If you’re going to run an A/B test…
… say on 2 different subject lines, or 2 headlines for a landing page, or 2 AdWords ads…
Never EVER test 2 slightly varied pieces. That’s really easy to do. And tempting. And it appeals to your inner-lazy. And you’ve heard the gurus blabber about how they once tested a eensy-weensy variation on a page… and got a 6,000% boost in conversion.
Despite all this…
Don’t do that. Don’t test what the Conversion Rate Experts call “meek tweaks.” If you do that, you’ve already failed the test before you’ve even run it.
Instead…
You should ALWAYS test radically different, big, BOLD variations on a headline. Or subject line. Or landing page. Because when you test big, bold changes… you have a much bigger, bolder chance of getting bigger, bolder results with one of the variations.
i.e…
This would be a terrible A/B test:
Subject line A: “I can’t believe I’m doing this…”
Subject line B: “I can’t believe I’m doing this sale…”
Why?
Because the variation is so small.
Just 1 word changed.
Better:
Subject line A: “I can’t believe I’m doing this…”
Subject line B: “People are telling me I’m insane for running this sale…”
Even better:
Subject line A: “I can’t believe I’m doing this…”
Subject line B: “I might go broke soon. Here’s why…”
Even better:
Subject line A: “I can’t believe I’m doing this…”
Subject line B: “The quest for true life ends here”
See?
Test big changes.
And by that last example…
You can probably see why it’s the most extreme change: the idea underneath the subject line… the “hook,” the angle I take to try to sell the email is 100% different than variation A.
Test changes like that. Not “meek tweaks.”
