Life after page views: Web analytics 2.0

Many agree that the page view is dead, but not about the next Web measurement metric

In a move that many said signaled the official death of the page view, Nielsen/NetRatings last July announced that it would no longer use page views as the primary metric for comparing Web sites. At the time, the Internet benchmarking firm cited the growing popularity of Asynchronous JavaScript and XML, or AJAX -- which can refresh content without completely reloading a Web page -- as the main reason for the change to measuring time spent on a site.

Eight months later, Nielsen now says that it overestimated the impact of AJAX on page-view metrics. Nielsen found that instead, online video is the key reason for the growing irrelevance of page views.

Nielsen is among a growing number of companies grappling with "life after page views," a new Web analytics arena where blog posts, widgets, online video and other emerging Web 2.0 media are changing how companies look to measure the effectiveness of Web efforts -- and the tools used to do the measuring.

"AJAX specifically is a little bit overestimated as far as the breadth of its impact," noted Scott Ross, product director of Nielsen's NetView Internet audience measurement tool. "A larger impact is probably being felt by the integration of more video on pages."

The new metric hasn't led to "widespread benefits" for those sites that heavily use AJAX, Ross added. Rather, the "time spent" metric has mostly benefited sites offering online video and/or slide shows, he noted. "It's a little surprising, but our position was, we didn't quite know where page views and all these technologies were going to go," Ross said. "We just knew as in-page technology would be more and more common that page views were less of an indicator of user activity."

Nuconomy, a new Tel Aviv-based Web analytics company that this month opened a private beta of its Nuconomy Studio technology to several hundred users in the U.S., agrees that the page view is increasingly irrelevant. However, it offers a different new metric than Nielsen does.

Rather than measuring minutes spent on a site, Nuconomy focuses on what CEO Shahar Nechmand describes as a Web site's "semantic layer." Nuconomy's analytics tool measures comments added to blogs, ratings, applications shared with friends, clicks on ads and online video use -- all of which can show how "engaged" a user is with a particular brand or product, he said.

"Page views don't really tell you the whole story," Nechmand noted. "If you have widgets, you don't really have page views. If you do videos, you don't have page views. We actually measure every action that every user is doing and not just the aggregate."

Each company can create its own measurement formula with the free tool, he added. For example, a site can use the tool to measure the number of comments it generates, the number of recommendations it receives or how many links back to it from other sites it generates.

In addition, Nuconomy has created a two-way API to allow users to take the information generated from the tool and use it to make adjustments to their content.

"You can actually take the information in an automated way and create better Web experiences [and] change your business based on this information," Nechmand said. "We believe that the biggest problem in analytics today is it is completely separated from operational systems. People view analytics for the sake of analytics. If I can show you that a user is more interested in sports than technology, and if you don't do any personalization to accommodate him, it doesn't mean anything to your business."

Nuconomy plans to release a plug-in of its technology for bloggers next month. In the next two months, it plans to release a public beta of the technology. It intends to make its Web analytics tool available for free to companies with fewer than 3 million unique visitors per month, Nechmand added.

Avinash Kaushik, a Web analytics consultant whose clients include Google Inc., for its Google Analytics product, cautions that despite the death of page views, companies shouldn't blindly embrace measuring "engagement" without a deep assessment of the concrete metrics behind such an effort.

"A lot of people think the page view is dying so we should measure engagement," he noted. "Just because the page view died, who … gave you the right to move to engagement? The Web is becoming more fluid in terms of how people interact with it. The fluidity does not mean the core questions you wanted to answer go away."

Kaushik, who is also author of the book Web Analytics: An Hour a Day, pointed out that while page-view counts traditionally have been obtained by monitoring server log files and JavaScript tags, such models are limiting because they rely on a page refresh occurring.

He noted that there are other methods for analyzing the effectiveness of streaming video, Flash applications or podcasts. For example, Google's event tracking, which launched in October, can track traffic interactive elements on a site.

A site that is advertising a movie with a series of 100% Flash Web sites can use event tracking to monitor how many users downloaded a desktop widget from that site after arriving, or how many people click on a link to order tickets to the movie, Kaushik explained.

Still, he dismisses "engagement" as a "crappy metric." Web site operators should shy away from engagement and instead focus on concrete metrics, like how many users read or commented on a blog, how many users have watched 35% vs. 100% of a video, or how many rewind a portion of a video, Kaushik said.

"I am not saying don't create engaging experiences," he added. "[Just] don't use the term engagement, because it has been bastardized to the point that it doesn't mean anything."

Quantitative Web analytics tools can tell a company the degree of engagement with elements on a Web site, but by themselves, they can't truly measure engagement, Kaushik argued. For example, he noted, just because users saw 16 pages of content, that doesn't mean those users had a positive experience. It could mean that the site has poor navigation. And just measuring the number of comments a blog generates doesn't take into account the sentiments contained in those comments, he added.

Instead, companies should use additional qualitative tools like site surveys, and market research to determine whether users had a negative or positive experience on a site.

Nielsen's Ross noted that while the company is satisfied for now that time spent is one of the best ways to measure the effectiveness of Web sites, it is open to re-evaluating its strategy as technology evolves.

"Can the industry move to a case where we decide certain in-page actions are representative of an advertisement opportunity?" Ross said. "We might get to a point where it's not really a page view anymore, it is an ad opportunity. As there are ad opportunities driven by in-page activities like watching a video -- where do we go with that? That is an interesting challenge that we see."

For example, if the industry could agree on new metrics, such as the number of comments added to a blog entry, Nielsen may consider adding that as a metric, Ross noted. In addition, Nielsen is looking closely at ways to measure content contained in widgets that may be embedded in multiple sites. Perhaps multiple sites would get credit for that content, he suggested.

"We're always considering what other needs there might be, [but] we don't see one today with a dying need to create new metrics," Ross said. "By and large, I'd argue that time spent is a good, impartial and broad view of engagement."

Copyright © 2008 IDG Communications, Inc.

  
Shop Tech Products at Amazon