A place for NSW Government digital creators to collaborate, share and learn.
There have been some discussions around adding a carousel to our pattern library. Speaking from our team’s experience, this particular pattern does not deliver a user experience we feel is conducive to what we want to deliver for a Design System, from a human-centered design, interaction and accessibility stand point.
However we also acknowledge that it is still a widely used pattern across many sites, so would love to hear your opinions, both for and against.
We typically recommend against any carousels. Generally users do not see anything past the first slide, most scroll past them before the transition even occurs. Many stakeholders view it as a way to cram in more information without realising how poor the usage and interaction analytics really are. My preference is that they are not included and are considered a bad practice by this team that others can use as leverage when a stakeholder tries to shove one into a website.
Nielsen Norman Group did a good study on them and clearly concluded they are bad:
Links through to multiple additional articles and studies:
http://shouldiuseacarousel.com/
If there isn’t a Design System carousel, they will be designed - as we have done.
We tried to ban carousels in our new site design but inevitably they snuck in there by the time the Alpha had done the rounds. For a central website, we recognise there will be a number of priority campaigns or programs and a whole of government site will need to support them. The better we can personalise the site and tailor content to the user, the less we need to “give them everything” on a carousel.
We also currently have them on the nsw.gov.au homepage and limit the number of items to a maximum of 3, as the analytics showed there were minimal clicks on slides beyond the third one.
The carousels that were designed in the alpha prototype was made as accessible as possible. It clearly gives the users an overview of how many frames there is and what the content is in each frame.
@viola.wong do you mean we should have an anti-pattern (i.e. a page about carousels and why not to use them)?
@rika.johnander I’m more concerned about the general UX at this stage, as we can make them accessible.
In my experience, the message from the first blade of the carousel, is all that’s ever interacted with. I don’t have the figures from CBA, but from memory of all interactions of their carousels, somewhere in the high 90’s were from the first blade, with circa 5% from the next blade and all of the rest of the blades added together were less than 1%.
I think stakeholders believe that if they put it up, people will see and interact with it, but they simply don’t.
This is a really interesting idea. Could be a valuable tool for designers to have and be able to send to stakeholders when needing to push back against adding in things we know to be bad. Especially if the anti-patterns pages can have supporting research, analytics, or other data to justify why they are bad and should not be used.
Completely agree with you both @mason.phan and @mike.hall
Carousels can actually be accessible from a pure a11y perspective, like you’ve said.
BUT
1. Suggest non-vital content has made it’s way in to pole position
Carousels usually take top spot on a page, yet they’re doing so to show content which has been deemed so unimportant that it’s OK for it to be hidden 1/5 (or more) of the time
2. Aren’t based on user needs
In 10 years and nobody has ever shown any insights that could be fixed with a carousel
3. Are salesy and a bit corporate (opinion based)
Government websites exist for a completely different reason to corporate websites. The context is different, the user needs are different, and the users themselves are different. I think carousels are the type of thing we inherit from the corp world because stakeholders (and some designers) have ‘seen it done’
Overall I think that using carousels suggests lazy content prioritisation, or a business-based compromise rather than following user needs.
This is interesting!
There’s also implications we’d need to test for users with intellectual impairments and conditions such as autism or other conditions who are severely effected by distractions or movement.
More than, “here’s why we don’t have a carousel and why”, the guidance should explain:
Yes good point Viola
You more clearly articulated my thought of “why”. This is what I would expect behind a why, I was just too lazy to type all that out. 
Carousels are poor UX. Users get little value from them, they eat a huge chunk of real estate and they’re terrible for screenreaders.
However, they do keep managers happy and are often a really good tool to distract those internal clients (with strong business knowledge but poor web knowledge) who would otherwise launch a campaign to destroy site styling and structure.
can you define carousels?
image carousels are very useful for displaying gallery content and reducing the amount of real estate used. carousels can easily be made accessible, it is a myth that they are not accessible. like everything they need to be coded correctly for compliance.
banner carousels are lazy design and should never have been introduced IMO
Hi @danwaters - thanks for your input, this topic is largely around banner carousels, sorry for not stipulating.
Hi all. I spoke recently with one of our developers who provided the following response. Keen to hear your thoughts.
"I have read that Neilsen report, it was written in 2013. So I would like to see some updated data, as we know that things do change ie the above the fold recommendation is no longer an issue. I do agree that content on the positions 2+ do not get the same interaction as the first slide, but I would like to see some more data on what current analytics show. I think it’s pretty important as this is the only argument against the carousel on that thread.
I think from a stakeholder management perspective that a carousel does solve a lot of issues.
So my thoughts are, that I’m not a massive fan but I think they are important from a stakeholder management perspective.
The other component that I think the Design System will need is an image slider or some way to display a set of images which was raised back in February https://community.digital.nsw.gov.au/t/new-patterns-any-ideas/262/2."