 return to the list of columns |
No.26 The Key to Improving Your Website-Web Analytics (Part 1) (March 28, 2005)
Everyone who operates a website is interested in web analytics. Since the access log is the only trace left behind on a site by all users who visit it, web operators should definitely take advantage of this resource.
Also, in the Plan-Do-Check-Act (PDCA) cycle of operating a site, web analytics is effective in the Check phase as a method of verifying results. For this reason, web analytics is very important in terms of overall site operation.
However, just because an access log contains the data of all users who access a site does not mean it is a universal solution. In the end, since you are working with numbers, you may go astray in your interpretation of those numbers if you don't perform your analysis based on a firm grasp of the nature and limits of the log.
 What you can understand with web analytics
|
Of course opinions and requests sent in by users can be a good reference, but you mustn't forget that these opinions are only those of the most assertive users, and behind them is the silent majority who don't say anything. Access logs provide information on the behavior patterns of on all users on your site, including this silent majority, such as where they came from and what pages they viewed.
Accordingly, a proper analysis of an access log can provide you with a factual picture of users' behavioral history and patterns.
By comparing the numbers from this data with such items as
- initially set target values,
- behavioral hypotheses,
- previous web analytics results,
- and the results of web analytics on other pages,
you can infer the problem areas of the site and the needs of its users. However, this is still simply speculation. It is important to understand that at this point all you have is a hypothesis.
|
What you can't understand with web analytics
|
Web analytics can help you to understand users' behavioral history and behavioral patterns-but it won't tell you the reasons behind their behavior.
Therefore, it is risky to think that simply by having done web analytics you now understand the reasons behind user behavior. Since what you obtain from web analytics is never more than just a hypothesis, the next step you must take is to test that hypothesis using another method.
In pinpointing the reason for the behavior, the most effective method is to conduct a usability test.
Here at beBit, we conduct usability tests several thousand times. We often find that the reason we had assumed to be behind the behavior is different from the actual reason driving that user's behavior.
Since the numbers can be interpreted in any way we like, it is important when operating a website to take seriously users' behavior and to listen carefully to their opinions, rather than simply relying on numbers.
|
In the second half of this article, I will explain some specific points related to actual web analytics.
Yukiko Takei
|
|
