Site-centric analyse is a measure based on the website. Two technological approaches are being used today:
- logfile analysis
- page tagging
A log is a text file containing all the events on a web server with requests and responses. Logfiles are files generated by web servers each of their transactions. We can change the format to make them more compatible with different analysis tools. The web server (Apache, IIS) which delivers a website also notes in a file (called log) all requests (hits) that are made. Log analysis is therefore to analyze these files to get the statistics of visiting a website.
Business-to-consumer user-centric analysis is a measurement based on a panel of Internet users. The measurement range is based on building a team with a tracking software that will record all the internet usage. To ensure the representativeness of such a panel, it is necessary to conduct a scoping study which will indicate the structure that the panel should have. Finally, to be able to produce results at the individual level, all household recruited members are integrated into the panel with a clean identifier.
The interest of the user-centric analysis is to provide a consistent basis for comparison of the main sites in terms of penetration in a specific area and their market, and to know the profile of Internet users.
Business-to-business user-centric analysis is “customer centrics” optimized and is a measure based on the revelation of the names of organizations that visit a website. The measure is therefore based on an analysis of 100% of the traffic and allows you to make a segmented view of visitors. This allows to conduct audits of websites against the objectives of the Digital Strategies B2B. And so it helps to further the development of specific analysis for web marketing teams.
Another major advantage of the “customer centrics” analysis is the ability to measure the web relational capital of visitors to your website and turn it into a competitive advantage for sales teams.
“Hybrid” mixed methods
Many system administrators use a log analysis tool in conjunction with a marking system.
The analysis of packet sniffing allows to observe the activity of a network web server and to deduce its traffic. Like a log type solution, it loses the concepts of events and interactivity.
The financial argument
Logs are generated automatically by the web server, but it can, after a while, represent a considerable amount of data (several gigabytes). It is better to generate logs by week or month of the year. These logs must be stored and be as accessible and transferable as possible.
Methods based on tags store their information directly into a database that is then used to generate reports and / or export data to various information systems.
Translated from Wikipedia