In April 2007, Web tracking company Compete.com added an "Attention" metric to their analytics, which takes into account time spent on a site. In July of last year, Nielsen//NetRatings followed suit, and began using time spent as the primary metric of popularity. In search, text-dominated search engine results pages are making way for universal search results that include more "engaged" formats for search listings.
Similarly, Yahoo and Google recently added video ads to their paid search programs--a change that favors time spent with the ad rather than impressions or clicks. And YouTube recently announced YouTube Insight--an analytics product that measures the length of users' interaction with videos. Over the last few months, time spent has clearly become an established metric, and it provides one gauge for engagement.
While it's valuable, time spent is not an adequate or complete measure of audience engagement.
First, tabbed browsing and online multitasking means people can frequently spend hours on a site without actually looking at it.
Second, the prevalence and growth of data portability has allowed people to receive and interact with content on their own terms, off-site. Tools like RSS, widgets, and "lifestreaming" via social aggregators like FriendFeed deliver site content straight to users, enabling them to interact with the content without ever visiting the site.
As the online experience becomes richer, it's also becoming difficult to define meaningful interactions, and even more difficult for analytics tools to aggregate all this data. Very soon, those tools will need to measure feed subscribers, Twitter followers, and the number of Facebook wall posts alongside more traditional site statistics such as unique visitors, page views and time spent. The challenge marketers face is making sense of all that data and applying that information to the way they buy ads now (currently page views and impressions--although this will change soon, too), both online and across their broader marketing mix.
Earlier this year, Microsoft added an "engagement mapping" feature to its reporting that seeks to give marketers a better idea of what pieces of their marketing mix actually resonate with users and lead to conversions, rather than just giving credit to "the final click." That kind of engagement metric doesn't rely on time spent at all, but instead wants to determine what people actually pay attention to, and what makes them buy.
In the end, we all need to realize that there won't just be one standard. Marketers have to pick metrics that most appropriately map back to their objectives. This may be a more abstract, nuanced, and complicated approach to setting up a campaign, but ultimately it will lead to better and more effective marketing.