A/B Testing Email Subject Lines: What Actually Moves the Needle

Subject lines are the highest-leverage copy in any email. A strong test can lift open rates by 10–30%, but most tests fail because they change too many variables at once or draw conclusions from tiny sample sizes. A disciplined approach turns guessing into a repeatable optimisation habit.

Test one variable per send

Compare curiosity versus clarity, short versus long, personalised versus generic, or emoji versus no emoji — but never all at once. If variant A says “Your March invoice is ready” and variant B says “🔥 Don’t miss our biggest sale!”, you learn nothing useful about what drove the difference.

Choose metrics beyond open rate

A subject line that boosts opens but kills clicks is a net loss. Track click-through rate and revenue per recipient alongside opens. For Australian ecommerce brands, revenue per email is the metric that matters most.

Use meaningful sample sizes

Split your list 50/50 only when each segment has enough subscribers to produce statistically significant results — typically at least 1,000 per variant for smaller lists, or use your platform’s built-in winner selection that sends the winning variant to the remainder after a test window.

High-impact variables to test

• Personalisation: first name vs no name vs location (e.g. “Sydney”)
• Length: under 40 characters vs longer descriptive lines
• Urgency: deadline-driven vs benefit-driven framing
• Questions vs statements: “Ready for tax season?” vs “Tax tips inside”
• Preview text pairing: does the preview text reinforce or repeat the subject?

Document winners and build a swipe file

Keep a running log of what worked for your audience. Patterns emerge over time — B2B service businesses in Australia often respond to specificity, while retail audiences may respond to urgency and offers. Your data beats generic best-practice lists.

Avoid common testing mistakes

Do not test on your smallest segment, declare a winner after 50 opens, or run tests on transactional emails where consistency matters more than optimisation. Save A/B tests for broadcast campaigns and key automation touchpoints where volume supports reliable results.