Part I: Hits & Page Views
Methodology
- We’re going to do this ordered similarly to an AWStats report, for good reason: Overall, the AWStats report is more linear than Google Analytics. Furthermore, Analytics provides many more options than AWStats, so we’ll default to the lowest common denominator for reports: AWStats.
- The sites in question receive a good amount of daily visitors, search engine traffic, and page views. Obviously, the amount of difference will vary depending on traffic. But I consider these to be moderate-traffic sites, one content-based and the other a forum site.
- I’m sure someone (several people?) will say we’re not comparing apples to apples here. I’ve done my best to isolate the apples and compare them (in this case, what both programs report as “visits” and “pageviews”). In the end, however, we depend on both pieces of software to tell us how much traffic is coming to our sites. So it’s important to know if there’s a difference, and how wide the difference is.
Visits and Page Views
These were taken from the AWStats “Summary” report and the Google Analytics Overview report.
SITE 1:Number of Visits
AWStats: 8324
GA: 2860
SITE 1:Pages
AWStats:39,506 (4.74/visit)
GA:16,128 (5.64/visit)
SITE 2: Number of Visits
AWStats: 27,286
GA: 4114
SITE 2: Pages
AWStats:210,680 (7.72/visit)
GA:30,237 (7.35/visit)
Visits and pageviews by day
For simplicity’s sake, I drew a sample from the 1st, 5th, 10th, 15th, 20th, 25th, and 30th of the month to compare stats.SITE1: (visits/pageviews)
- Dec.1
AW: 303/1193
GA: 94/494 - Dec.5
AW: 313/1366
GA: 97/503 - Dec.10
AW: 267/1023
GA: 88/441 - Dec.15
AW: 262/1288
GA: 81/522 - Dec.20
AW: 262/3573
GA: 96/543 - Dec.25
AW: 225/859
GA: 79/515 - Dec.30
AW: 262/1163
GA: 95/546
SITE2: (visits/pageviews)
- Dec.1
AW: 993/7842
GA: 742/5634 - Dec.5
AW: 1008/9186
GA: 826/6264 - Dec.10
AW: 801/9100
GA: 630/4980 - Dec.15
AW: 940/6934
GA: 666/4856 - Dec.20
AW: 956/6733
GA: 677/4152 - Dec.25
AW: 627/4293
GA: 437/2713 - Dec.30
AW: 908/7800
GA: 649/4846
Conclusion:
Frankly, I’m amazed at the huge variation between these two statistics programs. It’s like they’re not even measuring the same site!
Now, apples-vs.-oranges discussions aside, we look to these pieces of software to tell us how many visits and pageviews our sites are getting. That’s not a hypothetical question- there is a concrete number out there.
Given the chasm between these results, I have to think that one of these two sets of results is very, very wrong. Question is, which one?
Coming Soon:
I’ll continue the comparison of AWStats vs Google Analytics and compare some of the following:
Countries
Visits Duration
Pages-URL (Top 25)
Operating Systems
Browsers
Search Engine Traffic
Links from an External Page
Search Keyphrases
Search Keywords
Miscellaneous
I have looked using Google AdSense and another tracker program. I think what you are seeing is how many crawlers are moving through your site - and they’re not all search engine crawlers. IN AWStats, do you see some sites where the pages and files read are exactly the same? Those are crawlers.
Whenever javascript is turned off, as the robots do, you won’t have a count. In my opinion, the Google number is more worthwhile, even though it shows less traffic.
Geoff Fox
Left by Geoff Fox on January 2nd, 2006
[…] Deres siste post er en sammenligning av antall besøk og unike besøkende mellom Google Analytics og Awstats. Mitt eget inntrykk er at AwStats er mer off target en jeg trodde de var tidligere og at Google Analytics virker meget korrekt. […]
Left by Web Analytics Guide - Søkebloggen on January 3rd, 2006
Wow, the number is quite suprising indeed. I accepted it to differentiate by 500-1000 or so.
Left by mark on January 21st, 2006
It’s because AWStats tracks bots also. Each time a page request is generated (whether it be by a human or by a bot) AWStats logs it.
Since google analytics uses javascript to track hits, only real browsers show up (since bots can’t parse javascript) causing the huge difference in numbers.
I’ve actually compared some sites and i’ve found that the applications that track the bots can actually inflate your hits by 75-80%. Yeah, it’s scary.
Left by Patrick Weber on March 1st, 2006
Oh yeah, so to answer your question, both are absolutely right.
Left by Patrick Weber on March 1st, 2006
I see huge variances between the two as well. I tend to trust the stats generated from actual web logs moreso than an external service. It’s possible that google misses some visitors if the code doesn’t load properly. Some visitors that don’t have javascript turned on will not be counted at all. I prefer using google over awstats since it’s reports are more comprehensive, but when quoting actual traffic numbers to interested parties I will quote the stats generated from the web logs.
Left by Drag Racing on November 24th, 2006
If I had to choose between the two I would say Analytics because it has more features. With that being said, I still like using AwStats as not everyone has cookies enabled and well lets face it, it makes your site look more popular
Left by Wes Fernley on December 20th, 2006