The different TYPES of stats packages
Log Readers
Examples of a log file
Whenever a hit occurs on a website, the access made by the browser is stored in a raw text file called a log file. This log file contains a list of all access to all elements on a webpage (including the page itself, every graphic, style-sheet, etc), storing alongside it various information about the access including the date/time of access, the IP Address of the computer attempting the access and potentially other information including browser type (eg. Internet Explorer or FireFox), computer type (eg. Macintosh or PC), and (very importantly) the referer etc.
A log reader is a program that reads a series of log files and make sense of them! As you can see from the screenshot, the log file is like a spreadsheet of information which is next to useless without some level of organisation. The log reader will use the IP Address as a unique indicator of each user, and can therefore determine where any one visitor's browsing has started and ended. It can also use the other information from the logs, including the details such as Browser type, computer type, calculate average visit lengths, etc.
Once a log reader has read in all the data from a log file, or a series of log files, it will store this information in its own database where you will be able to query it in order to find answers to specific questions, such as the number of visitors your website had yesterday, which pages were looked at most, and where those visitors came from.
Potential Flaw Alert!
Since the log reader will use the IP Address alone as a way of calculating a unique visitor session, there is a flaw in the log reader's paradigm. Say another website was hot-linking an image, this image file access would appear in your log with an IP Address, and would be interpreted by the log reader as another visitor to your website. However, this access would NOT actually be an actual visitor, would not have seen an actual web page on your site, and would thus inflate your visitor numbers erroneously.
Some servers have different configurations for log files and some of the information available in the files may be restricted. If you are on a shared server (that is, your website is hosted by a company who hosts your website on the same server as many of its other customers) then you may not always have access to your log files. This varies from host to host but ideally you would want access to these raw log files since you then have a choice to use your own statistics program (Log Reader) to read these logs. Invariably if you do not have access to the logs themselves, you will be presented with some form of log analysis program and it is often a version of Webalizer. Some ISPs also restrict the numbers of logs which are stored for your website - which means that if you are downloading them for offline analysis, you MUST keep on top of this.
What is Integrated Tracking? »
Log Readers vs. Javascript »
|
|
|