Sometimes it is difficult to know which design is more optimal to increase conversions on a web page or in an email campaign, or we simply want to see if the new planned design will give us better results. The best way to know which of all our options is better is to perform an A / B test .
What is A / B testing?
A / B testing simply consists of testing at least two variants of some element of our marketing . It will allow us to compare the results obtained to make the best decision for the project. Although it does not have a fixed use, email marketing campaigns and web design are the points that most use these tests.
We see a curious case with this Chile Mobile Database methodology to optimize a web page with Amazon. Your web analytics department did one of these tests to select the color of the purchase button on your website. The winner was orange, with a conversion rate higher than the rest of the colors. This is a curiosity, since it has always been thought that green is the color that most encourages converting, but in this case it was not.
Why use an A / B test?
The use of A / B tests will us with provide a great number of advantages when it comes to achieving the final design of a web page or a newsletter. Below I would like to share with you some of these advantages.
Image selection : nowadays visual support is essential to convert, and it is one of the factors most highly valued by users. Through a test, if we carry out the analytics correctly, we can know which images are most interesting to us.
Title selection: this factor is very important in email marketing campaigns. An attractive title can make the difference between whether an email is opened or not. If we are not able to capture the user’s attention, it will be useless to have an incredible design within the email.
General design: although it seems silly, it is not the same to put the text on the left and the images on the right than vice versa. The analytics will help us choose the best option for our design.
Improve brand image: It seems silly, but it is. In the end, by carrying out these tests we are adapting to the tastes of the clients, that is, we think about improving their user experience on our site.
To innovate: in the end with each test you are trying to innovate. If an experiment goes wrong, few people have experienced it and your image will not suffer. You will not make any changes if most of your users are not satisfied with them.
To increase profitability: it is the main objective of performing A / B tests. The decisions you will make will be based on the KPIs you mark, which will logically be directly related to the profitability of your website and your business.
a / b testing exampleHow can I perform A / B testing?
Doing an A / B test is not as simple as it seems, but requires prior planning, to achieve all the advantages of performing an A / B test it is essential to be clear about this point. Here are some steps so that you are more clear about the route to follow.
Set goals
We have to be clear about what we want to achieve with these experiments. The most common is to seek an improvement in KPIs and increase conversions:
KPIs improvements
Average time on site.
Visited pages.
Bounce rate.
Shopping cart abandonment rate.
Average value of the purchase receipt.
Increase conversions
Contact forms.
Download brochures.
Subscriptions to the newsletter.
Decide what changes you want to make
Once we are clear about the objective we want to achieve, it is important to know which sections we want to change. Actually, anything can be modified to improve the ratios , size or color of the buttons, distribution of images and texts … the important thing is to make the changes one by one, since if you do them all at once, you will not optimize the maximum.
Crear la prueba
Existen varias herramientas para la realización de los test A/B, aunque yo hablaré de las dos que conozco, una para diseño web y otra para el envío de newsletters:
Google Analytics : in its experiments section we can do these tests. You will have to have two pages, and you will have to indicate to what proportion of the Brother Cell Phone List users you teach each one of them.
Mailchimp : in the free version, it already gives you the option to perform these tests. It will allow you to create two different email layouts. Once the campaign has finished, it will show you the data, so you can choose.
Expect
The A / B tests are not done by time, but by number of users . It is useless if the test is only carried out by 10 people. It is necessary to have a large number of users, the more the merrier.
Choose the solution
Once the comparison is over, it is time to make the best decision.
Re-measure
In the digital world, constant improvement is mandatory. Once you carry out a test, and its subsequent changes, you will have to continue to see if it is possible to further optimize your page or your new
