The fledgling online ad industry is going gangbusters, with growth this year expected to be at least 100 percent over last year and expenditures reaching beyond $9 billion.
    Much of the appeal of the web to media buyers is its promise of accountability. In theory, you know not just where your client's ad appears, you actually know how many people see that ad.
    That's in theory. Reality, it turns out, is something else entirely.
    The reality is that there are huge discrepancies in online ad data. Buyers routinely hear wildly different numbers reported for their clients' campaigns.
   When a banner ad for Product QRS runs on Site XYZ, the server at Site XYZ may tell the ad agency for QRS that 110,000 impressions were delivered.
    But the third-party company that actually serves the ad to the site—perhaps DoubleClick or AdKnowledge or Real Media—may report that only 80,000 impressions were delivered.
   Part of the problem is technological. The technology in place at many web sites is not adequate to the task of effectively measuring ad impressions.
    But an even bigger part of the problem is a complete lack of agreement upon what an ad impression is and how to measure it. Different kinds of software programs go about the job differently.
     In short, the world of new media is awash in numbers but more often than not those numbers are in conflict with each other, to the great consternation of media buyers.
    "I have seen differences as high as 90 percent," says Paul Benjou, senior vice president and group media director with Draft Worldwide "But no one's been paying attention to it because no one wants to pay attention to it."
    This could well be the single biggest problem facing new media today: ad discrepancies and how little anyone seems to be doing about them.
    "Only if there’s a very egregious discrepancy—say 40 percent or more—will anyone bother to call up the site," says Tig Tillinghast, director of agency strategy at Solbright, a company that sells ad management software. Before Solbright, Tillinghast worked in interactive departments at both Leo Burnett and J. Walter Thompson.
    Discrepancies that would cause alarm in other forms of media are basically accepted when it comes to the web, says Tillinghast. What’s more, he says, these discrepancies often stay within the confines of the new media departments.
   "Frequently the managers of agencies don’t realize there’s such loose stuff going on," he says. "They just assume that all discrepancies are resolved, as they are in spot TV or print."
   How serious a problem is all of this?
     Imagine for a moment if in television it was never exactly clear how many of an advertiser's commercials ran during a week's worth of "Who Wants to be a Millionaire?" Imagine if ABC thought it ran six spots and the advertiser thought only two had run. And imagine if there was no way of actually proving which number was right.
      "The industry is implying its own guilt when most sites request a 20 percent variance on their own impressions," says Benjou.
     That is to say, many web sites right up front acknowledge that they can’t exactly know what they are going to deliver for the client.
    As it turns out, the ad impression discrepancy problem is just one of a series of interrelated problems facing the online ad industry when it comes to measurement, discrepancy and accountability.
    To examine the entire issue, Media Life will be running a series of stories reporting on the new media discrepancy problem and its ramifications.
     We will look in depth at the underlying ad impression discrepancy issue in our next article.
    Another issue that arises directly out of the difficulty of measuring ad impressions accurately is how effective targeting can be done under such circumstances.
    Advertising on web is supposed to give advertisers the capacity to do all sorts of specialized targeting.
    If a perfume maker knew that its scent was most compelling to women in their 30s who work for major corporations, live in cities with at least 400,000 people, drive two-door cars and don’t have children, this company could in theory go to a major web site and see that its ad is run in front of this exact group of people.
    But this sort of targeting is not often happening, and one of the reasons is because the basic, underlying data about ad impressions is so unreliable. 
    If the site thinks it delivered 1,000,000 impressions, the ad server thought it was 600,000, and the client had paid for 1,200,000 in the first place, then who can really say where these impressions were delivered and where they weren’t, and whether the "right" people received them.
    Media Life will examine this issue more closely in part three in our series on new media and accountability.
    If the discrepancy issue is so vexing, why isn’t it being resolved?
    In no small part because there are a number of different forces in the market and they each disagree over which standards ought to prevail when it comes to what an ad impression is and how to measure it. As a result, no standards are prevailing, a situation we’ll check into in part four.
    Okay, but this is obviously a problem that will eventually be resolved, right? Not necessarily. 
   There are not only conflicting standards, there is a pitched battle going on for control with a lot of money at stake. Industry consortia have formed to help straighten the situation out but they haven’t gotten anywhere. We may well end up with another Nielsen situation, stuck in new media with a measurement system no one believes in or is happy with but no one can do anything about.
    Media Life will discuss this issue in article five.
    Next we’ll turn to the results of a poll Media Life conducted last week. We asked people involved in buying new media what they find most exasperating about the process. Some of the answers are what you’d expect, some surprised even us. In part six, we’re going to tell you what buyers find most troublesome about the world of new media.
     In part seven we’ll wrap the whole thing up. We’ll look at everything we’ve covered and come up with some recommendations about this seemingly insoluble conflict.
    The articles will be appearing regularly over the next two weeks. 
     Part two will appear here on Wednesday.

-Jeremy Schlosberg is the senior editor for new media..


                     Cover Page | Contact Us