Observation is one of the methods for conducting primary marketing research. This method is used by firms if they have their own website. The research carried out within its framework consists of the collection and subsequent analysis of data obtained from log files (log files) of the web server or from cookies. This information may pertain to visitor behavior, the sequence in which they traverse across pages, or web server visit statistics. In the event of placement on a search engine’s website, user queries might be gathered and evaluated in addition.
The main distinguishing feature and advantage of this type of marketing research over traditional methods of observation and Internet surveys that require the active participation of respondents is the ability to collect valuable information without involving visitors in action.
First, let us turn to the consideration of methods for identifying users since they determine the accuracy of conducting or evaluating certain activities.
There are three main ways to identify users:
By IP address of the visitor’s computer – this method has the maximum error compared to the others. This error is determined primarily by the fact that the site can be visited by several users with the same IP address, for example, working through a proxy server;
Cookies are small data files that help a web server to track users on website, when visited through a browser. Thus, during the next visit, the server knows that this user has already been its visitor before. This knowledge is used, for example, in advertising, when a banner that he has already seen is not shown to him. In more sophisticated systems, with the help of cookies, it is possible to study the preferences of the visitor and show him appropriate advertising on each visit. The main error in applying this method is created due to the fact that cookies identify exactly the user’s browser and not a specific person.
In this case, when visiting a site or accessing one of the Internet services, the user enters his name and password, and the system can uniquely identify him during the entire visit. This method has the least errors in counting users and their repeated visits, but, unfortunately, it is mainly applicable only to certain Internet services, access to which requires their authentication.
At the moment, the most common is the identification of unique users by IP address, which is done either by site log files or by various counters. At the same time, the number of systems that allow users to be identified using cookies is growing.
Methods for obtaining and processing statistical data on the behavior of visitors
The main methods of collecting information about the behavior of visitors to the site are counters and the use of statistics obtained by the Internet service provider.
Another visitor recording tool is using statistics collected by the Internet Service Provider. The service provider maintains log files that contain information for each request for a Web page or graphic from the server. The log file usually contains the following information: the visitor’s IP address, the date and time of the visit, the command, the requested file, the link from which it got to the server, the browser, and the platform used. It does not store information about the visitor’s email address, only the IP address by which the user’s domain can be determined. A domain may be of significant interest, for example, if it characterizes the geographic region of the user. The source from where the visitor got to the server is very informative, as this allows you to analyze the effectiveness of various sources of access to the website.