> ## Documentation Index
> Fetch the complete documentation index at: https://clapvo.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Randomize message variants (A/B test)

> Send different message variants to different recipients in the same campaign.

Randomize message is Clapvo's built-in A/B (or multi-variant) testing for broadcasts. Prepare multiple message variants for the same campaign and Clapvo distributes them randomly across recipients.

## Why this matters

* Helps you discover which copy, offer, or call-to-action actually performs best.
* Avoids guessing wording - real audience response decides.
* Reduces the risk that one tired or low-performing message tanks an entire campaign.
* Protects sender reputation - repeating the exact same text to many recipients is a common spam signal, while randomized variants look more human.

## Before you start

* Two or more message ideas you want to test
* Clear hypothesis: which element are you varying (headline, CTA, length, tone)?

## Steps

<Steps>
  <Step title="Open the campaign builder">
    Create or edit a campaign.
  </Step>

  <Step title="Add message variants">
    Enter each variant of the message - same audience, different copy.
  </Step>

  <Step title="Enable randomize message">
    Turn on the variant rotation option.
  </Step>

  <Step title="Preview">
    Use the [preview drawer](/docs/product/campaigns/preview-drawer) to confirm each variant reads well.
  </Step>

  <Step title="Save and send">
    Clapvo distributes variants randomly across recipients.
  </Step>
</Steps>

## Typical use cases

* Testing different headlines or opening lines on a promotional broadcast
* Comparing CTAs (for example, **Reply YES** vs **Click here** vs **Book a slot**)
* Comparing short vs long copy on the same audience
* Comparing message tone (formal vs casual) for the same offer
* Ongoing optimization of recurring templates used in sales follow-up or support nudges

## Tips

* Vary one element at a time so you can attribute the difference to the change.
* Pair with [Randomize sender](/docs/product/campaigns/randomize-sender) for the strongest combination - multiple senders and multiple variants per campaign.
* Keep the winning variant; iterate on the rest next round.

## Related

* [Randomize sender](/docs/product/campaigns/randomize-sender)
* [Preview drawer](/docs/product/campaigns/preview-drawer)
* [Create a campaign](/docs/product/campaigns/create-campaign)