Aiming at Perez, Gawker spits in the wind

Gossip & media blog pioneer Gawker has posted a long and winding exclamation about PerezHilton‘s “seemingly astronomical traffic numbers.” (Here’s the post.)

While there’s lots of bated-breath innuendo, Gawker’s analysis boils down to:
a) gee, 100 million impressions a month sounds like a lot. (At least ten times Gawker.com’s traffic, apparently.)
b) web analysis firms Nielsen/Netratings and Comscore guestimates of Perez’s US-only page impressions are lower.
c) gee, 100 million impressions a month sounds like a lot.
d) extrapolating from Comscore’s tally for unique visitors (US and abroad?), you’d need fully 26 visits a month to justify the 100 million impressions.
e) the only source for Perez’s traffic is Blogads.com.

Some Gawker commenters were bemused: “i feel like you just tried to explain wikipedia to my mom. why is this here?” Sensing simmering jealousy, another offered a hug: “Gawker, you are SO hotter than Perez. He may get more page views, but you’re way better looking. And a better cuddler, too.” Another sympathised, “This is some real inside baseball. My blog gets 200 visitors and 300 page views a day. My head hurts imagining all those other zeroes.” One fingered Gawker’s owner, Nick Denton, as the author, rather than the bylined “Doree.” (I couldn’t find Doree among Gawker’s Facebook peeps.)

The bottom line is that Perez’s page impressions are real. He’s now doing 5 million impressions on good days. That’s what our servers, which tally each time we serve a ad-filled javascript to a human who visits his pages, show.

Gawker writes: “…just how does Perez manage to rack up even that many page views anyway? By the ComScore numbers, each visitor would have to be looking at about 26 pages…”

Gasp. Think 26 visits per reader per month is unbelievable? Sadly for America’s employers, it’s likely many Perez readers refresh 26 times a week. Perez is addictive and his readers are insanely loyal. Ask your Perez-reading colleagues how many times they check the blog. Per hour.

So, what’s up with Comscore and Nielsen? As Gawker notes, their estimates are only for US traffic. Extrapolating from the lowest estimate, Gawker comes up with 55 million global impressions. (Gawker neglects to extrapolate from the higher estimate, Comscore’s, which would yield 80 million global impressions monthly.)

Why give credence to two services who have a nearly 50% variance between their estimates — Nielsen/Netrating’s 33 million US impressions versus Comscore 48 million? As opposed to Blogads servers’ methodical bean-counting, their traffic estimates are based on extrapolations of extrapolations. Why extrapolate from a sample only or 30,000 or 2 million people, when you can instead count the real thing?

We know that Comscore and Nielsen’s numbers are overweighted by at-home audiences, since that’s where they find folks willing to take their surveys and/or tolerate their traffic-tracking software installations on their computers. This at-home skew leads to lots of problems because, as our server logs show, the vast majority of blog reading gets done during the workday. (Like you, right now, right?)

Further discrediting their numbers, both Comscore and Nielsen think more than 50% of Perez’s readers are men, which is ridiculous. Do a quick survey in your office and you’ll find that 90% of his readers are women. And very likely 26 years old, plus or minus three years. And that’s exactly what our recent survey of Perez’s readers found.

Trying to deflate Perez, Gawker also hawls out Technorati’s “top 100” list, noting that Perez at #22 is lower than some smaller blogs. Technorati rates blogs on inbound links. How can you keep a straight face quoting Technorati, which lists one Beppe Grillo‘s blog as #16? (Ahead of Gawker itself at #30!)

Gawker’s final evidence that Perez’s popularity can’t be trusted is that Blogads is the “only one whose statistics seem to align with Perez’s.”

Ahh, that unseemly “seem.” This “seem” is easily fixed. Gawker Media relies on SiteMeter for its own stats, so I assume it thinks these are reliable. (They are here.)

If any Gawker staffer wants to put his/her money where her/his snark is, I’ll bet dinner in the Soho restaurant of her choice that Sitemeter validates Blogads monthly impressions rather than Comscore’s or Nielsen’s. Or, if you prefer, that in coming weeks Perez can match a month of Gawker.com traffic in three days or less.

A footnote for insider blogball fans: years ago in Budapest, Nick and I covered the same business beat as journalists, he for the FT, I for pubs like the Herald Tribune and Euromoney. Nick later introduced me to blogs, in the form of ObscureStore which has remained my favorite. And I met my wife through Nick. So I’m amply amused to be tilting with Gawker about business numbers.

Update: Andrew Krucoff does some digging with a Gawker staffer and declares “The entire crux of their argument which seems to be ‘Blogads numbers are bullshit’ falls apart on a painless no-ringing, low-vibrating note.“