Stop using carousels on your homepage

Sliders. Carousels. Hero rotators. Whatever name your designer uses, they're the same thing — a stack of slides that auto-rotate at the top of your homepage. They've been a staple of corporate websites since 2008.

Almost every UX study published since 2013 says the same thing: carousels reduce conversion rates. Yet they keep showing up in new builds, including ours occasionally when a client insists.

This is the data, the reasons it's so consistent across studies, and what to use instead.

The data

The Nielsen Norman Group, the most-cited UX research firm, has published multiple studies on carousels. The summary, repeated almost verbatim across their reports: only the first slide gets meaningful attention. Subsequent slides drop off a cliff.

Slide position	Click-through rate (avg)
Slide 1	~89%
Slide 2	~3%
Slide 3	~2%
Slide 4+	<1% combined

That's not a small effect. That's "the four slides you spent two weeks designing are essentially invisible to 96% of your visitors."

Why carousels fail

Auto-rotation = fake animation

The brain is wired to detect motion. Every time the carousel changes slides, your visitor's attention is yanked away from whatever they were reading. If they were skimming your value proposition, the slide change interrupts them. They might not even consciously notice — but their reading flow breaks.

Worse: if a slide changes just as they go to click on it, they end up clicking the wrong slide entirely. This is the "rage click" pattern that Hotjar's data shows on most homepage carousels.

The user is never on your timer

You set the slide to rotate every 5 seconds because that "feels right." Some visitors read at 2x that speed, some at half. The slide either changes before they're done reading or sits there forever after they've moved on. There is no rotation timing that fits everyone.

Mobile users hate them more

Touch interfaces interpret carousels as swipeable, but most carousel implementations don't quite work right on mobile. Either the swipe is laggy, or the dots are too small to tap, or the carousel auto-rotates while you're trying to read.

And in 2026, 80%+ of Indian web traffic is mobile. Whatever's broken on mobile is broken for most of your audience.

SEO penalties

Carousels usually:

• Have multiple H1 tags (one per slide), confusing search engines.
• Slow down LCP because they preload images for slides nobody will see.
• Add JavaScript that hurts INP scores.
• Hide content behind interaction (slide 3+ is invisible to crawlers in some implementations).

The "but my client wants one" objection

Stakeholders ask for carousels for one specific reason: they have multiple things to feature and don't want to choose. Marketing wants the campaign hero. Sales wants the product launch. Founder wants the press mention. Carousel is the political compromise that nobody dies for.

This is real. The work isn't about UX research — it's about helping the client make a hard choice. Which is the actual job we get paid for as web designers.

What to use instead

Option 1: A single, focused hero

Pick one. Make it the strongest one. Test others against it. This is what every site converting above industry average does.

Example: Stripe's homepage. One hero. One value prop. One CTA. No rotator. They don't lack things to say — they're choosing what to say first.

Option 2: A static feature grid below the hero

Three or four cards in a grid, each linking to a sub-page. Visitors can scan all options at once. No motion. No timer. Each card competes on its own merits.

Option 3: A horizontally scrollable shelf (mobile pattern)

Like Netflix shelves. The user controls when to scroll. Works because users have learned the pattern and expect static content until they touch it.

Crucially: no auto-scrolling. The minute you add auto-scroll, you're back to the carousel problem.

Option 4: Tabbed sections (if features are categorical)

If the "carousel" is really "we have 4 different products" — that's a tabbed interface, not a slider. Each tab loads its own content, and the user controls which one they see.

The exception that almost-but-not-quite proves the rule

There's one place carousels work reliably: product detail pages on e-commerce sites. A product carousel showing 5–8 photos of the same product (different angles, zoomed details) is genuinely useful. The user has already decided to look at this product; they want to see all of it.

This works because:

• It doesn't auto-rotate.
• The content is highly related (different views of one thing, not different competing offers).
• The user is in a focused, depth-seeking mindset, not a scanning mindset.

This is not the same animal as a homepage hero rotator. Don't conflate them.

The summary

Homepage carousels are a UX anti-pattern that survived because they politely solve an internal political problem. The cost of solving that political problem is a measurable hit to your conversion rate.

Pick one hero. Make it your best one. The other things you wanted to feature can live further down the page, in a feature grid, in a navigation menu, or on dedicated pages.

If you're still tempted: at minimum, kill the auto-rotation. Make it manually controlled. That single change recovers most of the lost conversions, even if you keep the slider.