Mull-it Over with Alexa: Bias Upfront, Party in the Back
Fred Wilson, the AC/DC of VC bloggers (my highest compliment), I'll see your nifty Alexa tech-bias analysis and raise you an even more glaring Gawker Media-sized example. Let's look at Silicon Valley tech tweenster Valleywag and the hepcat of hubcap-lovin' car junkies Jalopnik which provide the benefit of publicly available sitemeters.
Valleywag
Jalopnik
So Jalopnik gets 3-5 times more traffic (beep beep!) than Valleywag, right? Now here's what that roadwhore Alexa has to say about the size of their junk in the trunk.
That's quite a throbbing blue vein in the backseat of Valleywagon while Jalopnik's flat redlining can't even bump-draft its way off the cement blocks. That Dodge don't hunt.
Server logs are the obvious answer as long as you're sticking to pageview count which is the primary measurement dictating the online ad market (unique visitor analyses, best derived from representative panel-based methodology, are more useful for drilling down to audience usage patterns that answer other questions) except not everyone makes their internal stats package public for open comparison. Until that day comes, Fred is absolutely right in saying to obey the yellow lights with all third-party traffic monitors.
Of course, if you have the balls to match your bearings, go ahead and talk trashly but carry a big sitemeter.
Related: Pageviews Are Obsolete [evhead]
Valleywag
Jalopnik
So Jalopnik gets 3-5 times more traffic (beep beep!) than Valleywag, right? Now here's what that roadwhore Alexa has to say about the size of their junk in the trunk.
That's quite a throbbing blue vein in the backseat of Valleywagon while Jalopnik's flat redlining can't even bump-draft its way off the cement blocks. That Dodge don't hunt.
Server logs are the obvious answer as long as you're sticking to pageview count which is the primary measurement dictating the online ad market (unique visitor analyses, best derived from representative panel-based methodology, are more useful for drilling down to audience usage patterns that answer other questions) except not everyone makes their internal stats package public for open comparison. Until that day comes, Fred is absolutely right in saying to obey the yellow lights with all third-party traffic monitors.
Of course, if you have the balls to match your bearings, go ahead and talk trashly but carry a big sitemeter.
Related: Pageviews Are Obsolete [evhead]
