Forum Moderators: phranque

Message Too Old, No Replies

Stats Software

         

erobs

10:32 am on Apr 21, 2005 (gmt 0)

10+ Year Member



What stats software do you use that is the most accurate. I have a website that on Awstats, it shows about 8000 impressions per day. Webazlier shows about 800 impresssions per day.

Both stats programs show about 600 unique visitors per day. I would guess that webalizer is closer to accurate.

Does anyone have a suggestion on stats software to give me an accurate count?

netguy

12:59 pm on Apr 21, 2005 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



I have seen a lot of deviation across various log analysis packages, but an impressions disparity of 800 - to 8000 - is ridiculous.

Is it possible you are looking at 'hits' (the file and every image) on Awstats? If you have an average of about 9 images per page, that would account for the discrepancy.

Also, look in Awstats for 'pages per visit.' If it is showing 600 uniques, and something along the lines of 1.25 pages/visit, then the impressions count is obviously more like 800 rather than 8000.

Steve

Unless the mods move this for you, you'll find a lot more on logging/analysis in Forum 39:
[webmasterworld.com...]

erobs

2:08 pm on Apr 21, 2005 (gmt 0)

10+ Year Member



Thank you for the reply. Here is the info from Awstats:

Reported period Month Apr 2005
First visit 12 Apr 2005 - 09:58
Last visit 21 Apr 2005 - 09:38

Unique Visitors 3025

Number of Visits 3611

Pages 40463

Hits 62945

Webalizer Same time period

Visits 3220

Pages 5052

Files 54825

Hits 69014

There is the info from both programs. Not sure what to account for the difference and not sure what Webalizer is counting as Files vs Hits. I know hits is usually everything including all images. I do have a few banners on each page, so not sure if Awstats could be counting each page as more than one because of that.

monkeythumpa

12:28 am on Apr 22, 2005 (gmt 0)

10+ Year Member



Some stats pacakes count each file-call a pageview. So if one person viewed one page on my site, some stats would count that page, the css file, my header include file, my footer include file, and each graphic on the page a pageview. You can see why a discrepency of 10x is not that out of the question, even 100x is not rediculous. Either you have to figure out what each stat package is counting and do the math, or define what to count and what to ignore.

Then you have to account for caching . . . Ug!