Google PageRank and the local mediasphere
Local wonk David Brauer set off a minor shitstorm among Minnesota's media cognescenti yesterday with a post on his well-read blog at MinnPost.
Innocuously titled "Fun with Google PageRank," Brauer published a chart that seemed to turn local media on its head. Through this lens, MinnPost was now as big a player as the Star Tribune, after just one year and with a tiny fraction of the staff (and both were as big as the Drudge Report). Equally surprising, relatively smaller, lesser known blogs were placing higher than local TV stations, City Pages, and even the Pioneer Press. Unable to resist a touchdown dance, Brauer even listed his BrauBlog as the equal of WCCO-TV, Fox9, and KARE11--stations with far greater budgets than his humble one-man operation.
How could this be?
As Brauer was quick to correct several intelligent local observers who nevertheless leaped to wildly inaccurate conclusions after reading his post, this chart does not represent traffic. It is instead a list by Google PageRank, from highest to lowest.
What's a Google PageRank? And how is it related to pageviews--the most important metric for media companies that rely on advertising rather than donations? Even Brauer doesn't purport to know:
Google PageRank is one of those black boxes I've never fully understood, but it basically measures how many people link to you, rather than pure readership.
OK, that's a good starting point. Still, why would MinnPost have a higher Google PageRank than the Pioneer Press, which dwarfs the scrappy start-up's daily traffic?
Google PageRank refers to the strength of a page, not a site. If your homepage is a PageRank 6 then you would expect subpages within the site to have lower PageRank.
Translation: Only links to the homepage count on Brauer's chart. Few people link to the St. Paul paper's homepage at twincities.com because they can just link to the article of interest (as Brauer himself does in his Daily Glean), which doesn't benefit twincities.com's Google PageRank much if at all (though this last sentence will certainly benefit the paper's Google PageRank--no need to thank me, though a free subscription would not go to waste).
Same goes for MinnPost, of course. But clearly an outsized amount of people are linking to MinnPost's homepage. Why would that be?
I'll hazard a guess: MinnPost has been in the news a lot lately, receiving regular mention on media gossip site Romenesko and even appearing in the vaunted New York Times several times. These national observers aren't so much interested in MinnPost's reporting and writing but rather its business model: journalists are looking for a way to sustain good reporting with a revenue stream not dependent on advertising, and MinnPost is one of the first major attempts to do that, relying on donations and grants from foundations. Reporters are naturally interested in ideas for keeping their paychecks.
Which brings us to the second reason for MinnPost's success in this metric: Google PageRank grants a much higher amount of juice to links from major sites like the New York Times than from smaller Midwestern blogs. As discussed in this page Brauer linked to:
Google PageRank assumes that a page is important if another important page links to it. So if your page is linked from a BBC News article, for example, that link will give your site a 'vote of importance'.
Under Google's logic, one link from the Grey Lady may be worth more than 100 links from local bloggers/readers. Which means sites with millions of local readers can be dwarfed on Brauer's list by a much smaller site with one reader at the New York Times who happens to link to their homepage in an article about the future of journalism.
(Ironic sidenote: This also helps explain the Star Tribune's high ranking. It too has received a lot of links to its homepage lately from the New York Times and other major national media with a lot of Google PageRank juice. Except all those stories are about how the Strib is a failing business in bankruptcy--more evidence that Google PageRank isn't necessarily a good metric of a site's economic or journalistic success).
Deeper into the same story, sites are even instructed on how to inflate their Google PageRank through a variety of means:
First, you can try to get external sites to link to that page. Secondly, you can get lots of the other pages on your site to link to that page, in particular pages that themselves have high PageRank, such as your homepage.
The practice is known as "link trading." I say more power to 'em, but City Pages is more interested in growing pageviews in order to continue serving our readership and allowing advertisers to reach the customers they're interested in. That's why nobody has ever mentioned "Google PageRank" to me until yesterday, when I read Brauer's post, but I hear about our growth in web traffic once a week at our pitch meeting as well as receive a monthly status report. In the interest of transparency, I've asked Web Editor Jen Boyles to start sharing that information with readers in a regular Blotter post she's calling "The Traffic Jam." The first installment debuts Friday.
P.S. In fairness, I should note that savvy readers were able to read between the lines on how Google PageRank operates, and what it means, as evidenced by this comment by Heckman (aka Ryanol) on this MnSpeak post, which is far more succinct than my attempt to explain it:
I think what DB [David Brauer] is referring to is actually tool bar page rank.
Actual pagerank is the secret sauce.
That said TBPR [Tool Bar PageRank] has been devalued as of late because it's main use has been as a valuation tool in the exchange link market.
I wouldn't put too much stock in it.