When you log into Blogger, there's a nifty new section for traffic statistics. Cool.
But I don't understand why Blogger reports almost twice as many pageviews as Sitemeter. That would be very nice indeed, but Sitemeter is reporting a shockingly humdrum level of traffic.
Anyone have any idea why the discrepancy?
7 comments:
My guess is Sitemeter requires the HTTP client to follow a URL off the main page, something RSS readers, crawlers, and other non-browser thingy's won't likely do without good reason. But since Blogger can actually get the hit rate of you page at the HTTP server level, it can record each and every time a page is viewed regardless of client type.
No one knows.
Really.
There are LOTS of different methods to measure pageviews. How many you "count" depends a lot on the method. Consider:
1) A javascript fragment that calls back to the stats server when the page is loaded. Only counts pages loaded; misses page loads by browsers that don't support javascript or have it turned off. Gives somewhat unrealistically small numbers.
2) Have the web server count the number of times a particular page is loaded. Counts LOTS of hits from things that aren't humans (search engines, spambots, sometimes counts images or page "fragments" as part of a page). Gives unrealistically large numbers.
3) Add an image to the page which is loaded from the stats server. A little bit of a middle ground; catches people without javascript but with images turned on. Misses people with images turned off. Gives numbers close to 1.
To add to the confusion, some sites cache web access (turning multiple hits invisibly into a single hit or even no hit at all). Some browsers don't fully load a page (is your stat counter at the beginning or end of the HTML?). Some browsers preload pages the user never saw. Some stats packages make an effort to filter hits from search engines and spambots. Some spambots deliberately load pages in order to garner traffic from the blog admin looking at their referrer logs. Some stat packages try to filter multiple hits from the same user to the same page. Some blog packages or templates seek to refresh the page every few minutes to generate more "page views" than actual human attention warrants. Some readers (RSS/Atom readers) won't load images or javascript in your standard blog template, depending on what is included in your feed data.
What TriggerFinger said.
Server side counters will always be more accurate.
Did you turn off the "track your own pageviews" button?
One visitor my visit multiple pages. Your pageviews should always be more than your visits. It's a good thing. It often means new visitors are interested enough to read several things that you have written. Those that hit several pages in one visit are more likely to come back.
Simple, really. They count the front AND BACK of each web page!
And you call yourself a blogger!
Sheesh!
Maybe they mixed up my stats with yours. Mine have been higher than average and I think I've been off my game.
That would be assuming I was ever on game. Or in the game.
I've used statcounter and sitemeter and they never agree on the numbers with statcounter consistently 5-10% higher.
Post a Comment