Share This
Metrics
May 6, 2009
Research compiled by Clyde Bentley
Life was much simpler back when I thought "metrics" were foreign inches. The metrics we use to gauge Web success, well that's a much knottier subject for all of us in the news world.
Newspaper chicken-and-egg: For the past few years, most newspaper managers have conceded that online sites are the drivers of our news products. After all, those visitor counts and page views keep going up. All of that was thrown into question when Nielsen Online reported that its metrics show traffic to the new all-online Seattle Post-Intelligencer actually fell when the paper version was closed.
Wags saw this as clear evidence that online is not the king we thought and print drives digital. But then PI's Ed Steenman came back with statistics showing page views for the PI are actually up. Maybe not in Nielsen's computer, but certainly by the Alexa measurements.
Since then, the research, newspaper and advertising discussion sites have been filled with discussions about how we measure the Web and how much or little we really know.
It gets worse: Martin Langesveld of the Neiman Journalism Lab had earlier calculated that only 3% of newspaper reading was done online if you compared the NAA's figures for print readership and Nielsen's online page views. Last week he kicked up a firestorm with figures he said showed that newspaper online readership was piddling compared to MSNBC, CNN and Yahoo! News.
Back to the basics: So why should a newsroom denizen care? This is all about ads, isn't it? I wish it were so, but in today's climate knowing the right online metrics can keep your job. Those numbers represent our audience and the justification for our newsroom efforts.
Unfortunately most working journalists I know are still talking about the "hits" on their site. In the cyber marketing world, that's a laughable term. I refer my students to Google's metrics genius Avinash Kaushik, who says it is an acronym for "How Idiots Track Success." Hits just count miscellaneous cursor movements around the site.
Much more important are counts of visitors and their views of pages, but even those are confusing numbers. "Unique" visitors and unique page views try to track the action of one person, rather than crediting that person each time she or he clicks back on the page. And then there is "bounce," a count of how many visitors immediately left the site. Kaushik calls it "I came, I puked, I left." Last year Oreilly assembled a fairly clear primer on metrics that was specifically aimed at publishers. But newsroom types are also welcome.
What counts?: The fact that computers could tally points whenever someone called a news story to their screen was at first extremely appealing to both marketers and editors. But all that counting relies on identifying the computers at both ends of the circuit. As it turns out, that's not as easy at it might seem. Many counting systems rely on "cookies," bits of text placed on your hard drive when you visit a site. But cookies can and do get deleted by users. A 2006 study by comScore showed that 31% of users delete their cookies at least once a month and 7% wiped cookies at least four times a month. Gian Fulgoni from comScore estimated that common Web metrics are inflated by 150%.
Season tickets: Then there is the "stadium effect," described by Greg Harmon in 2004. Greg was then director of interactive services for Belden Associates and was concerned that Web analytics produced radically higher readership numbers than the direct surveys the company did. In a paper I co-authored with him, he likened it to a baseball stadium that counts attendance via a turnstile. Over a 10-game season, a 10,000-seat stadium could record 100,000 fans through the turnstile. But if the team had 9,000 season ticket holders, the actual season attendance could be as low as 19,000. Web sites have the same math problem if they continually count their most loyal readers as new visitors, which they do if cookies are cleared. Belden compared Web logs and direct surveys and found evidence that the traditional counts may be too high.
So what?: The upshot of the confusion over metrics is that ad buyers are taking a harder look at our numbers as the recession squeezes their budgets. The 2009 first quarter portion of thee-tailing group's 8th Annual Merchant Survey showed a third of online merchants predict their revenues will be down or flat this year. And 92% of those online merchants said the primary tools for making their ad buying decision are Web analytics ? a fancier name for metrics. And a related study showed that marketers think that the top challenge they face as they try to integrate traditional media and online media is having adequate metrics to guide them.
Advice for the Metric-worn: I tell my students and my colleagues that the best defense in the face of conflicting numbers is good reporting skill. Reduce the numbers to a story and look for the logic - or lack of it. A few years ago I received a note about a publisher who was ecstatic that his paper's online site had several times more visitors than his state had people. Something was wrong there that even a cub reporter should have questioned.
A simple test of readership is to compare stories in the print edition and the Web edition of a paper. I like to look at the top local story in the print edition, then the single-day unique page views for that story. Unique page views are not always accurate, but they give a reasonable estimate of readers who actually looked at the full text of that story. With the exception of disasters and other major news events, I find this figure about the same or lower than the daily circulation of the paper. Try it during a routine news cycle and let me know how well it works. It is akin to another metric used by the advertising world - click through rate.
Meanwhile, back at the cell phone: Like many researchers, I have been trying to look ahead and see where newspaper-style journalism will find a home in the next few decades. Online is certainly an option, but for several years I have been particularly curious about the potential of cell phones.
In August I will begin a full-academic-year fellowship at the Reynolds Journalism Institute, where I will try to develop a mobile strategy that even small newspapers can use. More than 80% of Americans carry cell phones, so they are the one communications device that is most often within easy reach. My job will be to see what it takes - technology, content tweaks, financial incentives - to give let readers pull a newspaper-like experience from their pockets or purses.
The fellowship means I will put away my PowerPoints and turn my courses over to other professors for the year. But I will still produce Research for the Newsroom about every two weeks while exploring the cell phone world.
And before you ask, I'm not ready to try texting this to you.
Contact Me:
E-mail - bentleycl@missouri.edu Twitter: http://twitter.com/MizzouBentley