People are trying to figure out why the Alexa statistics for a bunch of sites (including kottke.org) jumped sharply in mid-April. I don't buy the Digg explanation (for one thing, the timeline is off by a month)...it's gotta be some partnership or something that kicked in. Or how about Alexa's "facelift" on April 11?
There are 11 reader comments
• Jul 05 2006 • 12:07PM
Alexa has never made any sense for my site. The "people who link here also link to..." selections have not changed in years, although the thumbnail is relatively recent, so Alexa must be monitoring my domain. However, they show a 50% decrease in traffic for my site, which is not supported by any other source. My traffic has been flat for about two years. Maybe they tweaked the algorithm that calculates traffic?
• Jul 05 2006 • 12:15PM
Alexa started using "powered by windows live" graphics in late April. Maybe they started getting some extra traffic data from microsoft.
• Jul 05 2006 • 12:45PM
When did they start offering the Firefox plugin? If the timing is right (and if people are actually using it) the sites whose stats I've seen jump would certainly have a large FF audience.
• Jul 05 2006 • 1:05PM
A terrific MetaTalk thread about the phenomena.
I'm pretty sure it was a handful of duggalos installing the spyware, independently of each other, acting on digg's obsession with the stupid Alexa rankings. It wouldn't take very many diggtards to freep Alexa.
• Jul 05 2006 • 1:22PM
That is not a terrific MetaTalk thread. As I said in my post, the date of the Alexa post on Digg is off by an entire month with that of the spike.
• Jul 05 2006 • 2:56PM
Did you read the comments? That's kind of the point, you know. Digg's fascination with Alexa is not limited to that post on 3/19. They're obsessed.
Alexa's userbase is miniscule. The sites affected by the spike are primarily digg and slashdot, with sites that are read by some of that population (like boingboing, metafilter, and you) affected too, but nowhere near as much.
All sites not frequented by the geek-wannabe contingent were not affected at all. This is consistent with a handful of these people installing software that reports to Alexa.
• Jul 05 2006 • 5:10PM
Alexa numbers don't correlate with reality. Including Adbrite.
I wonder if it was a server "upgrade"? New sniffer/spider?
• Jul 05 2006 • 5:23PM
Building on the Powered by Windows Live...perhaps they are getting RSS subscription data from those who use the Windows Live homepage? Digg, Slashdot, and this site all have many users that use other sources to read the posts. Maybe Alexa accounted for how many people read it on Windows Live, then extrapolated to account for other RSS readers?
• Jul 05 2006 • 10:31PM
The Alexa rankings remind me of Amazon Sales Rank numbers and their correspondence to sales. If only there were a highly reliable way to map them back and forth.
• Jul 08 2006 • 12:15AM
I totally blundered the dates, but there are still a few interesting things going on here:
1) The fact that this spike occurs over the span of a day and not weeks or months suggests that there was some single spontaneous event that caused the whole thing (rather than a months worth of digg community obsession over Alexa). Was this a user driven event where something caused a bunch of tech oriented folks to install the toolbar or was it Alexa "balancing" their statistics to give more weight to an underrepresented tech crowd?
2) If it was an Alexa-side rebalancing, what could they have based this calculation on? Maybe they partnered with a few large sites or have access to ad service data to compare actual site users against the site's alexa count. It seems that if this were the case, they would have announced something, though... and why wouldn't they reinterpret the historical data to provide more coherent information? This leads me to continue to suspect it was a user caused event.
3) Why did digg get a higher benefit from the reranking than other sites? Are there other sites that benefited even more and what kind of clues could this provide?
Anyway, sorry for the horrendous date mixup. I'm really looking forward to someone figuring this thing out.
This thread is closed to new comments. Thanks to everyone who responded.
Paul Wilson • Jul 05 2006 • 11:38AM
Definitely not buying the digg explanation either. Alexa has always been fishy to me...