What is meant by testing an A / B website?

Question

Please explain (with examples) what is meant by A / B testing of a website.

Answer 1

In the context of the website, this means that some users of the / webapp website get slightly modified version (s), and you check how each modification affects the behavior of the user / user compared to the basic level of "control".

As always, when testing A / B, which is a statistical method, you need a large enough sample to correctly assess the impact of changes.

This was done by both Amazon and Google among well-known web companies .

VERY good article (with a marketing angle) on A / B web testing: http://www.clickz.com/3349901

In the interest of completeness, here is the full definition of A / B testing from Wikipedia:

A / B testing or bucket testing is a marketing testing method by which the original control sample is compared with a lot of tests with one variable samples to improve the response rate. A classic direct mail tactic, this method was recently adopted in an interactive space for testing tactics such as banner ads, emails, and landing pages.

Significant improvements can be seen through testing elements such as copying text, layouts, images and colors. However, not all elements produce the same improvements, and the results of various tests, it is possible to identify these elements which are consistently the best improvements.

Employers of this A / B testing method will distribute several test samples, including control, to see which single variable is effective in increasing response speed or other desired results. the test, in order to be effective, must reach an audience of sufficient size that there is a reasonable probability of revealing a significant difference between control and other tactics

Answer 2

Disclaimer: I have an A / B test run called Visual Website Optimizer http://visualwebsiteoptimizer.com/

A / B testing is a method by which you submit to visitors two or more different versions of a website. For example, suppose you have a website, and then your boss / partner suggests that moving the registration form from left to right can increase the total number of registrations because it reduces friction. Now either you can have faith in God, or immediately realize the proposal, or you may be curious and make a version of the page where the form is on the left. Then you redirect 50% of the traffic to the right form and 50% to the left form. Then you see which version actually gets more signatures.

The problem becomes a little complicated because you need to statistically prove that one version is better than the other. To do this, you can use online calculators, such as those available on my website http://visualwebsiteoptimizer.com/ab-split-significance-calculator/

If you are looking for examples of real situations, again there are many of them on my website http://visualwebsiteoptimizer.com/case-studies.php

If you have further questions, feel free to ask. I LOVE this topic :)