Okay kids, today, we’re going to talk about image sliders and conversion rates, otherwise known as World War III!
This is one of those areas where conversion optimizers and designers find themselves at variance, and occasionally break out into altercations.
On the one hand, you have conversion optimizers who hate sliders and everything related to them. On the other hand, you have designers, whose templated site designs and trend-sensitive design proclivities are steering them toward the siren song of sliders.
What should be done?
This article tells all.
What are sliders?
An image slider is a changing image, usually positioned at the top of a website’s homepage.
You’ve seen these before. You probably know exactly what I’m talking about. Even so, let’s get some examples in front of our eyeballs.
Here’s an example of an image slider from Rodania. They make watches, and feature a slider you can’t ignore.
A little clicky-click or waity-wait, and voila! The image changes.
Other sliders may have a bit more design snazzy, but the concept holds true. Here’s a German site, Maisen Gasse.
And here’s another image slider example in the website of a graphic designer. Yep, those are sliders you see.
On the other end of the profession spectrum, here’s a bakery website.
I threw in this website slider example, because, hey, dovecotes.
You get the picture.
There are slider hybrids and carousel offshoots. Some sites, such as Amazon, feature a row of images on a carousel that displays items that other customers bought.
Sliders are used for just about every type of site imaginable. Sliders are ubiquitous. They are sexy. They are cool.
Why would a website use image sliders?
The answer comes down to one word: design.
Less space. More visual.
Most of the time, designers use sliders because they allow maximum visual impact with minimum real estate. I mean, why would you not want to pack five images into the space of one, huh?
Let’s not forget that sliders use motion. Whether they have a fade, a crumble, a swipe, a rotate, or whatever, the slider uses some form of cut to change views. The transition itself, not to mention the new scenery, helps to engage the senses and attract the eyeballs.
Part of the problem emerges when we see designers using image sliders to elicit conversions. Although most designers use sliders to enhance a site’s design, they may sometimes think that the sliders contribute to improved conversions.
Most sliders are clickable, essentially comprising a CTA. To add more visual appeal, copy and focus on a CTA, then the designer places that CTA within the slider.
In the example above, this financial institution is attempting to gain conversions by use of the text on the slider. The conversion action is to apply for an auto loan. The slider switches a few seconds later, encouraging the user to apply for a home loan, etc.
This is usually not a very good idea. More on that later.
What are image sliders for?!
So, now we know what sliders are. We know why designers want to use them.
But here’s the thing: Nearly everyone in the conversion community declares sliders to be unequivocally evil.
Yes, they actually use the term evil.
Rotating banners are absolutely evil and should be removed immediately.
Who said it? Tim Ash of Site Tuners on ClickZ.
Other conversion people and developers have chimed in, hurling epithets at sliders. Notice some of the big guns who blasted sliders.
Don’t use them. Ever. (Lee Duddell)
Sliders are distracting. (Hiten Shah)
Sliders … deliver little to no value to the customers. (Avinash Kaushik)
Sliders suck. (Thijs de Valk, Yoast)
Sliders suck. (Bryan Eisenberg)
Carousels are completely ineffective. (Craig Kistler)
Remove carousels immediately! (Harrison Jones)
First off, we need to figure out why, where and when an image slider is being used. What is the purpose of the slider? Why does it exist?
You can’t legitimately hate something unless you first understand why it’s there to begin with. So, why are people using sliders?
A slider exists for one of three reasons:
- To show nice pictures.
- To present content.
- To navigate to a product, department, information or other place on the website.
Let’s consider those issues one-by-one.
To show nice pictures.
In this, the slider’s sole raison d’être is to look nice. There is no verbal value to the slider. It’s just a nice looking thing.
To present content.
Often, a slider is intended to give the user information. It shows a slogan, headline, etc.
To navigate to a product, department, information or other place on the website.
In this case, the slider’s sole purpose is to drive traffic and conversions to a given destination.
In order to answer the “should I use a slider or not?” question, you need to first answer the question, why are you using a slider? What is its purpose?
This is the critical drop-off point for most civil discussions. People hear “slider” and they immediately start getting angry.
Wait a minute! Hold on! What are you getting angry about?
First, figure out the purpose of the slider.
Only then can you start hurling epithets or shouting curse words, or whatever it is you’re doing.
The “study” that caused everyone to go astray.
Conversion optimizers love data and studies, so naturally, there’s a study that many people turn to in order to validate their stance against sliders.
The big “study” that everyone turns to?
Take, for example, this article by VWO.
I thought I’m the only one who couldn’t stand these automatic sliders. But thank god, there are others like me as well! A usability study by Neilson Norman group confirmed that auto-forwarding carousels annoy users and reduce visibility.
The fact that they misspelled Nielsen isn’t my biggest concern with the article. (It’s an i-before-e thing. Confusing, I know.) The problem has to do with the “study” itself. I’ll tell you why this study is way off in just a minute.
But first, take a look at everyone who is using this study to bolster their blather.
VWO cites the study, as does ConversionXL. They praise Jakob Nielsen as a guru, and even do him the honor of actually spelling his name correctly.
ConversionXL, too, seems to think that the study is actually legitimate.
Here’s yet another conversion consultant, Erik Runyon, citing Nielsen’s study, and slaughtering the name spelling with perfunctory totality.
Even the reputable SEL cites the venerable Nielsen in their take on the topic:
Yoast even goes so far as to call it “science.”
So what is this “study” everyone is talking about?
Read it for yourself. The “study” consists of one user on one website.
Go ahead, check it out here: http://www.nngroup.com/articles/auto-forwarding/
That’s not a study. That’s an anecdote.
Look at my highlights. Nielsen isn’t trying to hide the fact that he tested one user on one website. That should be obvious if you have low-to-medium reading skills.
What Nielsen does, however, is calls this a “study” and declares, therefore, that “carousels and accordions annoy users.”
Users, plural? Carousels plural?
Wait, man, your study was of one woman on one website, and you are making pluralized statements, and popping it up as a full-fledged usability study on your website?
And we’re going to draw some world-changing massive implications from that? What else would you like to tell me about your amazing powers of inductive reasoning?! Puh-LEEZ!
I’m reacting with some vigor, because I see so much emotion expended upon this subject. What it rests on, however, is extremely shaky “evidence.” It’s one person.
And who knows? Nielsen does have this thing about “banner blindness,” and maybe he saw this “study” as the perfect soapbox for more fire-and-brimstone sermonizing on the banner blindness.
But let’s be real. Nielsen came up with this banner blindness thing in 1997! That was literally 18 years ago! In digital years, it’s like 500!
I must be intellectually honest, and dismiss Nielsen’s “study” as pure crap.
Let’s look at some more “evidence.”
The folks at ServerTastic decided to test their image slider using VWO/GetClicky. They had a slider that was generating revenue. They split test the slider vs. no slider, and came up with the following assessment.
At the moment, the non-slider variant has a 36% chance of beating the original. However the sample size is fairly small at present so we will have to wait and see which one wins! I will report back once we have more data.
Sorry to disappoint you, but that’s one website, one slider, and one very small data set. So let’s go ahead and toss the data out the window.
There was no slider on Servertastic at the time I wrote this blog post.
Erik Runyon is one of the few data collectors that did a great job of getting some stats together.
He tracked traffic and CTRs for ND.edu, which at the time had sliders.
He tabulated the CTRs on the five sliders, and came up with a not-so-surprising finding:
Here are the raw numbers:
Homepage visits: 3,755,297
Percentage that clicked a feature: 1.07%
Percentage of total clicks for each position:
Position 1: 89.1%
Position 2: 3.1%
Position 3: 2.4%
Position 4: 2.8%
Position 5: 2.6%
Let’s ask the question: What were the image sliders for?
Runyon doesn’t explain, other than to say that the sliders were calls to actions. This is what he mentioned when he first ran the study from 2012-2013.
Since Runyon provides dates for his data-culling, I went to the Wayback Machine to find out what exactly he was measuring. What were the sliders for?
I’m still not sure. The page has two sliders. The first slider is the main background image, which cycles through pictures of the campus buildings. The second slider cycles through who-knows-what, but presumably these are the CTAs to which Runyon is referring.
I’m left without a full picture of the situation.
However, what I do know is interesting. 1.07% of all users clicked the CTA on the slider.
Honestly, that’s not a bad conversion rate. For an educational site that serves a wide variety of users and has a very expensive and complicated product, getting a whopping 1% to convert on a given slider, crowded as it is by so many other navigational and aesthetic elements, that’s not too shabby!
A final touch of evidence comes from Brian Massey at the Conversion Sciences. Surprisingly, they come out in favor of image sliders. Why? Because the data told them so.
But there are some pretty stringent qualifications for what constitutes a legitimate and nicely converting slider.
- Small images (519 x 319 pixels)
- Slow fade
- 10-second transitions.
They used four panels:
Here is the bottom line from their study:
In our test, the rotating header beat the static image by 61% with a confidence of over 99%.
So, the slider worked.
Let’s ask the right questions and get the right answers.
I wanted to wrap this article up with a bow, but as I dove into the data, the bow got tangled.
After some careful reflection, I have a few simple instructions. You’re wondering whether to use a slider.
Question: Should you use an image slider or not?
Consideration: Why do you want to use one? What is its purpose? To present information, to show a nice picture, or to elicit action from the user?
If the answer is to show a nice picture or share some information, you’re not risking your conversion rates. Don’t sweat it.
If, however, you are attempting to elicit CTRs from the slider, then do this: Test it.
The bottom line is simple: Always be testing.
In my experience, sliders usually don’t boost conversion rates. Why not? Because they are not necessary, and because they are used for the wrong reason — to elicit CTRs.
The more CTAs you target on a single page, the smaller percentage of users will click on any one of the multiple CTAs. Reduce your number of CTAs, and you will improve the CTRs for the remaining CTAs.
In rare situations, however, sliders work.
Back to you — what are you going to do? Figure out why it is you want to use a slider, then do a split test.