Skip to content

A/B test your opener or connection note

Send half your leads one version and half another, and HeyLeads tells you which one wins and how sure it is.

2 min read · Updated November 18, 2026

An A/B test step splits leads between versions and compares them. Use it to find out whether a note beats no note, or which first message gets more replies.

  1. Add an A/B test step

    Place it right before the message (or connection request) you want to test.

  2. Choose the goal

    Under Goal, pick what a version is judged on: reply rate or acceptance rate. Acceptance is decided by the connection request, so use it to compare requests and use reply rate to compare messages.

  3. Set the traffic split

    Give each version a share. An even split is the fairest. Name the versions so the results are easy to read.

  4. Write the versions

    Draw a line for each version to its own message or request. Change one thing at a time, so you know what made the difference.

#Reading the result

Open the campaign's Results tab and find A/B tests. You will see how each version did, and one of three outcomes: a clear winner, not enough data to call it, or no clear difference yet. With three or more versions the bar is raised a little, because more comparisons make a lucky streak more likely.

  • Do not check every day. Early numbers swing wildly.
  • When there is a winner, you can use it for everyone from the results page.
  • To stop sending new leads to a weaker version, switch it off in the step. Leads already in it carry on.
Was this article helpful?