What is sampling in Web Analytics and what are its consequences?

Congratulations!! The doctor tells the father. What is he going to call you? Well Bill. Oh, any Bills in the family? No no. Bill is a guarantee of success. What does he mean? Man, notice that many successful people are called Bill: Bill Clinton, Bill Cosby, Bill Gates, Bill the Kid… Little Bill doesn’t know what awaits him. In another part of the world, a psychologist listened to him narrate the reason why he asked his partner to go to therapy. “Our marriage is super boring, we barely do anything. Everything is day to day, obligations. I go on Facebook and see happy people, kissing, traveling, at weddings, partying, dining in nice restaurants or going to the movies. Every day you see happy people on Facebook, all people are happy and we, if anything, have a plan 4 or 5 times a month”…

What Google Analytics tells you when you exceed 10 Million page views per month

Imagine that you are there, celebrating that every month your metrics are growing, your top industry data traffic volume is constantly growing and, consequently, so is your business. One day, you enter Google Analytics (as you usually do) and the bittersweet message appears: “Your data volume (xM hits) exceeds the limit of 10 M hits per month specified in our Terms of Use of the Service. If you continue to exceed the limit, you could lose data in the future.” As? What what? Then you start searching for articles on the subject and you reach the help of Google Analytics, where it explains the case . It turns out that the free version of Google Analytics has a limit of 10M hits per month and you have exceeded it.

What does it mean that sampling will be applied?

To explain it simply, think about polls in elections. When it is said that a political party will obtain x seats, it is not actually the result of asking all citizens but rather taking a more or less representative sample and extrapolating it to the total . We ask 1,000 and what comes out, we multiply it and we intuit what will happen. And, well, they don’t always get it right nor do they come close. TRUE? Data sampling in web analytics is exactly the same. It means Phone Number Lt that Google Analytics will not take all of the data that we send, but rather a “more or less representative” sample of it. At the same time, we also have no guarantees that the sample is equitable. How much information could we be missing? How reliable are the conclusions we can reach with that data? We are not clear.

Leave a Reply

Your email address will not be published. Required fields are marked *