Stop 07 · Email & Marketing AutomationTopics · 5/9

A/B Testing Emails

Part of the Digital Marketing Roadmap

In short

You don't have to guess which subject line works. Send two versions to small groups, see which wins, then send the winner to everyone else.

Testing one variable at a time (subject line, send time, CTA) across a split of your list to learn what actually improves open and click rates for your specific audience.

The blob holds out two small bowls of green and red chutney for tasting.

Think of it like this

Testing two chutneys before the party

Before a big family function, Maa makes two chutneys and lets a few relatives taste both. The one they finish first gets made in bulk for 200 guests. A/B testing an email is that tasting round.

Watch it happen

A subject line test on a 10,000 list

List
Split
Wait
Winner
Send rest

1/510,000 subscribers. You have two subject lines and no idea which is better.

The big ideas

What you need to know

One variable

Test subject line or button text or send time. Never several at once.

Random groups

Both groups must be picked randomly from the same list, or the test is unfair.

Pick the right metric

Apple Mail's privacy feature can mark emails as opened. Clicks and sales are more reliable.

Enough people

On tiny lists, small differences are often luck. Test big changes, not tiny tweaks.

Save yourself the pain

Common mistakes

Oops

Calling a winner because A got 21% and B got 20% opens on 200 people.

Do this instead

That gap is likely noise. Use bigger groups or a bolder change.

Oops

Testing the subject line and the offer at the same time.

Do this instead

Change one thing so you know what made the difference.

Your mission

Plan an email A/B test

~15 min

Quick check

Why are clicks often a better way to pick an A/B winner than opens today?

Go deeper

  1. 1Design an A/B test for a subject line, specifying your success metric upfront.
  2. 2Learn why testing only one variable at a time gives you a clean, trustworthy result.
  3. 3Practice interpreting a hypothetical A/B test result and deciding what to roll out.

More topics at this stop

Stuck on this topic? Ask an Insider

Get 1:1 guidance from people who've walked this exact path — on the InsideEdge app.

Download
InsideEdge