Glossary
What is A/B test?
A/B test is an experiment where you show two versions to similar groups and keep the one that performs better.
What an A/B test is
You change one thing, such as a subject line or a button colour. Then you split your audience into two random groups. Group A sees version A. Group B sees version B. You compare the results against one goal, such as clicks or bookings.
How to run a fair one
Change only one variable at a time. Make the groups random and close in size. Decide the goal and the minimum sample size before you send. Do not stop early because one version looks ahead after an hour.
Common mistakes
Testing with too few people gives random results. Changing two things at once hides which change mattered. Calling a winner on a tiny difference wastes the next campaign.
Example
An Ahmedabad coaching institute sends two subject lines for an admissions reminder. “Last 5 seats for the April batch” and “Admissions close Friday” each go to 800 contacts. The first one gets clearly more clicks. The institute uses that style for the next reminder and notes the result for the team.