How can I A/B test email body copy?

Your open rates are fine but conversions aren't moving. That's usually a body copy problem, not a subject line problem. The fix isn't to guess what might work better. It's to test systematically so you actually know.

The golden rule of body copy testing is one variable at a time. Change your opening paragraph, or your offer framing, or your story approach. Not all three at once. If you test multiple things simultaneously, you won't know which change drove the result.

Keep everything else identical. Same subject line, same sender name, same images, same CTA button, same send time. The only thing that differs between version A and version B is the copy element you're testing. Everything else is a constant.

So which variable do you test first? If conversions are low, start with your offer framing. That's the most direct lever between reading and clicking. Try two genuinely different structures, not two ways of saying the same thing. "30% off everything" versus "Buy one, get one free" is a real test. "Free shipping included" versus "Your order ships free" is not. The difference has to be structural, not cosmetic.

If click rates are fine but purchases aren't happening, look at the opening paragraph next. The first two lines set the reader's frame for everything that follows. A problem-led open ("Still paying full price for...") often performs differently than a benefit-led open ("Here's how to save..."). That's worth knowing.

Measure the right things. Body copy doesn't affect opens because the reader has already opened. Your metrics here are clicks, conversions, revenue per email, and scroll depth if your platform tracks it. Open rate is irrelevant for this type of test.

Plan for patience. Body copy tests take longer than subject line tests. Opening an email is a one-second decision. Reading and clicking is not. You need people to actually engage, which means you need a bigger sample and more time before the numbers stabilize. Most ESPs like Mailchimp, Klaviyo, and Brevo will show you a confidence percentage alongside results. Don't call a winner until you're at 95% confidence or higher. Calling it at 70% is how you make changes that don't actually help.

Still a rough benchmark for list size: you generally need at least 1,000 recipients per variant to get meaningful click data. If your list is smaller, run the test over a longer window or accept that results will be directional rather than definitive. Either way, something is better than nothing.

Once you have a winner, that version becomes your new control. Then you test the next variable against it. That's how body copy improves over time. Not in one big overhaul, but in small, proven steps.

Curious what to test after body copy? Tone, length, and offer type each pull different levers. And once you've run a few tests, you'll want to know how to measure content test impact properly beyond just clicks.