According to the Neilsen Norman Group, no. They annoy users, reduce visibility and just plain do not work.
This is a very interesting topic for me because we get a LOT of requests for carousels. Usually, the client wants to promote a few things on their homepage and carousels are a great way to save space and add some movement.
Nielsen Norman’s study could not be more clear. Don’t do a carousel. Here are some highlights from the article:
- Users ignore design elements that look like ads
- Users do not respond well to clever marketing content
- Accordions and carousels should show a new panel only when users ask for it
- The probability that users will spot the item they want is drastically reduced when only one thing is displayed at any given time
- Because it (the carousel) moves, users automatically assume that it might be an advertisement, which makes them more likely to ignore it.
Read the full article here.
What should you do instead?
Nielsen Norman don’t give a lot of suggestions in the article, but here are our thoughts in order of simplicity:
- Pick on thing to promote at a time. Consider it to be a campaign. Run the banner, track the clicks and then assess the success of the banner at the end of the time period that it runs.
- Randomize your banners. Show a different banner on each page load.
- Employ split testing (aka AB test). Split your visitors into groups randomly. Show each a different banner and see which performs the best (gets the most clicks).
- Use a software like HubSpot to serve up the banners as a smart CTA or as smart content and target the user. For instance, if the user is already a customer, show banner A, if a user is in the middle of the sales funnel, show banner B and so on.
Tagged with: homepage