Answers
Sep 08, 2006 - 07:08 AM
Apparently others have experienced the same:
http://www.seomoz.org/blogdetail.php?...
Still don't understand why, though...
http://www.seomoz.org/blogdetail.php?...
Still don't understand why, though...
Sep 11, 2006 - 05:03 PM
Hey, Dude...
It looks to me like there's a difference in the "formula" that is used between the two - which, incidentally, also differs when using WebTrends which is another stat tracker that I've used on sites in the past. It also looks like the filters and enabling/disabling local traffic monitoring on the site will make a difference in the the numbers generated by Analytics. Granted, it's not much...in my case it made about a 5% increase which brought it closer to WebTrends, but still a good deal different.
I don't really have any suggestion past checking the filters to make sure you're not inadvertantly taking out valid "views".
It looks to me like we've got three different companies who are trying to sell us three different versions of "blue". They all have a different interpretation of just what "blue" is, and we really don't have any way to dispute their interpretation.
It looks to me like there's a difference in the "formula" that is used between the two - which, incidentally, also differs when using WebTrends which is another stat tracker that I've used on sites in the past. It also looks like the filters and enabling/disabling local traffic monitoring on the site will make a difference in the the numbers generated by Analytics. Granted, it's not much...in my case it made about a 5% increase which brought it closer to WebTrends, but still a good deal different.
I don't really have any suggestion past checking the filters to make sure you're not inadvertantly taking out valid "views".
It looks to me like we've got three different companies who are trying to sell us three different versions of "blue". They all have a different interpretation of just what "blue" is, and we really don't have any way to dispute their interpretation.
Sep 12, 2006 - 06:07 AM
Thanks, rcastagna.
As far as I could read it also boils down to a tag-based versus a log-based way of measuring.
In Google Analytics you include a script on each page, which could be disabled if javascript is not enabled on the users browser, if the page doesn't load completely etc. In Webalizer it takes the stats directly from the logfile generated on the server.
There also seems to be a difference in the crawlerbots that they're including or not per. default as visits.
And finally one visit is not counted equally always. Sometimes if the same person is back within an hour, it's counted as the same visit in other statprograms it's per 30 minutes etc.
What is quite strange though is that I'm using Google Adwords for the site, and the amount of clicks that Adwords tells me I have to pay for, is higher than the amount of clicks registered by Analytics as coming from the Adwords campaigns!
I guess my conclusion is to see all this as trends and if I really wanted a number on it, I should do a triangularization as was suggested somewhere, i.e. taking the average result from three different statistic programs.
As far as I could read it also boils down to a tag-based versus a log-based way of measuring.
In Google Analytics you include a script on each page, which could be disabled if javascript is not enabled on the users browser, if the page doesn't load completely etc. In Webalizer it takes the stats directly from the logfile generated on the server.
There also seems to be a difference in the crawlerbots that they're including or not per. default as visits.
And finally one visit is not counted equally always. Sometimes if the same person is back within an hour, it's counted as the same visit in other statprograms it's per 30 minutes etc.
What is quite strange though is that I'm using Google Adwords for the site, and the amount of clicks that Adwords tells me I have to pay for, is higher than the amount of clicks registered by Analytics as coming from the Adwords campaigns!
I guess my conclusion is to see all this as trends and if I really wanted a number on it, I should do a triangularization as was suggested somewhere, i.e. taking the average result from three different statistic programs.
Sep 12, 2006 - 01:32 PM
That makes a lot of sense...and I didn't even think of it. Webtrends is a log-based analyzer, as well, which would reinforce that theory.
Really, I can understand why Analytics doesn't include the crawler bots...I've "excepted" most of them from my Web Trends log, because you really can't consider a bot a visit since no "eyes" are being put on the page.
As for it considering a repeat visit as 1hour instead of 30 minutes, I think 1hour is a bit too generous...especially with the big ISP's like AOHell that recycles IP address pretty frequently.
I haven't started using Adwords yet, but that is a very curious point. You would think that those numbers would coincide with each other.
Well, after all, we're just after statistics...and maybe you've heard the old adage: "There's three types of lies. There's lies, there's damn lies, and there are statistics."
Good luck!
Really, I can understand why Analytics doesn't include the crawler bots...I've "excepted" most of them from my Web Trends log, because you really can't consider a bot a visit since no "eyes" are being put on the page.
As for it considering a repeat visit as 1hour instead of 30 minutes, I think 1hour is a bit too generous...especially with the big ISP's like AOHell that recycles IP address pretty frequently.
I haven't started using Adwords yet, but that is a very curious point. You would think that those numbers would coincide with each other.
Well, after all, we're just after statistics...and maybe you've heard the old adage: "There's three types of lies. There's lies, there's damn lies, and there are statistics."
Good luck!
Sep 19, 2006 - 07:15 AM
sorry, i forgot to close the question.
Thanks for the help and yes, I like the phrase you quoted in the end, it seems very appropiate in this case as well, hehe.
Thanks for the help and yes, I like the phrase you quoted in the end, it seems very appropiate in this case as well, hehe.
Add New Comment