How-to guide · Campaigns
A/B Testing
Write two versions of a campaign message in Text Blaster. Statxt splits your recipients 50/50 between them, then shows which version earned more replies.
How it works
- 1
Pick an audience
Choose the contact groups for the campaign, just like a normal blast.
- 2
Write two variants
Turn on A/B test and write Variant A and Variant B side by side.
- 3
Split 50/50
Statxt assigns recipients evenly across the two variants at send time.
- 4
Compare results
Reply rate, opt-outs, sentiment and reply speed for each variant.
STEP 01
Open Text Blaster and pick your audience
Start a campaign the usual way: give it a name and select the contact groups you want to reach.
- A/B testing works best with a few hundred recipients or more, so each variant gets a meaningful sample.
- DNC and opted-out contacts are excluded automatically.
- Use groups to choose the audience. Tags are for filtering later, not for targeting.
STEP 02
Turn on A/B test
In the Message section, click the AB button. A second message box appears so you can write Variant B next to Variant A.
- Click the same control again to exit A/B mode and go back to a single message.
- Your Variant A draft stays as it is when you switch A/B on.
- Variant B is required before you can send in A/B mode.
STEP 03
Write Variant A and Variant B
Change one thing between the two versions, like the opener, the offer, or the call to action, so you know what made the difference.
- Both variants can use merge fields such as {first_name}, {sender_first_name} and {brand_name}.
- Each variant shows its own character and segment count. Keep them close in length so cost doesn’t skew the test.
- Put opt-out language in both variants. Each one is checked against compliance rules on its own.
STEP 04
Preview, then send or schedule
Open the preview to see personalized messages for real contacts in your groups, shown separately for each variant, then launch.
- Recipients are assigned evenly across the two variants.
- Check that merge fields filled in correctly for both versions.
- Send now or schedule for later, the same as any Text Blaster campaign.
STEP 05
Compare the results
Open the campaign from the All Campaigns tab. The A/B Performance panel puts both variants side by side.
- Reply rate is the headline number, with replies, delivered and opt-outs underneath.
- Median reply latency shows how quickly each version got answers.
- Sentiment splits replies into positive, neutral and negative.
- Statxt calls a leader once each variant has at least 30 delivered messages. Before that, treat results as directional.
Pro tips
Get more out of A/B Testing
Test one thing at a time
If both the opener and the offer change, you won’t know which one moved the reply rate.
Wait for enough data
Small samples swing wildly. Let each variant reach at least 30 delivered messages before picking a winner.
Watch opt-outs too
A variant that gets more replies but also more opt-outs may not be the better message.
Match the length
Keep both variants in the same number of segments so one isn’t cheaper or more expensive to send.
Reuse the winner
Save the winning copy as a template so the next campaign starts from your best message.
Keep a test log
Note what you tested and what won. Patterns across campaigns are worth more than any single result.
Common questions
How does Statxt split the audience?
Recipients are divided evenly between Variant A and Variant B, so each version reaches about half of the eligible contacts.
Can I test more than two messages?
A/B mode in Text Blaster compares two variants. To compare more ideas, run follow-up tests against the previous winner.
Do both variants cost the same?
Each message is billed by its own segment count. Keep the variants close in length to keep costs and results comparable.
Last reviewed October 8, 2026. Names, phone numbers, contacts, and stats in the illustrations are fictional examples.
More guides
Try it in Statxt
Set up your workspace, bring your contacts, and send your first campaign today.