SPECIAL REPORT: Page Views OK --'Time Spent' Better?

By: Jennifer Saba As newspapers continue to beef up their sites with allurements ? videos, blogs, user-generated content, and the like ? the goal is not just simply garnering more eyeballs. No one ignores the benefits, selling power, and prestige of heavy traffic, but other metrics aside from unique visitors and page views are playing a greater role in determining an online destination's popularity ? and therefore how many dollars it can capture.

Key among these metrics is the amount of time spent, per user, on a site. One "unique" might surf in and out for 30 seconds total in any given month ? or might read, dawdle, and watch videos for hours.

Time spent can be defined in a number of ways. Like every other online data set, including the tracking of unique users, there are many ways to look at time spent, depending on who's minding the clock.

Nielsen Online, for example, collects its data mainly through a panel of users who voluntarily install tracking software on their computers. The company follows the average time spent per person at a site during a given reporting period (usually one month). In August, for example, Nielsen reported that the average time spent per person on NYTimes.com throughout the course of the month was almost 30 minutes.

Publishers also can measure, down to the user, how much time a person is spending on the site during each visit by analyzing server logs. For instance, at the Detroit Media Partnership, which publishes The Detroit News and Free Press, internal data show the average time spent per visit during the course of a week was 17 minutes, according to John C. Jackson, the Detroit Media Partnership's vice president of digital sales.

"Most of the 'power users' are coming back three, four times a week," says Jackson, who has the data segmented to tell if someone visits the site more than 10 times a month. That would be a pretty loyal person, he notes, compared with someone who only stops by occasionally.

Many advertisers still buy based on CPMs (cost per thousand) per page view. "The cold hard reality now is still the more page views we get, the more ads we serve. That translates into more dollars," notes Anthony Moor, deputy managing editor/interactive at The Dallas Morning News. But he adds that executives also pay close attention to time spent: "As an editor, I like seeing time spent go up ? that's a good thing."

Jackson knows that it's critical for him to educate advertisers on the value of that metric. "Our position on time spent is that it is a very important measurement," he says. "It measures the 'stickiness' on your site. If someone comes to your homepage and doesn't go anywhere else from there, there is less of an opportunity to interact." To keep people longer, the Detroit sites emphasize video, photo galleries, and longer stories.

Chris Jennewein, senior vice president and publisher of Greenspun Interactive, thinks that time spent is a metric that has gotten shortchanged ? and one that will eventually carry more currency. "As this medium matures," he says, "sites that can show greater time spent per user are going to be able to command higher CPMs."

While the number of page views still drives ad buys, he explains, time spent acts as a deal-breaker. Given two different Web sites with similar traffic, the one that shows the greater time spent will be able to command higher rates, Jennewein contends. He believes that reader interactivity will drive that time up. "It's a plus when readers can comment," he adds, noting that sidebars around popular stories also aid in getting readers to stick around longer.

More important, those metrics should be fleshed out for local or "in-market" readers as well: Typically, there is an inverse relationship between traffic growth and time spent. For example, if a newspaper Web site covers a popular national story, people outside the market tend to spend less time on the site; they visit, read the story they came for, and then leave. So do marginal visitors, even if they're within the site's market.

Using the San Diego wildfires to illustrate his point, Jennewein, who previously led Internet operations at The San Diego Union-Tribune, explains that a spike in traffic caused by a big news event can make it difficult to make accurate year-over-year comparisons in page views or time spent. "A national story can swamp everything else," he says.

The U-T's site enjoys high traffic, but trends toward lower time spent: For example in June, Nielsen Online (owned by E&P's parent company, The Nielsen Co.) reported that SignOnSanDiego.com grew its traffic 51% year-over-year. The average time spent per person dropped, however, from almost seven minutes in June 2007 to about four minutes in June 2008.

"If you do a great job being search-engine friendly with stories that have national attention, it can bring down time spent," Jennewein adds. "Doing well can have the ironic impact of making you look bad."

Getting engaged
Publishers tend to look at page views and time spent together in order to get a better sense of their user engagement.

At the Dallas Morning News, the site has been growing traffic, thanks in part to its alliance with Yahoo. The Morning News crowed that because of that boost, Yahoo placements have accounted for up to 27% of the day's page views, and 65% of the day's unique visitors.

Moor says the site has actually been attracting many Dallas-based users, people who typically don't visit DallasNews.com: "Yahoo has terrific reach locally ? even better than we do." What's curious, though, is that the time per user is growing. "Those new users are less likely to [deliver] more page views," he says. "What is interesting for us is the increased time spent but not page views. My gut tells me it's video."

Time spent also plays a role in the level of "engagement" ? another term that can be difficult to even define, much less measure. There is no standard for measuring engagement, says Domenic Venuto, vice president of the media and entertainment practice at digital marketing firm Avenue A/Razorfish in New York: "That is what the industry is grappling with."

The dilemma facing many publishers: engagement is absolutely necessary, but there are many ways to go about it. "I think it's critical that we drive engagement," says Leon Levitt, vice president/digital media at Cox Newspapers. But to Levitt, that can be determined with several different data points, including time spent on site, page views per visit, and frequency. In his mind, frequency trumps everything else.

"How do we get people to come back to our sites every day?" he asks. "That is a better level of engagement than time spent."

Levitt stresses the importance of noting what time of day people visit, and correlating that data to the amount of time they spend on the site. A user may come in the morning for the day's breaking news, while that same person may linger longer during his lunch hour to check out a video or two. Cox's AJC.com, the online arm of The Atlanta Journal-Constitution, features a section called "LOL" (e-mail shorthand for "laughing out loud") that spotlights various comedians. People spend more time with those videos. "I don't think every 'daypart' is the same once it comes to time on site," he adds.

Howard Owens, director of digital publishing at GateHouse Media, looks at page views and, to a lesser extent, time spent to gauge a reader's engagement. A popular story or post on his personal blog, he notes, can drive up time spent per user: "Time spent can tell you, in a way that page views can't, if people are spending a lot of time on a big investigative report."

Jason Silverstein, director of the Charlotte (N.C.) Observer's interactive division, says time spent is a metric that definitely warrants attention. Silverstein, who joined the newspaper from Yahoo last year, contends that page views can indicate a thousand different things. A Web site can put up a slide show with a 100 slides and 10 people can breeze through all of them, producing 1,000 page views alone ? but just what does that signal?

Silverstein, who just led a redesign of CharlotteObserver.com, believes in breaking content into "easily digestible" pieces so people can stay and consume, and keep coming back. Sometimes all the bells and whistles ? even slide shows ? try to do too much, he says. Keeping it simple helps. "It's not about one 10- minute chunk," he adds. "It's about two five-minute chunks."

On the flip side, an increase in time spent can also indicate that a site is poorly constructed, that people have to hunt and dig deep for information. Another common belief is that a user can walk away from the computer, leaving a site (or sites) up for hours. Nielsen, however, says that it has measures to determine if a user is inactive for 30 minutes or more based on keyboard and mouse activity, and the time-spent count halts after that.

An executive with Omniture, one of the leading companies for Web analytic software, argues that it's virtually impossible to know if someone is idling or not. Christopher Parkin, Omniture's senior director/Genesis Solutions, claims the time-spent metric is being peddled by panel-based companies Nielsen and comScore. "Measuring the value of visitor engagement based upon time spent on page is absolute folly," he says.

One reason is that a person can have multiple tabs open in a browser in which only one is in use at any given time, but it appears that the visitor is still on the other pages. So while the person may be "active" in only one tab, the other tabs open also get the credit, he asserts.

Scott Ross, director of product marketing for Nielsen's NetView service, says that only an "active tab" gets credit: "Server-based solutions, including Omniture, are limited in their ability to measure active time as they are only able to see activity at the site level, and not what is active on the desktop."

Parkin feels that engagement is better measured by monitoring what content people are reading ? in effect, what is most popular ? and then repositioning that content during the course of a day to generate the highest yield of page views. The popular content will keep people coming back to the site.

Time ads up?
According to the Online Publishers Association (OPA), newspapers are extremely well-positioned to take advantage of advertisers looking for branded and trusted sites.

The association tracks how people spend time on the Internet in five separate categories: community (which includes social networking), commerce, communications, search, and content. By far, more people use the Internet for content ? the category captured a 42.3% share in June. The category with the second-largest piece of the pie was communications, with a 28.9% share.

OPA President Pam Horan admits that the online industry at large is still in the process of getting its head around measuring engagement, and time spent is a good place to begin. "It's certainly a great start," she says. She adds that OPA found advertisers also like to be associated with news sites with recognizable and trusted names: "We see this 'brand halo effect' captured for advertisers."

At the macro level, this is what marketers are looking for, says Andy Ellenthal, newly appointed CEO of the national online ad network quadrantONE ? which counts The New York Times Co., Tribune, Hearst, and Gannett as investors. But the conversation usually has to go to a deeper level.

"You are not talking about a user going to their Hotmail account," he says. "You are talking about users that are there for the actual content. I think if you ask any advertiser, they would rather be in content than low-premium type inventory. It's a question of the user paying attention."