If at first you don’t succeed, test, test… and then test some more.

My wife and I often get the same piece of mail (separate lesson: de-duping literallysaves money).

Occasionally, it gives the marketer in me a chance to take a close look at how smart brands are handling multivariate (A/B) testing to improve performance.

Before I dive in, let me say this: if you aren’t testing, you’re doing it wrong. No matter what the communication is or through which channel it is delivered, there is a way to be testing how your audience responds to different variations. They don’t always have to be significant or earth-shaking, but if you aren’t figuring out how to squeeze a few extra percentage points of return out of a communication, you’re losing out.

Take the two Charles Tyrwhitt magazines pictured above. Now indulge me (and/or your inner child) in playing the classic children’s game “Spot the Differences.” How many do you count between the two covers?

I’ll wait…

What did you get?

I count 7 meaningful differences in how the covers of these two magazines are designed. Mind you, they’re selling the same products at the same prices.

That tells me that my friends at Tyrwhitt are thinking seriously about how to make sure people open the magazine, meaning they fear they’re losing conversions from people not even flipping through.

There are some standard tests, including how prominently the sale is displayed (the green and blue circle on the left does a much better job, to my eye) and inclusion of the website (though I don’t know what they hope to prove through that test).

More intriguing to me are the treatment of their logo and their headlines.

The significantly larger logo on the right makes me suspect they have some fear that they just aren’t getting noticed in the stack of mail. Political mail traditionally uses large-format pieces with big type and photos to make sure it gets noticed. It works.

The headline test is the most interesting. Both convey basically the same information — new season, new styles — but the subhead on the right truly caught my eye. Using “Plus Our New How To Wear Guide” suggests that they are trying to position this not simply as a marketing piece but as a true “read” for customers. That’s a way of conveying value to the end user not just in traditional monetary terms.

These tests show that Tyrwhitt is serious about figuring out what makes customers most likely to dive into the magazine and then, ultimately, make a purchase. That’s what makes having sent the magazine worthwhile.

Email marketing, social media campaigns, advertising, even traditional media relations… all of these channels can yield better results through testing.

Are you trying different subject lines? Different post photos? Different times of day?  Different senders? Segmented audiences?

Every additional conversion is another sale… another activist engaging… another vote cast… another relationship cemented.

And the data is all there to prove what worked and what didn’t, so you can test again the next time and keep improving results.

All of this makes me really wish I could get my hands on Tyrwhitt’s response rates. Oh, and a couple of nice new shirts.

I guess their testing worked.


Cross-posted from LinkedIn:


Leave a Reply

Your email address will not be published. Required fields are marked *