As a media buyer deeply entrenched in the online world, I’m often struck by the paradox that exists between the near-universal consensus that the Internet is the most measurable medium in history and the sheer volume of conflicting data generated by online measurement companies. How is the web really counted?
As chair of the Audit Bureau of Circulations’ new Canadian Digital Advisory Committee, I’d like to offer some perspective on the state of online audience measurement. And I’d also like to comment on some exciting trends that aim to end the historical disparity—and confusion—between the differing measurement methods, trends that will greatly benefit online marketers.
There are three generally accepted methodologies used to measure Internet usage: panel, census and hybrid.
The panel-based measurement method, such as that used by comScore MediaMetrix, recruits a sample of Internet users, tracks their online behaviour and then applies their habits to the larger Internet population with complicated weighting and algorithms. Because panels track the habits of actual people, they provide valuable user demographics, an important component when evaluating sites for media plans.
There are some areas where panels may falter, including coverage of niche sites and work traffic. And publishers are occasionally concerned about how a firm like comScore defines a website, i.e., which URLs are part of the domain definition.
Websites that target a unique or regional market may not be accurately represented on a panel. These websites still serve up valuable audiences, but if only a few panel participants visit them, the data may be inconsistent from month to month and not accurately reflect the sites’ true audience, causing it to be overlooked by buyers.
There is also some concern that panels may have less-than-complete coverage of the traffic generated by people visiting sites while at work. Given the volume of workers that spend their days connected to the Internet, this could have a significant impact on a site’s overall traffic stats.
The last area is related to the domain names that are counted as part of a site’s overall traffic. When panel-based measurement firms gather the data for a website, they must determine which domains to include in the aggregate numbers for that site. This may seem like common sense, but in reality many website publishers use a variety of third-party companies to deliver content throughout their site. The third-party content may be branded to mimic the original website and a visitor may never realize they’ve navigated away from the initial domain. So, who gets to count that traffic? The website the user sought out or the website that ultimately delivered the content, even if the user never realizes they left the original site?
Websites that use analytics tools, such as Omniture, are measured according to the census-based methodology. This method does not provide any of the valuable demographic information, but it does have the distinct advantage of tracking almost 100% of a site’s traffic, regardless of the visitors’ locations, including home, work and mobile.
The census method’s greatest downfall, however, is that it focuses on tracking machines, not actual people. This leaves us without beneficial demographics and can lead to some duplication in the data. For example, if someone deletes his cookies every day, that person could be counted as a different visitor every day. If someone uses three devices to access the same website (home computer, work computer, mobile phone), that person could be counted as three visitors.
This is why it’s important to recognize a clear distinction between panel measurement and analytics. The former measures people; the latter measures machines. Apples and oranges, I’m afraid.
The hybrid, or unified measurement, method is the next-generation technology available for measuring online audience and it shows significant promise. Purveyors of the hybrid method say that it incorporates the positives from the panels and analytics software and combines them to create a new web measurement tool that tracks nearly 100% of traffic and provides detailed user demographics.
While only recently introduced by comScore MediaMetrix, the unified approach has been well-received by many of the website publishers and digital advertisers I’ve spoken to.
ABC’s new digital committee has created a forum for leaders in our industry to gather and educate ourselves on the future of digital media, the methodologies used to measure it and the best way to ensure digital media delivers for our clients. As Neal Lulofs, ABC’s senior VP, communications and strategic planning said, the need for a standardized and up-to-date model of digital measurement was the primary motivator for creating the committee. Because I believe it serves such an important role for our industry, I volunteered to serve as its chair.
In addition to the state of online measurement, the committee also discussed ad operations workflow challenges; the explosion in e-reader and mobile devices; and ABC’s current and future reporting options, including the recent change that allows magazines to better accommodate new devices like the iPad.
If you’d like to share your thoughts about online measurement and anything else related to digital media, please drop me an e-mail at [email protected]. To learn more about the Audit Bureau of Circulations’ Digital Advisory Committee which meets regularly in New York, please contact Neal Lulofs, SVP Communications and Strategic Planning at [email protected].
Originally Published in Canada’s Marketing Magazine.
