To send high-performance campaigns and boost clicks on your e-mails, optimise your conversion rate with A/B tests.
The A/B tests compare two variants of the same page until the successful version is defined, revealing users’ reactions to the changes. The control and test variants are sent to two different groups of users registered on the same database. The results determine the most effective version to be sent to the remaining users. Each user will receive just one e-mail.
How to run an effective A/B test, in 5 steps:
1. Establish your goal. Define the parameter to be assessed and the metric to be used to measure the results: for example, variations in the opening rate are used to test the object of the e-mail, while the click and conversion rates point out the impact of changes to the layout.
2. Choose the elements to be tested: change only one variable at a time. You can test:
– Layout: the position of the elements on the page, or the combination of colours.
– Visual media: including videos is more engaging, but they are also heavier to download.
– Copy: the headline and the style of the message, the length of the content and how it is positioned.
– Call-to-action: text, dimensions, colour and position on the page.
– Form: where to position it and which fields are required. Run a test on a mobile device to check that the form can also be filled in easily on smaller screens. Include autocomplete textboxes.
– Subject: use the recipient’s name to immediately establish a contact (“Anna, here’s the latest for you”), or offer a teaser indicating why the user should open your e-mail, with a number of keywords (“Slow smartphone? Try the new tool for free”).
3. Establish a statistically significant sample to send the test to: a sufficiently large number of random people must be involved for the results to be considered valid for the whole reference market. Address at least 20% of the contacts in your database.
4. Send versions A and B at the same time: sending them on different days or times may distort the test results, unless the day and time of sending are the variables you are seeking to test.
5. Read the results obtained, listen to the reactions of your users, identify the most effective version and send!
To maximise your conversion rates, test your messages with us: we’ll help you identify your successful strategy.
This post is also available in: Italian