Awstats vs Google Analytics vs Alexa
Have you ever thought how accurate your statistics are? Most of you should have noticed the big difference between Awstats, Google Analytics and Alexa. In fact I took a few minutes to actually go through my stats and visually compare the data. Bellow is a quick (Not so quality) graph comparing Analytics to Awstats and Alexa. Click on the image to enlarge.
The graph is not to scale. So you should only compare the patterns. Another graph bellow however IS to scale and compares Awstats to Google Analytics for this month.
As you can see there is a big difference. But why?
Alexa, Google Analytics and Awstats use different technologies to collect the data. Awstats is server based and crunches the numbers it finds in server logs. With Awstats every call made to a server is recorded. There is no way of going around it. All actions will eventually end up in your server logs. And the best thing about Awstats is that you are dealing with your own server directly. There is no other 3rd party to corrupt your data. However there are a few things on the client side Awstats can miss such as screen resolution. It does gather some data using JavaScript but this requires a compatible client and is said to be somewhat unreliable.
Google Analytics on the other hand uses JavaScript. If you forgotten to add the code to your newly created page it will never be logged. Another problem with JavaScript is that some browsers might not support it or accept it for one reason or another. You are pretty much relying on the client to provide you with the precious data. If however they happen to use cache they might not even need to request another copy of your page containing that lovely Analytics JavaScript. In which case you will never know you had a hit.
And then there is good old Alexa which can be described as a “joke”. It relies on data collected by the toolbar some people decide to install on their browsers to “make a better internet”. Even though it can give you a rough idea of the traffic a particular website receives, Alexa lacks accuracy. In fact they pretty much admitted how much they sucked recently by changing their algorithm.
Overall none of the big 3 will have 100% accuracy. In fact is there anything in this world we can measure with 100% accuracy?
May 23rd, 2008 at 4:17 pm
Maybe you can put one of those old skool internet counters that looks like a car odometer… lol…
May 23rd, 2008 at 11:32 pm
Some people still use them
May 24th, 2008 at 7:12 pm
I didnt know Awstats so far, always used google analytics. But i really hate giving all this info to the big “octopus”. So Awstats looks good to me, I,ll check out the possibilities.
May 25th, 2008 at 11:43 am
I know none of these are accurate. Probably nothing is. But if you have heard of woopra, that seems to go to the test. It is still in beta mode, but it will come out soon. The downfall is you have to purchase it, or part of it. But it points to you that if it can track visitors live (it can) then that is pretty good. You do have to put the code on your site, and since so many people have joined it may be hard to get in and approved.
May 25th, 2008 at 5:52 pm
I was always curious about this. Now I know why good post.
May 26th, 2008 at 3:19 pm
I’ve been using Google Analytics all the time, and I’m quite happy with the stats that it brings (the keywords, visitors, geographic locations, browser types, contents, referring urls, etc)
May 26th, 2008 at 3:22 pm
Nice general approach to blogging SEO, well presented, many interesting features and link exchange ideas of interest to all bloggers. Helpful.
Cheers,
JA.
May 28th, 2008 at 1:19 am
I have always been partial to Awstats myself. Google analytics seems to be missing some data that Awstats always caught. I also like that Awstats points you directly to the URL where you traffic came from, Google analytics only points to domain.
May 29th, 2008 at 2:05 pm
How about statcounter? In my observation, analytics and statcounter provide almost similar stat (with discrepancy +-50)
May 31st, 2008 at 5:19 pm
@ Nathaniel
I have had Woopra running on this blog for the last month. I should really blog about it. I’m loving it!
@ zaki blogjer
To be honest I would not know. Never really used it myself.
June 5th, 2008 at 9:22 am
[...] too curious on how accurate is my Google Analytics and Statcounter data after Seolad made a comparison between Google Analytics, Awstat and Alexa statistics before. As pointed out by him, the different [...]
June 8th, 2008 at 11:51 am
I use statcounter (pr:9) And it is the best I think.
June 19th, 2008 at 5:32 pm
I love awstats and it is much accurate than google analytics…
Regarding alexa, they have changed their toolbar based formula long ago…and their current ranking is pretty good…you will be surprised to find out how little traffic popular SEO marketers like John chow and maki (doshdosh) generate when compared to sited like labnol.org…
You will eventually find out how good these internet marketers are, as they have made you believe that they get big traffic,,,
June 19th, 2008 at 5:33 pm
I meant internet marketers and not SEO marketers…a typo…
July 27th, 2008 at 11:31 am
The problem with Awstats is that by recording almost everything it also distorts the data. For instance, if you have images pointing back to your server that will be recorded. Anyone said hotlinkers here? There are many other factors that you can observe by looking at your data and comparing it to GA.
For user-agent and ip detection Awstats really sucks!
At the end is better to have different sets of analytic software than rely on just one. Then, you have to be worry about duplicate tracking or software compatibility issues. Anyone tried mint?