A Picture Is Worth A Thousand Words. Right?

In my early days of “blogging for dollars,” I got some advice about affiliate marketing:

If you’re going to promote a product like, say, a standing desk, you should ALWAYS include a “clickable” picture that links to the product.

The reasoning being:

People like to click on pictures. So if you make sure to link the pictures, you’ll get more clicks and more affiliate revenue.

Makes sense, and it seemed to work well enough when I tried it in blog posts.

This technique popped into my head recently when I was brainstorming ways I could improve the click-through rate of several of email courses I’ve created.

Adding an image of the product seemed like a slam-dunk way to increase the number of clicks that my emails generated.

Last week I decided to test this idea out.

At Simple Programmer, we were running a special offer for $7 off on the new T-shirt that my business partner had designed.

I put together an A/B test for the final “this deal ends at midnight” email.

Half of our subscribers got my usual sale-ending email:

A short and sweet reminder of the deadline with a clear call to action, typically less than 100 words.

For the other half of the subscribers, I used the exact same email copy — except that I inserted a linked photo of the new T-shirt.

I’d have bet anyone $20 that the image would spike our click-through rate.

The email went out at 8 p.m. as usual.

And when I checked the stats the next morning…

.

.

.

.

.

.

.

.

… the plain-jane no-image email had stomped the version with the embedded image!

In fact, the text-only version got 44% more clicks than the version with a click-friendly image.

Why?!?

Honestly I’m not 100% sure.

Maybe showing the T-shirt lowered the “curiosity factor” and fewer people clicked through to see what the design looked like.

Or maybe the image pushed the call-to-action link too far “below the fold” and fewer subscribers scrolled past it.

(My money is on that second one.)

All I know is, I’m sure glad I tested this and didn’t just blindly add images to all of our sales emails.

Just goes to show how even what you “know” can be wrong.

P.S. If you’ve ever wondered why I don’t use a “nice” HTML template for these emails… This is why.

Design and images CAN drastically affect the response — and in ways that are hard to predict.

So start simple, and don’t add visual elements just for the sake of “making the design ‘pop’”.

And of course, test test test.

Here’s a good example of where a/b testing went wrong — Wikipedia. Read more here.


Originally published at joshuaearl.com.