EMAIL A/B TESTING, DONE PROPERLY

Let your customers pick the real winner, not a stopwatch.

Every campaign can go out as several designed versions: different subject lines, openings, offers or layouts. They rotate across your whole list, not a slice of it, so the winner is the one your actual customers chose.

Create your free account See how it works

Variants rotate across your whole list, not a sample
Human opens reported separately from automated ones
You choose the winner, nothing crowned on early numbers

Three designs in. One true winner out.

Hang on, did a bot pick that winner?

Most A/B tests crown the fastest click, not the best email.

Here's how it usually goes. You write two subject lines, the platform sends each to a slice of your list, say fifty people, waits an hour, and declares a winner. Problem is, the first opens in that window are rarely real customers. They're security scanners and preview bots, checking the link before your actual reader ever sees the inbox.

By the time your tradie regulars or your café locals open their phone over a cuppa, the "winning" subject line has already been locked in, sent to everyone else, based on numbers that were never about people at all. You never actually learn what your customers preferred. You learn what a scanner clicked first.

And it's not just subject lines. A different opening line, a different hero offer, a different layout, all of it matters, and all of it gets flattened into one guess because most platforms only ever test one thing, on a slice of your list, and only for an hour.

HOW IT WORKS

From one brief to a fair result

Design two or three versions

Change what you think matters: the subject line, the opening, the offer, or the whole layout. Each variant is a complete, designed email, not a one-word swap.

Launch to your whole list

Your variants rotate evenly across everyone on the send, not a sample held back for later. Every recipient is part of the test.

Watch results per variant

Opens, clicks and complaints are reported against each version separately, with human opens kept apart from the automated ones.

Pick your winner

Nothing gets crowned for you on an early, noisy number. You look at the results and decide which version to keep running or reuse.

MORE THAN A SUBJECT LINE

Variants that differ where it actually matters

A proper test needs more than two subject lines racing each other. A campaign can launch with several designed variants that differ in whatever you think will move the needle: the subject line, the opening paragraph, the offer itself, or the whole layout.

That means you can test "10% off today" against "free delivery this week" as complete, finished emails, not a guess dressed up as a hypothesis. Every variant looks like a real campaign because it is one.

Different subject lines, side by side
Different openings and offers
Different layouts, fully designed

Campaign · 3 variants drafted

Variant AVariant BVariant C

Subject ATuesday tune-up: 10% off today

Subject BYour Tuesday coffee, sorted

Layout CPhoto-led hero, same offer

Illustration · sample data

Audience split · even rotation

Variant A33%

Variant B33%

Variant C34%

Total recipientsFull list, every subscriber

Rotation Randomised, no sampling

Illustration · sample data

FULL AUDIENCE, NOT A SAMPLE

Every variant gets a fair share of your whole list

Variants rotate evenly across the whole audience, not a sample held aside to decide for the rest. If you launch a campaign to four hundred people, each version reaches roughly the same number of real inboxes, at the same time, so the test is fair at full size, not a guess made on fifty people and applied to the other three hundred and fifty.

It pays off beyond the test itself, too. Sending a handful of genuinely different versions, rather than one identical email to everyone, means your campaign doesn't present as a single, uniform blast to the inbox providers reading it on the other end.

Rotated across everyone, not a slice
A fair test, at full send size
Fewer identical-looking sends to inbox providers

RESULTS YOU CAN TRUST

See what real people did with each version

Results are reported per variant, with human opens kept in a separate column from automated ones, so a flood of security scanners checking a link can't dress itself up as enthusiasm. You're comparing people, not software.

And the decision stays with you. Nothing gets crowned a winner automatically on an early, noisy number. You read the results per variant, when you're ready, and pick which one to keep sending or reuse next time.

Per-variant opens and clicks
Human opens separated from automated
You choose the winner, on your schedule

Results by variant

Variant AHuman opens 41 · clicks 9

Variant BHuman opens 58 · clicks 17

Variant CHuman opens 37 · clicks 6

Automated opens excluded

Choose winner

Illustration · sample data

Variant B's winning, and that's no scanner.

WHAT IT MEANS FOR YOU

Knowing what actually worked, every send

Know what actually moves your customers

Stop guessing between two subject lines. Test the offer, the opening and the layout, and see which combination your own list responds to.

Full-size tests, every single send

No held-back sample deciding for everyone else. Every variant reaches a fair share of your whole list, so the result reflects your real audience.

See past the bots to the people

Automated opens sit in their own column, apart from human opens. The version with more real readers is the one that actually counts.

Mail that doesn't look like one blast

Several genuinely different versions going out keep your send from presenting as one identical email to every inbox provider reading it.

You pick the winner, on your terms

No early, noisy number gets to crown anything automatically. You look at the results when you're ready and decide what runs next.

Spot the difference? We certainly did.

COMPARED

A fairer way to test what works

A typical email platform

Tests a small sample, then guesses for the rest
Crowns a winner within the first hour
Counts every open, bots included
One identical email to every inbox
A/B means two options, if you're lucky

Senderoo

Full audience. Every variant reaches an even share of everyone, every time.
You decide. Results sit there until you choose to act on them.
Human opens only. Automated opens are reported separately, never as a win.
Several designs. Variants in play, so your send doesn't read as one blast.
Several designed versions. Subject, opening, offer or layout, tested together.

QUESTIONS

Creative variants, explained plainly

Is this the same as A/B testing?

It's a fairer version of it. Every variant goes to an even share of your whole audience rather than a held-back sample, and the results count real people's opens, not security scanners.

How many versions can one campaign have?

A campaign can launch with several designed variants, differing by subject line, opening, offer or layout. Each one is a complete, finished email, not a single word swapped out.

Does the test use my whole list, or just part of it?

Your whole list. Variants rotate evenly across everyone on the send, so the result reflects your actual audience at full size, not a guess made on fifty people and applied to the rest.

How do you know who really opened an email?

Every open is classified. The ones that arrive the way automated scanners and preview tools open mail are reported separately from the opens that come from real people reading on their phone or laptop.

Who actually picks the winning variant?

You do. Results are there to look at whenever you're ready, but nothing is crowned automatically on an early, noisy number. You choose what to keep running.

Does sending different versions hurt how my mail is received?

The opposite. A handful of genuinely different versions keeps your send from looking like one identical blast to the inbox providers reading it on the other end.

Can I test more than one thing at once?

Yes. Each variant is a whole designed email, so the subject line, the opening, the offer and the layout can all change together between versions, not one element at a time.

Do I need design skill to build each version?

No. Pair this with the AI writing in your campaign brief to draft each variant in your voice, then adjust what you like before you launch.

WORKS WITH