Segment variants of one email
Use when you have one good email and several audience segments that need genuinely different angles.
Fill in before running
Replace each placeholder with your own detail. The more specific you are, the less the model invents.
- {{BASE_EMAIL}}
- {{OFFER}}
- {{SEGMENT_DEFINITIONS}}
- {{SEGMENT}}
Getting a better result
- Define segments by behavior or situation, not by job title alone. Titles rarely predict what someone cares about.
- Let it tell you a segment is not worth a variant. Six near-identical versions cost more to maintain than they earn.
- Confirm the merge fields exist before sending. A blank token in the first line is worse than no personalisation.
Questions about this prompt
When is a variant per segment better than one email to everyone?
When segments differ in situation or behaviour rather than in label. If the only difference is job title, one email with a relevant subject line usually performs as well. This is for cases where the argument itself changes, such as a first-time buyer against someone renewing for the third year.
What do the segment definitions need to contain?
What defines each group and what it cares about, alongside the base email and the offer. Those definitions also license the merge fields, since the prompt will not use a token unless your definition proves you hold that data or track that behaviour.
What comes back for each segment?
Subject, preheader, body, call to action, what changed from the base and why, and what your data must contain to send it, then a ranking of segments by expected lift. The ranking is the useful part, particularly where it names a segment that should just receive the base email.
What is the mistake with segment variants?
Keeping every variant because you asked for them. Six near-identical versions cost more to maintain than they earn, and each is another thing to update when the offer changes. The other is shipping a merge field nobody confirmed, which prints a blank in the first line of the email.