Design by Committee: Why Carousels Are Evil
Image courtesy of Tim Sanders
This piece was written by Jon Powell, and Jasper Kuria both consultants at The Conversion Wizards.
As the adage goes, a camel is a horse designed by committee. Sure, beauty may be in the eye of the beholder, but by most objective standards, a horse is good-looking while a camel is not! [1]
Many poorly performing pages we come across have an element of this ‘design by committee’. Different departments sat around a table and the result was a homepage (or landing page) designed to appease each stakeholder.
This tendency is a variation of the well-known Conway’s Law, so insidious in large organizations that Microsoft’s CEO, Satya Nadella, recently warned against it, in an email to all his employees. The law is often stated as:
"Organizations are constrained to produce designs which are copies of the communication [or power] structures of these organizations [rather than catering to the customer’s need]"
Anyone remember a product from the mid-2000's called Microsoft (R) Windows (R) Live (tm) Hotmail(tm) Passport (tm)?
It was supposed to be the one login to rule them all! Rather than having multiple passwords for all your internet sites, you would have just one digital master-key. A neat idea, but what a name?
Any company division that managed to pull off such an idea would naturally have a lot of power. I can picture divisional vice-presidents at Microsoft jostling for position, each wanting a piece of the action--and thus the name! This is likely what Satya was warning against.
But I digress. As digital marketers and online designers, we are charged with increasing the bottom line and finding those reliable conversion gains, and yet we are constantly having to play negotiator, and referee, while doing so.
In fact, the larger the organization you’re in, it seems the bigger the war for space on the website, and the more difficult it is to do everything you need to get those conversion gains.
Well let’s talk about one of the easiest opportunities for a solid conversion gain on your website. It’s likely on your homepage (or your friend’s homepage). It’s the called the carousel.
Name not ringing a bell? Here’s an example:
(image found on: https://www.smileycat.com/carousels-design-gallery/ )
And why do these carousels exist?
Someone must have thought they were cool. And the rest of the internet followed.
OR… someone thought they would please multiple, warring departments in their company by giving them EACH a rotating spot in the prime eye-path. This is likely the case with a company like AT&T Wireless:
I know this because as part of my CRO practice, I worked with a very large wireless company in the U.S. and this is exactly how they operated.
This was also true with the mid-size Industrial IoT company I directed digital marketing for last year. Every product department thought they were the most important or had very little empathy for the next.
They would fight, go over heads, demanding priority treatment and pointing at some piece of evidence that revealed they weren’t getting their fair shake.
And so what would my then VP of Marketing boss do? Give them their space in the carousel!
The only problem is that those turns on the carousel meant nothing for anyone. Why? Because users aren’t seeing the information in them, except maybe the first one.
Look at view and click data for any series of homepage carousels and what you’ll find is a dramatic drop-off starting with the 2nd panel of the rotation. Some users don’t even get as far as the end before already having clicked on something or scrolled down the page.
For those of you who have already read a dozen different carousel articles, one study you will find commonly cited is the one conducted by ND.edu (University of Notre Dame).
At the time (2013), it is told that www.ND.edu was running one of these infamous information carousels in the hero spot on the homepage.
And here was the percentage of clicks for each position:
(source: https://erikrunyon.com/2013/07/carousel-interaction-stats/ )
And if you read the source article, you will find more proof as the author examines what the data looks like across very different sites (an aggregate news site, departmental site, executive site).
In fact, I thought I saved those stats from the IoT company I worked for but misplaced them. The result, however, was the same: first position gets the lion’s share, everyone else get’s crumbles.
But still, some people digress. Fortunately, I have actually put it to the test in years past:
New Data: My Own Carousel Test
When I was working with an organization called RegOnline (before they were acquired by CVent), one of the first series of tests we were running was on their homepage.
We had recently been through a workshop with business unit owner concerning their new messaging and wanted to determine if it would be more effective to spread the message out in a carousel or to cram it all on one static view.
Here was the static view:
Yes, it seems a bit wordy, and jam-packed. But the entire message was there!
So the test would be to determine if carousel design could attract more conversions as a result of spreading out the message, giving people more room to read and absorb, and also showing more employees—one of their goals was to humanize the organization.
Here are all 4 panels as part of that treatment.
Treatment 1
Treatment 2
Treatment 3
Treatment 4
And here’s a glimpse of the result: Can you guess which version is the orange line and which is the yellow?
Yellow is the carousel, and orange is the static. The blue line represents the original page without the new messaging.
As you can see, the carousel ended up converting worse than if we did nothing, compared to the static, seemingly wordy approach.
If you were to compare carousel versus non-carousel, there was a 28% relative gain in end-conversions (sorry, we cannot share the original data points).
What are Your Alternatives?
If you think about the original situation, remember that undoing a carousel often means re-negotiating a peace treaty.
One of the best pieces of advice I received from a colleague and friend was this: “In you want people to want to do business with you, they need to like you.”
And that’s the ONE thing I don’t see any of these other carousel articles addressing in detail. (I can’t go on writing an article without having a value proposition, right?)
So, before you go and end the life of the carousel in the name of CRO (Conversion Rate Optimization)--a great way to create enemies and sometimes a CLM (Career Limiting Move)--consider the content first and its owners. Rarely, if EVER, will you encounter a carousel that doesn’t have at least two different people’s interests at stake.
Based on my experience working directly with multiple product and division owners, all in different parts of the globe, the great wisdom of those Star Trek TNG writers from the early 90’s sticks in my head.
Quoting the famed negotiator in the episode Loud as a Whisper, here is the secret:
“My technique was to look for something, no matter how small, that was common to both groups, and then to begin a process where one person or one group expresses themselves to each other.”
(source: http://www.chakoteya.net/NextGen/132.htm)
Imagine you get two or three of the owners and pull them onto a call. And then imagine you present them with “newly discovered data” (just look in your Google Analytics) that not everyone is getting a fair share.
Then the conversation turns to the key performance indicators that matter to them, eventually leading to those that also matter to you.
From there? You can get their help sorting through the different ideas. Because if you don’t, you’re either going to be doing a lot of re-work, or, you will make an enemy or damage an alliance.
Maybe the solution is to designate one month to a specific product type in that hero area whilst killing the carousel functionality.
By executing that idea with the team, we took advantage of the static approach and dedicated even more of the homepage to what mattered to each product owner. We just made sure to line up their ‘five minutes of fame’ (more like a month in this case) with the timeframe that mattered most to them.
Each stakeholder’s goals were represented, and in the end, we determine how valuable the homepage was for anyone at its peak of focus. This provided me with the ammunition to re-focus test different strategies supported by the data.
So, if you’ve got a carousel, or a friend that needs to eliminate one, now you know that:
a) It is a bad idea, even for a SINGLE content owner (just look at the test I ran)
b) You’re going to need more than a designer and some CSS / HTML / whatever code to pull it off.
You’re going to need to reach into your inner negotiator and use the opportunity to deepen good ties, figure out just how valuable that space is (even in static form), and ultimately, get those green arrows going up!
Are you looking to generate more sales, leads or app downloads? Steal our proven cheat sheet for Landing Page Optimization here.
--------------------
[1] In the original quote the goal was presumably beauty and so compromise yields a less than ideal design. For landing pages, the goal is usually revenue, leads or app downloads and the most aesthetically pleasing design may not be the highest converting one.
Point is, clearly define the goal and then optimize ruthlessly for that. Another quotation on this topic that frequently comes to mind is from one of the greatest actors of our time, Sidney Poitier:
“Never do an artistic favor for a friend. By god loan them money and be there for them in hard times, but If you want to do great work, there must be at least one area of your life where there is no room for compromise.”
(He received this advice from an old, wise, director friend early in his career, when he reluctantly agreed to do a part that he didn’t really like to help this director friend)
