I have a new favorite website. It's Anne Holland's Which Test Won. Every week the site features an A/B conversion test and you guess which version won. Once you've voted, you can see how other folks voted, then see the actual results.
My favorite part about testing is learning not only what wins, but why.
In a recent test – Can the Photo Accompanying a Testimonial Make a Difference in Conversions? Happy Baby vs. Sleepy Newborn a company was testing to see if testimonial images had any impact on conversion.
Version A featured a smiling baby. Version B featured a sleeping infant.
I immediately chose version B. I was indeed correct, but when I voted, almost 80% of the participants chose version A.
Why did so many folks get it wrong? I can see how some people would think the big smile on the baby in Version A would be persuasive. The results even surprised the client.
Here's what the article had to say about why version B performed the best:
Surprising to Dr. Gary Berger, changing the testimonial image from a toddler to a newborn caused a 12.6% conversion lift for a special report promoting a health center’s outpatient tubal reversal procedure.
This test shows the critical importance of emotionally-relevant images. A photo of a newborn taps into more would-be mothers’ yearnings than a photo of an older baby.
I agree. When you choose images, you must know what is emotionally relevant to your customer. The key point there being the focus on the customer. Too many times even seasoned professionals go off of their gut. The further away you are from your target audience, the harder you have to work to get out of your head and into theirs.
But I also wonder if there was anything else that affected the results. I wish the two testimonials featured the same sex baby. Could the fact that one was a girl and one was a boy have influenced the results? I don't know.
I also would have loved to see the voting results broken down by gender. Did more women guess the correct image since they can more easily understand mothers or are mothers themselves? Was I tapping into my inner "mother" when I made my choice?
Another thing that struck me was that the photo of the newborn looked more "real." The image of the older baby was obviously more professionally done and might have more of a "stock photo" look to it.
Beware of stock photos
By "stock photo" I mean generic professional stock images that many websites use rather than pictures of real people and real customers. I don't have any proof of this yet, but anecdotally, I'm seeing that women don't respond as well to stock images because they look "fake" to them.
Since women are biologically wired to read human faces, they may be more likely than men to detect when something is real vs. posed.
Thanks Anne Holland for a fabulous website. Go check out Which Test Won.