Are You Getting Your Money's Worth from Online Ads?
In a survey conducted by the Association of National Advertisers, New York, more than half of consumer brand managers interviewed cited the lack of reliable audience measures as a key issue keeping them from buying online advertising.
Indeed, recent ABC Interactive audits show brand managers' concerns are real. The results also underscore the value of independent third-party audits when buying and evaluating online ads.
From April 1, 1998, through March 31, 1999, ABC Interactive audited more than 13.7 billion page impressions and more than 61 billion ad impressions. In 99 percent of the audits, adjustments were necessary.
Among Web site audits of page impressions, 40 percent of audits involved impression adjustments of more than 5 percent and 18 percent uncovered a need for adjustments of more than 10 percent. Some Web site adjustments were as high as 30 percent.
Among advertising audits for ad impressions, 25 percent of audits revealed ad-impression counts that had to be adjusted by more than 5 percent. Fifty-five percent of ad audits reporting click-throughs required click adjustments of more than 5 percent and 31 percent of ad audits reporting click-throughs required adjustments of more than 10 percent. The average adjustment for click-throughs was almost 10 percent.
These figures illustrate why an objective third party is essential for advertisers who need assurances they're getting what they paid for. Only the kind of accountability that comes from independent third-party auditing will increase participation in the online media marketplace.
However, auditing is not to be confused with log-based activity totals or panel-based measurement. Auditing refers to the checking, evaluating and verifying of what's already been counted. It helps eliminate the risks of accepting log-file measurement at face value.
There are three general categories of risk -- noncompliance, mis-statement and manipulation -- and all apply to any advertiser-supported, census-based medium.
Noncompliance is where information is not reported according to industry-accepted standards. Noncompliance makes it impossible to make apples-to-apples comparisons. Even when data from a Web site's server log file appears to comply, these numbers are not necessarily comparable across sites without an independent audit. Data that doesn't permit direct comparisons loses credibility and value.
Misstatement is where a log file is misrepresented to buyers either deliberately or due to a lack of understanding of industry-accepted standards.
Manipulation occurs when log-file data has been altered internally or externally. An example is when a computer is programmed to request certain Web site pages continuously for a search engine. This so-called spider or robot clicks on pages. However, no human sees the pages or the advertising they contain.
Spiders are just one of a growing number of ways that log-based tabulations can be magnified and manipulated. Internal manipulations, in particular, are almost impossible to detect by examining the original log file.
Print advertisers have long recognized that media properties performing their own accounting have an interest in reporting the best possible results. Using the same organization to perform the counting and auditing is a clear conflict of interest.
Numbers audited by a third party allow media buyers and sellers to make informed business decisions on an objective, quantitative basis.
An increasing number of Web sites rely on measurement-and-analysis tools to identify visitor buying habits and behaviors. Auditing this data reinforces a site's accountability and gives advertisers increased buying confidence. Seamless integration of these two functions -- measurement and auditing -- is a given for any site committed to providing its advertisers with the level of accountability they desire.
Evelyn Hepner is vice president, sales and marketing for ABC Interactive, Schaumburg, IL, the online auditing arm of the Audit Bureau of Circulations. Reach her at firstname.lastname@example.org.