Server side vs. Client Statistics: Correct Me if I am Wrong!
I am calling for a little help from our viewers…If you know anything about this topic, please reply or direct me to a great source.
So, I was delivering a report to one of my clients. We adapted our reports to reflect the traffic in a month from Google Analytics. I was reporting that his visitor count was around 5,500 people in the last month. He started contridicting me; giving me different numbers that were very different than what Google was reporting. I got a little nervous. How come these numbers are so obviously different and who is right?
I told my client I would check into this and he gave me access to his information. I got off the phone and turned to my co-worker: “Help! How do I explain this? And why is there such a big difference?” I was under the impression that Google didn’t need a toolbar or something to track visitors. It was simply the traffic of the entire world wide web.
Hence, I am told about Server Side vs Client Statistics. My client and I could be reporting two different traffic reports. I believe I am reporting on the client side, or real time visits; while, my customer is reporting on the server side.
The table below, which has a link to the original article, stats why different results may be produced. What would be more accurate? Or what would one recommend to a client?
Why Log File Analysers and Real Time Statistics May Produce Different Results
| Issue | Log Analyzer | Real-Time Statistics |
| Counting Method | Counts Hits from the Web Server Log Log unique by IP address only | Counts Visitor (Unique by Cookie and IP address) and page views |
| Reporting Method | Periodic processing and extraction from log file | Available Real-Time from any browser |
| Crawlers/Spiders | Tracked via hits | May not be counted if HTML is not viewed |
| Proxy Server Page Caching | Cached pages not counted | All Pages counted |
| Non HTML files | Non HTML files such as Graphics and downloads may be counted as page views. | Unless specifically tracked will not be double counted |
| Error Pages | Tracked in separate tables of Log File | Only counted if Java Script is able to execute |
| Redirects | Requires special configuration to track | Automatically tracks the first page viewed |
Anybody have any thoughts on this? Thank you for your help!
September 8th, 2007 at 10:53 am
Jules, thats a good table of differences. I use both and review monthly, but I don’t compare them. The log analysis has its own value and indicators, like file downloads, error messages, and overall health. The client-side data shows true activity of site visitors. The use of javascript eliminates crawlers and other ‘white noise’, so the actual activity comes into focus. An a smaller site, this white noise can make a significant portion of log files, making true trends and activities hard to discern.
September 8th, 2007 at 11:44 pm
The whole javascript thing is probably part of the problem. There’s been a surge of people disabling, javascript, cookies etc, for privacy sake and fear of hack attempts. The latter is probably more justified especially if using IE.
September 10th, 2007 at 8:57 am
Thank you Dave and my Anonymous Source. Both of your comments were super helpful!
I looked into the stats that my client was getting, they were definitely server-side statistics. So it will be useful to look into both resources.
It was interesting to see the varied results.
Thanks again for your comments. They are always welcomed!
Jules
February 12th, 2008 at 11:28 pm
If you are looking forward to unsecured loans, Unsecured personal loans,unsecured bussiness loans,Unsecured personal loans UK,Unsecured Loans UK and Unsecured tenant loans than logon to http://unsecuredpersonalloansonline.blogspot.com