This is true not only for websites but for all types of data analysis. It goes without saying that unless the research target is large enough, the reliability of the data analysis results will be very low.
For example, it would not be professional to conclude that “this is what Americans think” or “this is America” based on the results of interviews with only 10 people in the United States, which has over 300 million people. Unless you survey, say, at least 10,000 people, you won’t know the overall trends.
Also, just because there are 3 women in a company with only 5 employees, it is not wrong to say that “this company has a high proportion of women,” but it is a meaningless analysis. If this is a company with 50,000 employees and 30,000 of them are women, it can be confidently said that the company has a high proportion of women.
The same goes for website access analysis. To be extreme, when the number of accesses per month is only about 10 page views, saying “30% of the accesses are from Osaka” and “60% of the accesses are from women” is not very meaningful data analysis. The value is different from the percentage of accesses of a large website with 100,000 monthly page views.
It is true that analyzing data and understanding trends of a small website can bring you some ideas, but it is also important to understand that analyzing trends on small websites with few accesses has limited meaning.
