🎉 Limited Time Offer: 70% OFF on all Monthly Plans for First Month

Unlimited Variants in All Plans

Test every version.
Keep the one that wins.

Run two or more subject lines and email versions in one campaign. Contacts are split for you, and every variant gets its own results.

Start First Month with 70% OFF
Five variants of one email with sent, opened and replied counts

What's A/Z testing?

A/Z testing means sending a few versions of the same email to random parts of your list, then keeping the one that gets the most replies. Instead of guessing what works, you find out.

  1. Step 1

    Write a few versions

    Change one part of the email in each.

    Subject lineOpening lineOfferCall to actionLengthTone
  2. Step 2

    Split your list at random

    Each contact gets one version, chosen at random.

    6,194 contacts, about 1,239 per version

  3. Step 3

    Keep the winner

    Compare replies and switch off the rest.

    • B4.9% replied
    • E3.3% replied
    • A3.1% replied

More replies

A stronger subject line or opener lifts replies across the whole campaign.

Data, not opinions

Replies decide which version stays, not a hunch or a debate.

Learn your audience

See which angles your prospects respond to and reuse them.

Keep improving

Keep the winner, test a new idea against it and repeat.

One campaign. Five versions.
A clear winner.

Contacts are split randomly between your variants, and each one sees only its variant for the whole sequence. The results tell you what to keep.

6,194 contacts · Q4 SaaS Founders
  • A

    Quick question, {{First Name}}

    3.1%replied

    1,240 sent41.3% opened38 replies
  • BWinner

    Saw your post on hiring SDRs

    4.9%replied

    1,238 sent48.8% opened61 replies
  • C

    Idea for {{Company Name}}

    2.3%replied

    1,241 sent39.4% opened29 replies
  • DOff

    Cut your bounce rate in half?

    1.8%replied

    1,236 sent36.8% opened22 replies
  • E

    {{First Name}}, 2 min read

    3.3%replied

    1,239 sent42.8% opened41 replies

Click rate is tracked per variant too, for emails with links.

Tests that give you
a clear answer.

  1. 01

    Change one thing at a time

    Test one subject line against another. When only one thing changes, you know what made the difference.

    A

    Quick question, {{First Name}}

    B

    Quick idea, {{First Name}}

  2. 02

    Start with 2 to 4 variants

    Fewer variants means more contacts in each, so a winner shows up sooner. Add more later.

    3 variants400 each
    10 variants120 each
  3. 03

    Switch off what isn't working

    The rest keep running. A disabled variant can't come back, so decide with enough data.

    B61 replied
    D22 replied
  4. 04

    Test follow-ups too

    Every step in a sequence can have its own variants, so follow-ups get tested as well.

    Day 1Email 1ABC
    Day 4Follow-up 1AB

Add a variant.
Let the results decide.

  1. 1

    Add variants to a step

    Open a step and click Add Variant. Change the subject line, the content, or both.

  2. 2

    Launch the campaign

    Contacts are split randomly. Each one gets the same variant for the whole sequence.

  3. 3

    Compare the results

    See sent, opened and replied counts for every variant, with rates side by side.

  4. 4

    Keep the winners

    Disable the variants that fall behind. The stronger ones carry the rest of the campaign.

An email sequence with a wait in days between each step

Why built-in A/Z testing beats splitting lists by hand

Testing by hand means cutting a list into pieces, building a campaign for each version and comparing reports yourself. In Deliveryman, every version lives in one campaign.

Splitting by handDeliveryman
SplittingDivide the list yourself and build a campaign per versionContacts split randomly and automatically
VersionsMore versions means more campaigns to manageTwo or more variants in one step, all plans
ConsistencyEasy to email one contact two versionsEach contact sees one variant for the whole sequence
ResultsCompare separate campaign reports by handOpen, reply and click rate per variant, side by side
Dropping a loserPause or rebuild a whole campaignSwitch one variant off, the rest keep sending

Related features

All features →

Stop guessing which email works.

A/Z testing comes with every Deliveryman plan, with as many variants as you need on every step of your sequence.

Start First Month with 70% OFF

Frequently Asked Questions

What is A/Z testing?

It's A/B testing without the two-version limit. You create several variations of a sequence step, like different subject lines or different content, and Deliveryman splits your contacts between them so you can see which one performs best.

How are contacts split between variations?

Randomly and automatically. Each contact is assigned one variation and gets only that variation for the whole sequence, so your results aren't mixed up.

What can I test?

Subject lines, email content and follow-ups. Every step in a sequence can have its own variations.

How many variations should I use?

We recommend 2 to 4 variations per step, and changing one element at a time. That way each variation gets enough contacts and you know what caused the difference.

Which numbers can I compare?

Analytics are shown per variation, including open rate, reply rate and click rate, so you can compare every version of a step at a glance.

Can I stop a variation that isn't working?

Yes. You can disable a variation at any time and the others keep running. A disabled variation can't be re-enabled, so give it enough sends before you decide.