
Wall Street Journal — October 24, 2007 —
By EMILY STEEL, Wall Street Journal
As online advertising matures, so does the job of measuring results
When marketers first started advertising on the Internet about a dozen years ago, the Web was heralded as the Holy Grail of efficient marketing.
Unlike television commercials or print ads, where marketers never could be sure exactly how many people saw their message, the Web seemed easy: If someone clicked on the ad, then it worked.
Advertisers increasingly want more-detailed feedback on the effectiveness of their online marketing efforts, both to justify their increased spending on the Web and to help guide their future allocations across all available media.
They want to know not only what ads consumers click on but also how those clicks translate into purchases. They are more focused on understanding how their mix of online marketing works together and how it fits into their broader marketing campaigns. They want to find out how online advertising is affecting the image of their brands, and how to target ads to specific audiences.
In response, advertising agencies are turning to companies that use advanced analytic tools to make sense of a flood of data. And in many cases these tools are being combined with old-fashioned devices like surveys and focus groups to give advertisers a clearer picture of what's working and what isn't -- and why.
"It is more or less an evolution of metrics and measurement that has been happening," says Jason Bigler, vice president of advertiser solutions at Internet advertising firm DoubleClick Inc.
Focus on Results
Part of that evolution is an effort to analyze the wealth of data that are available more efficiently and effectively. When a company buys an online ad campaign, it may work with several partners besides the agency that created the campaign, including Web sites where the company places ads directly and a variety of ad brokers. Each of these partners collects data -- including, for instance, how many people are choosing to view a video that runs with an ad, or, in the case of a broker, which Web sites are generating the most clicks -- and delivers that information to the marketer and the ad agency.
Until recently, the agency would have to collect all these data and cobble them together to help the marketer understand what they mean. With information often coming in from dozens of sources, this is a laborious and time-consuming process, and one that discourages the kind of extensive analysis that can take full advantage of the available data. But now, several companies are using new software that gives computers, rather than people, the job of gathering and analyzing the data. That promises to get the job done faster and more thoroughly.
Omniture Inc., of Orem, Utah, is one such company. CEO Josh James says one of the advantages of the analyses his company performs is that they go beyond how well an advertising campaign is reaching consumers, showing as well what each element of the campaign is contributing to sales. "Nobody really cares about clicks," says Mr. James. "What you really care about at the end of the day is how many people purchased."
Digital marketing firm Organic, a unit of advertising giant Omnicom Group Inc., works with Omniture to understand how ads for clients perform. Using data from Omniture, Organic is able to learn, among other things, which sites customers visit before landing on an advertiser's site. "What it allows us to do is not only see the most successful [ads] that users are coming from but also the most successful places where users are coming from as well," says Rick Corteville, executive director of media for Organic.
Omniture also helps Organic see how much time each visitor spent on a client's site, the average length of session per user, the number of pages visited, the number of elements downloaded and what makes people leave a Web page.
All of this allows Organic to make continual adjustments to its advertising campaigns and its clients' Web sites. "It is such a living thing," says Mr. Corteville.
Omniture is also starting to offer technology that automatically changes a marketer's advertising mix based on the computer analysis of the data.
The Big Picture
Companies also are looking to better understand how their digital campaigns and their offline marketing initiatives complement each other. This not only helps ensure that their ad spending is properly distributed, it also puts companies in a better position to adjust their campaigns in reaction to changes in consumer habits, broader economic trends or business objectives.
For this kind of big-picture analysis, companies like MediaBank LLC combine online-data crunching with more-traditional methods of directly surveying consumers' reactions and attitudes. For example, MediaBank will talk to consumers to see how perceptions of a brand or a product differ among those who have seen a certain ad and those who haven't.
e-Storm International Inc., a privately held interactive marketing firm based in San Francisco, uses analytics from MediaBank to help figure out the best marketing plans for a number of its clients. In the past, e-Storm would look at multiple sources of data to understand how its clients' online ads were performing. But that data didn't include any information on how traditional offline marketing was influencing purchasing decisions. The data e-Storm now receives from MediaBank give it insight into the effect of all of a client's advertising techniques, including, for instance, billboards, coupons received in the mail, print ads, search-related advertising and a banner ad that appears on multiple Web sites.
"I've been starving for this kind of tool," says William Gaultier, e-Storm's CEO. "As marketers, we've been flying blind for a long time."
Such broad analysis has the potential to shift advertisers' spending patterns by clarifying the value of various elements of a marketing campaign. For instance, historically, if someone clicked on an ad online and then bought the product, that ad received all of the credit for generating the sale. But marketers are starting to find that there is a multiplier effect: When people are exposed to more advertising both online and off, they are more likely to purchase a product. That leaves marketers trying to figure out how much they should spend on pieces of their ad campaigns that don't immediately lead to a sale.
Jon Raj, vice president of advertising and emerging media at Visa Inc., emphasizes the importance of knowing whether his company's online marketing is communicating to consumers the attributes that Visa wants associated with its credit cards -- for instance, that they are secure for use on the Internet. "If we don't get the message across" on those brand attributes, he says, "then we've failed."
A Sense of Urgency
There are other challenges still to be tackled. For one thing, with so many different sets of data coming from different companies, marketers are looking for more standardization. For instance, the figures for unique visitors to a Web site tracked by comScore Inc.'s Media Metrix unit rarely align exactly with numbers from Nielsen Co.'s Nielsen/NetRatings service, and often these numbers differ from a company's own internal data. In addition, as new advertising opportunities emerge in fields like mobile phones and gaming sites, companies will be looking for even more information to track their ad campaigns.
For some digital-marketing executives who remember the fallout after the burst of the dot-com bubble in the late '90s, the effort to measure the effectiveness of online advertising has some urgency.
"My concern is that we are going to experience a kick if we don't figure out these measurement questions," says Organic's Mr. Corteville, suggesting that the growth of online ad spending could be slowed. "We don't want to take steps backward from all that hard work."
For the original article on Wall Street Journal, click here

