Change one thing in an A/B text test
Text Blaster can split a campaign 50/50 between two messages. The reply rate is only readable after each variant has enough deliveries, and opt-outs count too.
A/B testing in Text Blaster is two versions of one campaign, not a separate product. You pick groups the usual way, turn on A/B, write both messages, and Statxt assigns eligible recipients evenly at send time. DNC and opted-out contacts are already gone. The test is more useful with a few hundred recipients or more, so each half is large enough to read.
Turn it on in the message section
The AB control opens a second box for Variant B. Click it again to go back to a single message. Variant A stays as you left it. Variant B has to be filled in before an A/B send will go out. With the test on, Use template becomes an A button and a B button, so each side can start from a saved template.
Change one thing: the opener, the offer, or the ask. If both the opener and the offer change, a higher reply rate does not tell you which line did the work. Both variants can use merge fields. Each has its own character and segment count. Keep them in the same number of segments so one version is not simply cheaper, or more expensive, to deliver. Opt-out language belongs in both. Each variant is checked on its own.
Wait for 30 delivered messages
Preview shows personalized copies for real contacts, separately for each variant, before you send or schedule. After the send, open the campaign from All Campaigns. The A/B Performance panel puts the variants side by side. Reply rate is the headline, with replies, delivered, and opt-outs under it. Median reply latency is how fast answers came back. Sentiment splits replies into positive, neutral, and negative.
Statxt will call a leader once each variant has at least 30 delivered messages. Before that, treat the numbers as a hint. A variant that wins replies and also wins opt-outs is not automatically the better message. When you do have a winner, save that copy as a template so the next campaign starts there. Text Blaster only compares two variants. A third idea is a follow-up test against the winner, not a third box.
The click-by-click version lives in the A/B Testing guide.
Walk through it in the product
The guides cover these same workflows step by step, with the screens you will actually use.