There’s no service that provides 100% accuracy unless you have found the perfect world which I doubt. In the analytics world you will always find differences between services. An example of this is Google Analytics versus StatCounter. Both companies provide web analytics, but the data is different for both when looking at Unique Visitors, Pageviews, etc. They are very close which is surprising because some services are so far apart you don’t know which one to believe.
Google Analytics Data (February 23, 2011)
- 2,433 Visits
- 2,894 Pageviews
StatCounter Data (February 23, 2011)
- 2,458 Visits
- 3,017 Page Loads
Google and StatCounter are very close when reporting data, but I found a huge disparity on two services that check Markup Validation.
W3C (http://validator.w3.org/) and Nibbler (http://nibbler.silktide.com/) both provide validation which checks the markup of HTML and XHTML of Web Documents for errors.
The difference in errors reported is astonishing. I know and trust W3C (The World Wide Web Consortium) who develop Web standards, led by Web inventor Tim Berners-Lee and CEO Jeffrey Jaffe. How can you not trust them, the man who invented the web and not Al Gore is part of this organization.
I ran the markup validation on both sites and here’s what was reported;
W3C
Result: 4 Errors
Address: https://www.techjaws.com
Encoding: utf-8
Doctype: XHTML 1.0 Transitional
Root Element: html
Root Namespace: http://www.w3.org/1999/xhtml
Nibbler
W3C compliance
- A total of 83 errors and 0 warnings were found on the 5 pages tested.
- No pages are W3C compliant. Because there are errors in the code, some web browsers may not be able to read this website correctly and it may not always display correctly.
- This website was validated as XHTML 1.0 Transitional
W3C reported 4 errors versus 83 errors reported by nibbler. I know that Nibbler is still in alpha and they may stay that way until their validation tool is a bit more accurate. Nibbler also reports on many other analytics and I found more inconsistencies here as well.
Nibbler misses the mark on many of its reporting metrics including;
Social media (techjaws.com)
We found 0 Tweets about this website on Twitter.
The number of tweets in the last 4 days is 30!
There are 0 articles on Digg relating to this website.
The actual number of Diggs is over 1200!
Incoming links
About 12,261 other pages were found linking to this website.
The actual number of incoming links is 16.971.
The overall score 3 days ago was 6.7 and today 5.0. How can an overall score change that much? I guess we will have to wait for beta to see if nibbler improves in accuracy.
I have often wondered about the disparity between web analytics services which in some cases can be significantly different. You demonstrate the disparity quite clearly although what causes it remains pretty mysterious to me.
I am pretty sure we should be aware for google analytics, there are lots of website offering analytics to track our websites improvements or visits – the traffic information. Great post. I value this article, it make sense.
I have long noticed the disparity in stat counters on my technology blog particularly in relation to Google Analytics even when I compare with Adsense impressions. Overall, I believe GA is quite conservative when compared to the likes of AWStats and there are several reasons for this. At the end of the day, I would advise anyone to focus on just one statistical tool to avoid confusion.
Jay,
i use to analytic services and as I mentioned in this article, both are close but not exact and I can live with that.