Taifa MailTaifa Mail Docs
Broadcasts

A/B Testing

Test subject lines and content with broadcast variants in Taifa Mail.

A/B testing lets you send different versions of a broadcast to different slices of your audience, so you can see which subject line or content performs best.

Variants

A broadcast with no variants is a single send: everyone gets the broadcast's own subject and body. To run an A/B test, add one or more variants. Each variant has its own subject and email content, and a weight.

FieldDescription
LabelA name for the variant, e.g. "A" or "Short subject". Labels must be unique within a broadcast.
SubjectThe subject line for this variant (required).
HTML body / text bodyThe email content for this variant (optional).
WeightA relative whole number, minimum 1. Higher means a larger share of recipients.

You can add, edit, and delete variants only while the broadcast is a draft or scheduled.

How recipients are split

Weights are relative, not percentages. When the broadcast sends, each recipient is assigned a variant by weighted random selection. A variant's share is its weight divided by the sum of all weights.

For example, with two variants weighted 3 and 1, roughly 75% of the list receives the first and 25% the second. Weights of 60 and 40 produce the same split as 3 and 2.

Weights do not need to add up to any particular total. Each recipient receives exactly one variant.

Reading the results

Each variant carries its own live counters: recipient, delivered, bounced, failed, opened, and clicked counts. Compare them to decide which variant performed better.

The broadcast's top links report ranks the most-clicked URLs across the whole broadcast (all variants combined), with each link's click count and its share relative to the top link.

Tips

  • Test one thing at a time. If you change both the subject and the body between variants, you will not know which change drove the difference.
  • Give the test enough audience. Very small lists produce noisy results - a few opens either way can swing the percentages.
  • Apply the winner. Once you know which variant wins, use that subject and content as your default for the next broadcast.

On this page