The most frequently used data in web analytics is page views (PV). This data refers to the number of HTML documents (web pages) that a customer views. Since users normally view multiple pages within a site, the page view count is several times greater than the number of visits.
In general, a higher number of page views means more visitors viewed more pages, or that a site or its contents are very popular. However, let's delve a little further and look at user behavior. Users are not satisfied with every page they view. Sometimes a visitor will wander around a site looking for the page they want to see. So, in order to access the page they want to see, they go through some pages in which they are not interested at all.
In this way, a variety of hidden truths exist behind the data. Thus, you cannot see the whole picture merely by looking at the page view count.
In web analytics, this type of data duality and other factors make it important to look at the numbers with a hypothesis in hand, and to do so with an understanding of the nature and limits of access logs.
Science involves performing hypothesis testing multiple times – and web analytics is no different.
In order to ensure that your web analytics is meaningful, the following points are particularly important:
[1] Set numerical targets
[2] Have behavioral hypotheses based on a set of user behaviors
[3] Compare data with past data or data from other pages
[4] Design a site that facilitates web analytics (results verification)
[5] Investigate problems while comparing with the actual site
- [1] Set numerical targets
In order to understand the meaning of your data you must first have a hypothesis, in the form of reference values. Therefore, start by setting reference values, which will be your numerical targets. Set numerical targets according to such points as:
- The purpose of your site
- A behavioral hypothesis of users (described in[2])
- Previous PV counts and number of visitors to the site
- A measurement of each type of data using other media (number of customers, cost, etc.)
Once set, these targets should be revised from time to time. It is enough if you get a rough idea of the monthly number of visitors and PVs for a start.
It is important to remember here that the primary purpose of your site is not simply to increase the number of page views.
The numbers are nothing more than an indicator used to measure to what degree the site is achieving its purpose. It is only when your purpose is something like acquiring prospective customers or reducing support costs that you can talk about numerical targets such as what the numbers will look like when your goal has been reached.
This approach also allows you to verify the main users of the site and the volume of interest.
|
|
The duality of data