Reform in the air

The controversy over defining unique visitors — and in general over the very different pictures of the online world painted by various measurement firms — has provoked calls for an organized, industry‐wide reform. The state of affairs was captured well in a column by Steve Smith, digital media editor at the Media Industry Newsletter: “We are a decade and a half into the life of the ‘most accountable medium,’ the Web, and just this week we see some of the major online measurement firms still tweaking their models and arguing over methodology. I have major media companies reporting their monthly numbers to me, and I see staggering differences between the stats they claim from their internal logs via Google Analytics or Omniture and the third‐party numbers. Itʹs not just mobile, either. Itʹs still a mess all over”¹⁶a sign of the times, the Internet Advertising Bureau — a trade group “dedicated to the growth of the interactive advertising marketplace” — made Smith’s condemnation the opening slide of a presentation in early February to the Association of National Advertisers. (The presentation went on to highlight the wide differences in the monthly rankings released by Nielsen and comScore, reproduced above.) IAB appears to be leading the reform charge, at least in the world of panel‐based measurement. A confidential McKinsey & Co. study commissioned by the IAB and the American Association of Advertising Agencies concluded in 2009 that confusion over metrics stood in the way of greater online ad spending. Based on this report, the IAB has proposed a “cross‐industry task force of senior marketers, agency executives and media leaders to reach consensus on standardizing and simplifying basic audience measurement.” The Newspaper Association of America has committed to support this task force and to recruit participants from the newspaper industry. A parallel standardization effort is underway from the Media Ratings Council, a group comprising media companies, advertisers, and agencies which dates from the 1960s and whose mission is to accredit and audit ratings firms. The MRC has worked with the IAB to coordinate definitions (for instance of IAB’s controversial 2009 guidelines for “unique visitors”). However the MRC is also in the process of accrediting comScore and Nielsen, an effort likely to continue into next year. It is conceivable that these ongoing audits will make the two panel‐based measures more compatible, and perhaps more transparent to publishers and advertisers using them. In the world of server‐side measurement, any number of organizations offer third‐party auditing of traffic data from Omniture and other Web analytics tools. Both leading circulation auditors, ABC and BPA, operate interactive arms that produce verified online readership figures. ABC has been particularly active here: Its “Audience‐FAX” service, developed in partnership with the NAA and Scarborough, purports to measure a newspaper’s net readership across print and the Internet. However, this service depends on newspapers to submit readership data in various categories; the competitive picture may be skewed if newspapers calculate these differently. The emergence of new platforms and devices also poses a growing problem. For instance, as yet no consensus exists around how to measure streaming media and online video. How is an impression defined when content is continuously streamed? When it is short‐form versus long form? When it plays in the background, as most online radio does today? How these terms are defined is no small matter – streaming music service Pandora today stops playback if a listener doesn’t engage with their page in a certain amount of time. Each time a user has to click back to that page to hit “play,” a new session is initiated, thereby boosting Pandora’s traffic numbers.