We need to be very careful when talking about contexts — design decisions related to context must be based on fact rather than assumption. For example, a context-based assumption would be to change the language of a Web page in response to the location of a device — a terrible idea because the reader might speak a language that isn’t native to the detected country.
Actually it was just me and my colleagues having a rant. 🙂
People don’t like carousels, don’t care what they say
and don’t bother after the first panel.
…so, for the ux-inclined, it’s frustrating that we still use them the way we do.
[B]oy does that homepage look like a juicy piece of prime real estate to a roomful of stakeholders. It’s hard to navigate these mini turf wars, so tools like carousels are used as appeasers to keep everyone from beating the shit out of each other.
This ain’t a cry for blood, though. Typically, a design fails because of a combination of things – the medium, content and audience are incompatible. In the case of the humble carousel, we get the perfect storm of disharmony.
Suppose your amazing marketing message is 4th in line on a carousel that spends 2 seconds per banner. This means the user will spend at least 6 seconds on filler before they get to your gold. That’s a long time in internet years.
The modern website carousel is like a TV ad break, only your favourite show won’t be back after those messages. Is this really the best way get users to pay attention and engage?
After all, e-books and real paper books behave in the same way, sequentially feeding the audience one piece of information at a time – and most of us know how hard it is to put down a good book. This highlights the importance of storytelling in this space, something we tend to miss in the mish-mash of unrelated banners on a rotation.
Brian Krogsgard’s Sliders Suck says it could be totally appropriate to use a carousel when using the carousel is the end goal, not the means to the end.
When a slider item is active, it’s more important than everything else on the page. When a slider item is inactive, it’s completely unimportant.
The implicit assumption of a carousel is that the item on display will be relevant to the person viewing the page. I’m sure I’m preaching to the choir here, saying how one size does not fit all.
And we don’t have to look too deep into behavioural stuff (like banner blindness, tunnel vision and selective attention) or user agent stuff (like hardware capabilities, or compatibility with screen readers) to figure we might be slightly sub-optimal in catering for our users’ needs with sliding banners.
So, what do we do?
The tests say no to carousels, and you’ll find many, many voices in chorus about why. But sometimes we’ll have to use them, and sometimes we’ll actually want to use them – it’ll feel right, even though it seems wrong. We all have nights like this.
The problem isn’t in the doing, but in how it’s done. And we could be doing it less wrong.
Start this way:
- Make the content more interesting. (source 1, source 2) Can’t stress this enough.
- Limit the number of items in a carousel. (source)
- Make the navigation obvious, so users can skip. (source)
- Mind the interaction design stuff, don’t do the annoying crap. (source)
Thank you to @niaalist for sharing the links in this post.
The fundamental attribution error describes the tendency to explain someone’s behaviour based on their disposition or personality, while not considering the situation affecting them.
|Observation||Explain by disposition||Explain by situation|
|Mark jiggles his double chin.||Mark is letting himself go.||Mark looks fat because he’s pregnant.|
|David trips over.||David is clumsy.||Warren was sleeping behind David’s chair.|
|Josh has big hair today.||Josh is a crazy man.||Josh’s hairdryer is a jet engine.|
|Nick is rude to someone.||Nick is a jerk.||Nick is jaded, has been here 12 years and is trying to give up coffee.|
Naturally, this challenges some of our assumptions about web users and the way we as website-makers appeal to them. Are users actually stupid and/or lazy – or do poor design, copy and workflow put them in a position to respond this way?
By the way, the Curse of Knowledge phenomenon makes this entirely plausible.
So – what other observations could we reconsider, in light of the attribution error, during our design process?