandising Avenue's Trubey, Coen believes the Web is better for direct marketing than for branding. The proof, he says, can be found in the fact that while roughly 75 percent of Internet ad inventory goes unsold, his site has sold all its available ad space for months ahead. He attributes this success to knowing his customers.

"Because they have a chance to win money, people are more than willing to share information about themselves," he says. Site visitors take part in pop-up surveys in which they're offered an incentive for sharing their opinion or providing information, such as what kind of offers they'd like to see in the future. (Coen emphasizes that the goal is not to sell users' information to a third party.) Then, based on what users select, the site dynamically generates relevant offers. "That is the definition of direct marketing." For Web advertising, this is only the end of the beginning, so think outside the rectangle at the top of the Web page.

In terms of online advertising, the banner is only the beginning. Miller/ Huber's Myers urges thinking outside the rectangle at the top of the Web page. "We have some clients who are new to online marketing and they keep saying, 'Banner, banner, banner.' But when I hear that, I think of absolutely anything. It means anything that is in a marketing position or negotiable position on a page, so you need to be open-minded and creative from that perspective."

Tom Hespos, Internet strategist at Mezzina Brown & Partners, describes one such campaign, run through his firm's ad serving partner AdKnowledge. Hespos says it has cut client Finlandia Vodka's costs for acquiring mailing list names down to about 7 percent of what it previously paid.

Hespos attributes the success of the Finlandia campaign primarily to the decision to use Unicast's Superstitial format — a form of Flash animation that pops up in a new window and plays after downloading in the background, without any interruption in the user's Web browsing experience. He also credits the campaign's precise lifestyle targeting: Mezzina Brown placed Finlandia ads only on sites visited by the hip, young audience known to drink it, rather than blanketing all the most highly trafficked Web sites.

Everyone seems to agree that Internet advertising is still new, and that experimentation is a must. "The big thing for the Internet more than other media is really about testing and trying different things," says marketing consultant Laredo. "There is no one who has done it right every single time. Keep trying to find what's working, and why, and see if those things work elsewhere."

When planning your company's online advertising buys, look into banners in all their variations; look into Flash animation, streaming video, and e-mail marketing. Investigate sponsorship opportunities. Integrate advertising with appealing content like games, contests, or prizes. And don't get too stuck on click-through rates.

Most important of all, don't overlook the fact that the Web is an interactive medium, and in many cases your customers may be willing to do your advertising for you. Amazon.com and a number of other companies gain an advantage with affiliate networks. Seth Godin, author of Unleashing the Idea virus (Do You Zoom, 2000), suggests that because effective marketing today requires a huge budget to be noticed, the best strategy is to sell products that sell themselves.

"Instead of marketing directly to consumers," he says, "companies should create strategies and environments where consumers can market to each other. This kind of marketing has been around for thousands of years, but it has become extremely important in recent years because of electronic networks," Godin says. He's talking about old-fashioned word of mouth.

"One of my favorite examples is the Polaroid camera," he continues. "Nobody bought a Polaroid camera because they saw an ad. They bought a Polaroid camera because the thing was designed to be used at a birthday party. Someone there would say, 'Look at this!' and so you got sold by someone else. Today we can see examples from Hotmail to, more recently, Modern Humorist or AmIHotOrNot, two Web sites that are attracting hundreds of thousands or millions of people with no advertising budgets whatsoever. Contrast that with the millions a Web site might have spent as recently as a year ago to buy a big portal deal that sent them no traffic at all."

Of course, developing products that will sell themselves is easier said than done, but the motivation to do so points to an ugly truth: Money spent on advertising does not always deliver results. For companies without massive marketing budgets, that means stretching ad dollars by leveraging inexpensive or free sources of publicity. Yale's Judson advises companies not to overlook their placement in search engines. "An enormous percentage of traffic comes from search engines," he says. "That's free."

So before blaming the banner, question your campaign and its planners. Did you choose the right medium for your message? Did you have the best ad creative? Did you reach the right audience? Did you choose the best tools and technology for the job? Take the time to understand what works for your company and learn from your mistakes. Online advertising is only six years old. This isn't the end. It isn't even the beginning of the end. It is, perhaps, the end of the beginning. The banner — although in different forms and powered by improved technology — is likely to fly for many years to come.

Additional reporting by Chelsea Junget.

Making Web ads work
Plan your attack. Know what you're aiming for when you advertise online. Are you trying to increase awareness of your brand or sell your products? Gather customer information or strengthen customer relationships? Different targets require different approaches.
Mark your target. Knowing who you're aiming for is equally important. Insist on seeing tracking data for your online ads. Even the cleverest direct-marketing pitch will fail if aimed at the wrong group of people.
Measure the results. Because performance — click-through rate, customer conversion, or other data — can so easily be measured on the Web, it's essential not to ignore the wealth of knowledge that ad tracking technology can provide. Ad networks should be able to tell you whether your campaign is working almost instantly and, if necessary, make changes. And remember that if they can measure results, your ad rates should be based on them as well.
Don't be afraid to experiment. What works for some companies in online advertising may not work for yours, so don't be afraid to experiment. Try banners, sponsorships, affiliate programs, e-mail newsletters, contests, games, animation, interstitial pop-up ads, and streaming video. Better still, try something no one has thought of yet.