Last month we talked about the importance of the first fold of a web. We saw that most of the visitors who land on your website see the first block and, depending on whether they like it or not, continue investigating the rest of the content or leave without looking back.
Among the elements of the first block of a website that most affect visitors’ perception, the featured image is a key factor. It is the first thing they see, so if the feeling that generates is of rejection, we’ll lose our visitors. Sad to say so, but surely you sometimes bought a book just because you liked its cover. Well, with the featured images of your pages the same thing happens.
Further, if you share your content on social networks (if you don’t, I don’t know what you’re waiting for), featured images help to get the attention of the users who receive the message. A tweet linking to a blog post with no featured image is like a garden without flowers.
Testing Featured Images in WordPress
One of the most important pages on our website is this, where we present one of our WordPress plugins, Nelio Content. Visitors rely on this page to end up visiting the pricing page and, even more, purchasing our plugin.
On this page, the first block contains a featured image that shows a smiling man with a post-it on his forehead and a lush beard, in front of a computer. On his monitor we see the editorial calendar of Nelio Content. A rather curious image, chosen by David, as you may expect 😉.
In this test we are going to show different featured images on this same page. The current image is quite peculiar, so as variants we chose two different options. The first alternative is a more conventional image showing a team of people working in front of a laptop. In the second option we have a screenshot of the editorial calendar, to which we have applied a blur effect so that the text above it is properly shown.
To create this test, as always, we use Nelio A/B Testing on our website in WordPress. In this case it is a page test with three variants: the control version and two more, each with a different featured image.
In the screenshot above you can see how we defined the test. Note that we want to measure how many visitors to this page see each featured image variant and end up visiting the pricing page. To do so we set that the conversion action is precisely the action of visiting the pricing page of Nelio Content.
We have edited each variant and the only thing we have changed from the original version is the featured image. The first variant has as a featured image the group of people in front of the computer, while the second has the editorial calendar view. The control version is unmodified, as you may imagine.
After this, we start the test and wait to see how its results evolve. As you can see, the process is not complicated and won’t take you more than a couple of minutes. We do the same in the Spanish version of our website.
I encourage you to try it on your website with an important page to see what you discover and thus improve with real data from your visitors.
Analyzing The Results
Did the bearded man beat the other two featured image variants? Well, apparently he didn’t. Here’s the full test results page:
As you can see, we have had the test running for almost 9 weeks, until we stopped it at the time of writing this article. In all this time, visitors have been seeing the same page but with different featured images. And, without them being aware of it, Nelio A/B Testing has been tracking their behavior (anonymously, of course) to see if they ended up visiting the pricing page of Nelio Content, which was the goal of the test.
In the results we see that both variants are better than the control version. But one is better than the other. Specifically, the image with the group of people in front of the computer is almost 15% better than that of the bearded man. And the results have a statistical confidence over 95%.
Another aspect to highlight here is that we have been able to obtain interesting results in this test, even though the number of visitors analyzed is low. One of the recurring excuses for not doing A/B testing is that the your website does not have enough traffic. Well, now you see that even so, you can always test. Radical changes work much better for low-traffic sites, like testing different featured images.
In the previous screenshot you have the results of the identical test that we have carried out in the Spanish version of our page. It should be noted that in this case the number of visitors analyzed is half the visitors of the Spanish version. Also, the conversion rates of each version are lower in this case for each version. This makes us see something we already know: there is more interest in the product in English than in Spanish.
As you can see, the number of visitors to this website at this time is around 600. We have said this before, but the traffic that our website gets in Spanish with respect to our premium plugins is lower than in English (above you have the results of the test in English, where you can verify that this is so).
Note that the improvement percentage is over 25% in the winning version, which is the same as in the English test. The image with the group of people also wins here, now with a lower statistical reliability, almost 88%.
Even though in previous cases we have not taken results as valid if they did not have a minimum of 95% statistical reliability, in this case we make an exception because the statistical algorithm that we use in Nelio A/B Testing is stricter than those used by other competitors.
If you doubt this, you just have to go to the competition’s websites, look for their A/B test calculators and enter the numbers you see in the results. You will see that in these cases they give much more confidence to these same results. Still, we believe that our algorithm has a strong statistical foundation that validates our numbers.
Seeing this, the only thing left for us to do is use the featured image of the group of people as the default featured image for our Nelio Content page. This way, all visitors will always see it and we will improve our website: more visitors will visit the pricing page of Nelio Content.
The cost of having an A/B test running is low, but the benefits you can get from it are great. Laziness or inaction are enemies of continuous improvement. Don’t let them defeat you and never stop testing changes on your website. Data is always stronger than opinions.
Therefore, after seeing the results obtained with this test, we can affirm that yes, an image is worth a thousand words. Review the featured images you use on your WordPress and test them with a simple A/B test. The results you’ll get may surprise you.
PS: Sorry David, the bearded man you liked so much goes to a better place.