And here’s one we did on our home page:
You can see that the page on the right had an 8% lower bounce rate (the
bounce rate is the percentage of people who navigated away from our site
without visiting any other pages), so it was the clear winner. Our CEO preferred
the page on the left by a lot. She was shocked the other page won! A split test
helps you look at the cold hard facts; the design you like better won’t necessarily
be the one that works best.
Split testing works the same way with direct mail, albeit with slightly less
guaranteed precision. It’s the best way to continually improve your ROI. Here’s
how it works: Split your mailing list in half. Send your existing postcard to one
half of the list (that’s your “control” or “A”) and send a second postcard to the
other half (that’s your “variable” or “B”). You can split test completely diferent
designs or designs that are just slightly diferent. If you’re pleased with the ROI
on your current design, I suggest testing one change at the time, starting with
the ofer, the headline or text, or the colors.
WORKS THE SAME
WAY WITH DIRECT
THE BEST WAY
IMPROVE YOUR ROI.