A/B (split) testing is a tool to help marketers to enhance their campaigns. With its help you can test the subject, the content and time of campaigns and check which ones work the best for your subscribers.
With a/b testing 2 versions of an email, which differ only in 1 parameter, are sent, then the results are compared.
There are two variants of a/b tests depending on your test goals:
1. Test on a small quantity of subscribers
The goal is to determine the best email template.
A small sample of subscribers is divided into two parts, each receive their own template. The template with more clicks is sent to the rest of the subscribers.
A size of a sample is chosen in % of the total database size. For example 10% of the database, 5% for each template. The best template may be chosen manually or automatically depending on the feedback.
Using a manual choice you can analyze extra factors and make a decision.
2. 50/50 test
The goal is to test hypotheses and identifying patterns.
A group of subscribers is divided into two equal parts. The first part sees the A template, the second part sees the B template. The feedback is used in further campaigns.
A and B templates differ in one parameter: either headers (sender and subject) or in content or in time of the campaigns.