About the service
When a newsletter "doesn't land," it's rarely about the offer itself — it's usually the subject line no one opened or the text no one clicked from. I set up complete A/B testing for email newsletters: I prepare two versions of the subject, preheader, or email content, send them to a small test portion of the database, and based on open and click results, automatically send the winning version to the remaining, larger portion of subscribers. No manual second send and guessing is left.
I work with the mailing service the client already uses and set up the split so the test sample size and result wait time are statistically justified for this specific database, not taken fifty-fifty for the sake of it. If content is more important than subject, I test entire email versions — different openings, calls to action, offers — and look not just at opens but at what version actually leads recipients further down the funnel.
After the mailing is complete, I prepare a clear report: open and click rates for each version, statistical significance of the result, and an explanation of why this version won. Typical work composition:
— setting up two versions of the email subject or content in the mailing service used;
— calculating test sample size and automatically sending the winner to the rest of the database;
— setting up open and click tracking, including link marking;
— final report with figures and recommendations for the next mailing.
I work with the mailing service the client already uses and set up the split so the test sample size and result wait time are statistically justified for this specific database, not taken fifty-fifty for the sake of it. If content is more important than subject, I test entire email versions — different openings, calls to action, offers — and look not just at opens but at what version actually leads recipients further down the funnel.
After the mailing is complete, I prepare a clear report: open and click rates for each version, statistical significance of the result, and an explanation of why this version won. Typical work composition:
— setting up two versions of the email subject or content in the mailing service used;
— calculating test sample size and automatically sending the winner to the rest of the database;
— setting up open and click tracking, including link marking;
— final report with figures and recommendations for the next mailing.
