Should I Use a Carousel?
shouldiuseacarousel.com
shouldiuseacarousel.com
It truly is the cause of and solution to all of your website's problems.
Or to view it more cynically, SEO is the Swiss Army knife that can simultaneously excuse away poor performance while justifying additional expenditure.
It was written probably a decade ago but is unfortunately a classic.
<p hidden>Senior management's latest idea</p> if (user != seniorManagement)
display = hidden if (user == onBehalfOfSeniorMgmt)
doGrayball> Greyball was used by ride-hailing company Uber to evade city regulators and deny service to some customers...
What the hell, Uber?
> Uber did not receive any formal punishment or restrictions from the city.
What the hell, government?
[0]https://www.imdb.com/title/tt11173006/?ref_=nv_sr_srsg_0
Edit: Travis is a horrible human being during the time of Uber. Maybe he has matured and mellowed out now? However, his stink is still in my mind and has forever tainted Uber for me. I will never use Uber on any of my devices, ever.
One example is that it was used to prevent taxi drivers in South American countries from targeting Uber drivers for beatings, car jackings or worse. Regardless of what you think about Uber or the taxi monopoly, it was a good thing to prevent physical harm to drivers on the service.
Even the supposedly bad uses of Greyball are, well, in the grey. Government regulators, often working at the behest of medallion owners (who generally don't drive taxis*), preferred to use fear as a tactic. They would target the drivers with whatever traffic tickets they could find, knowing that it is difficult for these drivers to fight. Greyball made this tactic harder, which forced cities into court directly with Uber, a much fairer matchup.
The really odd thing about greyball, though, is that there's actually no user requirement that the cars in the Uber app be real. They could have just "greyballed" everyone, and use standard anti-fraud practices to prevent someone requesting many trips and never starting them.
* Even where medallion owners are required to be owner-operators, they can often just sit in the taxi queue at the airport doing a crossword for a few hours.
After some confused troubleshooting, I discovered he was looking was printouts of the website he had his secretary make so he could look at it at home.
Had it load a completely different CSS when the logged in user was her account. Good times.
This is when the app forces you to click on one tip after another before you are allowed to use the app.
It just frustrates the user who just wants to get to the button that does the job. He doesn't want to pay attention to your tips. If your app is hard to figure out without those tips, redesign it.
[1] https://www.nngroup.com/articles/mobile-instructional-overla...
Even if I wanted tips, brain is full.
He says much the same, but has some good ways to fix it in a game world. I think it's much harder to have the environmental queues in an app, but I'd be curious what it'd look like.
Example:
You are a contractor who makes their living off of their reputation. You have a set of testimonials (maybe 5-8).
A user visiting your website may be browsing for several reasons - and may or may not be interested in what others have to say about you. Collapsing all of the testimonials down into a single carousel shows you have testimonials and allows the user to browse through them if they’d like without forcing them to scroll through each one.
This content is repetitive - if you’ve seen one that conveys enough information - but each one potentially provides incremental reassurance for a user if they need that.
This is one of those things that really needs A/B testing. I'd bet the other direction: that sites get more conversions if they show a few testimonials on the landing page, than if there's a link to a page full of them but none directly on the landing page. Of course, you could just do both.
This page executed flawlessly on this requirement. Bonus points for having one slide with very short text to make me click through and the next slide with very long text.
And so since its not optimizing for engagement, it is purely aesthetic, but aesthetic in the sense that it can have poor features that other serious companies use. A Fortune 500 website would have notoriously shitty aspects, do that if you want audiences that think that is clout.
Fortunately my non-North American and younger North American audiences don't even care about websites. All commerce is driven straight through chat apps. And do I really want the clientele that thinks I need a website that they will accidentally find on a search engine, or by accidentally typing a .com in the address bar by habit? no. I've considered make my fonts smaller and thinner hoping they think thats a problem and bounce, I like the aesthetics of thinner fonts but its more like thinking maybe I can just ignore the issue from the people who are less likely to be able to read it since they're not the target audience anyway.
I'm pretty much never doing things for SEO. Everyone's just going to click through from a chat app, or a twitter feed because someone else was talking about it. Since a lot of people swear by other e-commerce books that never made them a dime and have opposite advice, maybe I should release my own.
This sounds either very labor-intensive if you use humans, or like a bad customer experience if you use bots.
People get their information from heavily populated channels (one way communication chat rooms run by personalities people like, defacto "influencers"), which are forwarded to people that didn't read it. Permeating many communities and private group chats in minutes.
If it is labor intensive, its only that way for a week or two as you coordinate all of the posts with many channels.
Your bio talks about fintech and digital assets, so perhaps I was imagining the wrong scenarios.
https://www.nytimes.com/interactive/2020/12/17/world/asia/in...
half way down the page - to suddenly reduce the size of the page content.
The stats don't lie.. but what is a better way to present information?
If you must hide some information for whatever reason (to preserve space, to increase user engagement metrics), perhaps tabs would work. They can have useful labels which are better than the navigation dots on the carousel.
if you don't have space, then you have too much information and need to make prioritization decisions
carousels are often the reflection of the inability to make a decision
1. Write a short summary of every content phase, and initially just straight up list those summaries each linking to details[0].
2. As an improvement, consider if you want the user to click/tap to expand a point without leaving the page. Pay special attention to accessibility, tabindex, aria-describedby et al. as you do that.
3. At certain viewport sizes, you may put details to the side, and highlight the expanded point.
4. Then, as a final step, you could also cycle expanded points until first user input. Respect prefers reduced motion flag.
Voilà, you have a carousel but thought-through, accessible and purposeful.
[0] There is a natural limit on the number of points that can be shown. Make sure to be clear about the purpose of the widget. If a business is attempting to list more than 4–5 points, either there is a problem with focus or there are nearly non-overlapping audiences (e.g., seller/buyer) and resource should correspondingly be split into multiple entry points. These are aspects where design cannot be decoupled from business decisions.
They’re not so good for text heavy or ‘critical’ content that you want people to definitely read.
It helps if the content is visual and users can see the next slide peeking through, it hints that there’s more to see. Autoplay is always bad imo
You can code this with basic CSS too unlike most carousel behaviours that have complex code for simulating scrolling and looping. If you check the open/closed GitHub issues on most carousel plugins, you'll see what a big source of bugs they are.
https://www.nngroup.com/articles/horizontal-scrolling/ (2014) says to avoid horizontal scrolling on desktop but is there anything more recent and with metrics?
Not all carousels are horrendous, but they are all bad. Even worse, all use cases have a significantly better, simpler solution.
Thanks for the share.
Mad Men, Carousel pitch: https://youtu.be/suRDUFpsHus
And the Carousel scene from Logan’s Run: https://youtu.be/viWT4JWWfTg
Forty is the new thirty…
Some previous discussions:
2 years ago https://news.ycombinator.com/item?id=23754676
9 years ago https://news.ycombinator.com/item?id=6018316
Although a carousel to display a group of images ( like car shots, the interior, etc ) seems to be the best of what I know.
Coincidence?
Or maybe the intent is just for the subliminal nature. So when you see the same or related images later, you'll pause just a bit longer.