Alexa: Eng and Giz, Sept07
Gotta love Denton’s showmanship and ability to spin, the man’s a master of what he does. Unfortunately, in his post today about Gizmodo overtaking Engadget, he got a number of facts wrong, not the least of which being that Engadget’s moved to the number two slot. Granted, that mostly had to do with the fact that all our traffic systems are internal, so he, like everyone else, only has tools like compete.com and Alexa to work with. (Actually, he probably didn’t want to draw any attention to Alexa — we all know it’s inaccurate, but right now it shows Engadget as being massively ahead of Gizmodo in page views.) The bottom line is: we don’t participate in compete.com, so like every other traffic estimating tool out there, it just makes educated guesses that are often totally off. Specifically, according to Denton:

“In August, according to Sitemeter, Gizmodo did 52m pageviews.”
This is true, Gizmodo did that much traffic — across Gizmodo’s entire international network. The way Sitemeter works is: you put a Sitemeter script on every page you want to be counted. If we put Gizmodo’s Sitemeter on our page, their meter would include our traffic. So 52m is the number across ALL Gizmodo translations. The real number Gizmodo.com (i.e. the US site) did in August, their best month to date, was 42m. That number is really awesome for any site — props to Brian.

Still, we did millions of pages more on our classic site, and including Engadget Mobile, HD, Japanese, our two Chinese sites, and our Spanish site, some of which had a record month in August, the number still doesn’t stick. In other words, the assumptions Nick makes about our stats are way off, and both domestically and internationally we were ahead in August. This month, US site to US site, we are ahead by well over 60% of Gizmodo’s traffic, as well.

“Yesterday, in coverage of Apple’s new iPod, Gizmodo was minutes ahead of Engadget at every step.”
Eh, that much I can’t speak specifically to. It’s easy to change the time stamps on posts and make claims that this or that site was first — it’s not as easily quantifiable. Both sites have learned how to get news up almost instantaneously, which is amazing for our respective readerships. Still, we were all pretty involved in what was going on and so far as I can tell we were neck and neck and nailing the news first on as many — if not more — stories than our competition. If Techmeme and Digg were anything to go by, our coverage was either first, better regarded, just plain more linked to, or some combination therein.

“Engadget had… in AOL, a corporate parent which funneled through new visitors.”
Oh Nick, how I wish. According to our internal traffic systems, yesterday, our second biggest day for traffic ever, AOL referrers accounted for 0.2% — yes, less than one half percent — of our total page views. On our biggest day ever, the day the iPhone was announced, our AOL referrers accounted for a record 2.1%; in all of August, they represented less than 0.5% of all traffic. The vast majority, I believe, are just coming from AOL search (which uses Google, where we’re routinely higher ranked than our competition), too — not from promo.

Obviously this isn’t something we’re super proud of, but figuring out how to better integrate Engadget and push the AOL firehose on us is something that we’re hard at work hammering out with the gang in Virginia. But to imply AOL is Engadget’s crutch and despite our best efforts we can’t keep up with Gizmodo, well, that’s patently untrue.

Of course, again, most of this has to do with the fact that Denton doesn’t have access to our traffic systems, so he goes by the best guesses these estimating services can make, and then spins them to his own liking. Unfortunately those guesses just aren’t accurate — but it’s not too hard to see through the spin. That’s not to take anything away from the job Gizmodo is doing, though — they’ve really differentiated themselves, narrowed the margin, and who knows, maybe the next time Denton posts about how Gizmodo has passed Engadget, it might actually be true.