Next month I'm on a keynote panel at the Advertising Research Foundation's third annual Audience Measurement Symposium (AM 3.0), titled "The TV/Video Audience Measurement Challenge." I've been thinking a lot about video lately, and talking about video with interactive agencies, traditional agencies, major TV companies porting their content online, and pure-play Internet companies expanding into video. I've come to believe that video is perhaps the most important development in the online metrics space, and that the rapid rise in importance of video will be -- and for many, has already been -- a game-changer.
At the risk of oversimplification, once upon a time Websites were made out of pages. Today, they are made out of pages and videos (and other rich media and dynamic content.) This is a fundamental sea change in the way online content is presented, distributed and consumed, and it is happening quickly. Consider: at comScore, we reported that in February 2008, Americans consumed 13% more videos than in the previous month; and the year over year growth was a somewhat astounding 64%. Perhaps even more astounding: for the most recent month reported (March 2008), Americans spent 77% more time with online video than a year ago. By any criterion, online video is a growth medium.
From the perspective of traditional media -- the big TV-focused media agencies, and media companies who are in the business of distributing TV programming on broadcast or cable -- online video is not so much a flavor of the Internet, as it is a flavor of TV. While this may seem disheartening to the collective digerati, keep in mind that even now, for many advertisers TV continues to command the lion's share of budget. If the traditional media buyers and sellers see that bucket of money as sloshing over to the Internet, I'm betting every digital boat will rise with the tide.
I think there are two important implications for measurement.
The importance of duration. Video is fundamentally different than page consumption. Our metrics are going to have to evolve to reflect the way consumers interact with video content, and quite specifically, that interaction occurs with respect to time. Buyers and sellers want to see video content aggregated into meaningful entities that allow for cross-media comparison, which will invariably mean that -- at least for long-form, ad-supported program content -- we will migrate to duration-based metrics that look a lot like traditional broadcast metrics (maybe even demographic quarter-hour ratings.)
The emergence of online video will serve to hasten a continued shift in online metrics toward duration-based metrics for all online content types. Advertisers and publishers won't be able to use Page Views (or, for that matter, number of streams) to compare the audience engagement of an online newspaper article, a stream of an NCAA basketball game, a gaming environment, a social network, and a widget. But all these content types accrue duration.
Video hastens the migration to duration-based metrics for all online content.
People centric. Advertising evaluation is about schedule performance; a schedule is an aggregation of media and media vehicles in the service of a marketing communications strategy. The Internet is evolving from a standalone medium to an integrated platform on which all media reside (newspapers, magazines, TV shows, and radio all live online, and what is a banner ad if not a tiny little billboard I can click?) As a result, publishers are packaging content across platforms -- for example, a newspaper's Web site expands the reach of a buy in the paper itself. Advertisers will increasingly demand information on the holistic, cross-platform performance of schedules. This means that machine-based metrics -- site-centric data -- will have less and less efficacy over time in an ad sales context... at least on its own. Many of the agency researchers I've met with recently are trying to understand how to place value on online video as part of a package that includes on-network advertising. And it won't be long before mobile video is added to the mix.
I've made the distinction before in this space about the difference between Audience Measurement and Web Analytic data. Site-centric, so-called census data is pretty good at telling us what machines are doing (especially if those machines are servers.) But advertising is viewed by people, not machines. Machine-based site centric data can't disentangle reach from frequency for a single Web site (i.e., cookie deletion and the controversy about site-centric estimates of reach, or Uniques.) For advertisers needing to understand the performance of their campaigns across Web entities, across content formats (pages, video, etc.), and even across platforms (broadcast, cable, streamed, mobile), the only conceivable workable solutions remain person-based as opposed to site-based. And the reason is simple: because advertising targets people, not machines, and individual people will be using more than one machine to see your ads.
And, too, I would be remiss if I failed to point out that the fundamental building blocks of duration-based metrics are start times and stop times. Site-centric data sources have yet to solve the problem of figuring out stop times -- Omniture doesn't know when I've left your site -- which means they cannot yet be of use in building actionable duration-based metrics for use in planning and buying advertising.
I don't usually write about my own company in this column; if anything I've bent over backwards to avoid pieces that might come across to MediaPost readers as unduly commercial. But I will tell you that we have a lot of exciting things going on at comScore, all generally designed to provide a 360 degree view of the digital media landscape for publishers, advertisers, and agencies. We are going to do this by keeping the consumer -- or as we call him/her, the panelist -- at the center of the equation. You may have heard about the strategy "follow the video." In today's increasingly fragmented environment, successful advertisers will be the ones who follow the people, who place the consumer at the center and surround him/her with relevant and appropriate messaging, in context, at the right time in the purchase cycle. The only way to do justice to such a complex landscape is with person-centric metrics.
Does all this go beyond the purview of online metrics, of online video metrics? I don't think so. Because online video might just be the thing that brings digital crashing through to the mainstream -- or that redirects the mainstream straight through the online space.