Ecommerce product hero showcases
Product photo carousels are the one carousel pattern that always works: visitors are already in evaluation mode and want to see every angle. Pair with thumbnail navigation for max engagement.
Pause-on-hover, manual swipe, lazy-loaded images, ARIA roles, zero CLS. Built for product galleries and portfolios where carousels genuinely lift engagement, not just because the homepage looked empty.
Trusted by 11,000+ brands








































Live demo, not a screenshot. Style it, set rotation, embed it. What you see here is what ships to your site.
How to use it
Three steps. Under two minutes. No developer needed, no JavaScript library to babysit.
Drag images into the Poper builder, reorder by drag, add captions, alt text, and click-through links per slide. Original aspect ratios are preserved so nothing gets cropped surprisingly on mobile.
Pick auto-rotate interval (or off), pause-on-hover, pause-on-focus, slide-on-swipe, and visible navigation controls. Each setting has a tooltip explaining the UX research behind it.
Copy the one-line script tag and paste it into your site. Works on WordPress, Shopify, Wix, Squarespace, Webflow, Framer, Ghost, and any HTML stack. Image dimensions are reserved before the carousel renders, so CLS stays at zero.
Works everywhere
Drop-in install on WordPress, Shopify, Wix, Squarespace, Webflow, Framer, Ghost, and any HTML-friendly stack. No build step, no developer needed.
Plus 250+ other platforms via the embed snippet. If your site can render a <script> tag, it can render the Image Carousel Widget: Hero Sliders, Product Galleries, Portfolio Carousels.
Six things that matter when you embed a carousel: usability, accessibility, performance, not 30 features no one uses.
Nielsen Norman Group research is unambiguous: users ignore auto-advancing carousels and frequently miss the slide they were reading because it disappears mid-sentence. Poper ships pause-on-hover, pause-on-focus, swipe-to-advance, and visible prev/next controls turned on by default. Auto-rotation is off unless you opt in. The result is a carousel visitors actually navigate instead of resenting.
Implements W3C ARIA Authoring Practices for carousels: role=region with aria-roledescription=carousel, slide labels, live-region announcements on slide change, prev/next buttons reachable by Tab, arrow-key navigation, and a Pause control when auto-rotation is on. Screen readers announce position (Slide 3 of 7) on every change.
Every slide reserves its width and height before the image decodes, so the page never shifts when the carousel paints. Cumulative Layout Shift stays at 0. Most carousel libraries score 0.15+ here, which Google flags as Poor in Core Web Vitals.
Only the first slide loads eagerly. Subsequent slides decode as users navigate to them.
Real momentum scrolling on mobile, not a fake JS swipe. Tracks finger physics correctly.
Four transition effects, each respectful of prefers-reduced-motion. No motion-sickness gimmicks.
Use cases
Four buyer types who get the most lift from embedding Image Carousel Widget: Hero Sliders, Product Galleries, Portfolio Carousels on their site.
Product photo carousels are the one carousel pattern that always works: visitors are already in evaluation mode and want to see every angle. Pair with thumbnail navigation for max engagement.
Use a manual-swipe carousel (no auto-rotate) to walk visitors through 3-5 product screenshots. Avoid for above-the-fold hero where conversion lives, single static hero with one clear CTA wins there.
Photographers, design agencies, and architecture firms genuinely benefit from carousels because the visual is the value. Use full-width with thumbnail strip and lightbox enabled.
Rotating multiple value props in a homepage hero is the most-criticized carousel pattern in UX research. If you do it, keep auto-rotate off and let visitors swipe. Better: pick the strongest message and ship a static hero.
Most carousel libraries default to user-hostile behavior (auto-rotation on, no pause-on-hover, broken keyboard support). Here is how the popular options stack up.
| Recommended Poper | Elfsight Image Slider | POWR Slider | Slick.js (DIY) | |
|---|---|---|---|---|
| Free plan available | Limited (200 views/mo) | Limited (5K views/mo) | Free (open source) | |
| Pause-on-hover by default | Manual config | |||
| Pause-on-focus (keyboard users) | Manual config | |||
| ARIA role=region + aria-roledescription=carousel | Partial | Partial | DIY | |
| Live-region slide announcements | DIY | |||
| Reserved image dimensions (zero CLS) | Partial | Manual | ||
| Native touch swipe with momentum | Plugin needed | |||
| Lazy-loaded slides past the first | Paid only | Plugin needed | ||
| AVIF + WebP auto-conversion | DIY pipeline | |||
| prefers-reduced-motion respected | DIY | |||
| Bundle size (gzipped) | ~38 KB | ~92 KB | ~110 KB | ~28 KB (no a11y) |
| Bundled with popups, forms, quizzes, more widgets | ||||
| Pricing for unlimited views | $19/mo (Starter) | $5-25/mo per widget | $10-90/mo | Free + dev time |
Comparison reflects publicly listed pricing and library behavior as of 2026. Slick.js is unmaintained since 2017 but still widely used; we list it because it is what most DIY teams reach for first.
Marketers and engineers who ditched user-hostile carousels for one that respects the research.
“Poper has improved our website's user engagement! Since integrating Poper's personalized popups, we've seen a dramatic surge in conversions and user interactions. The platform's intuitive design makes creating and customizing popups a breeze, even for those with minimal tech skills. What truly sets Poper apart are its…”

“Fantastic app and support team, they are very quick to help and provide a solution or improvement to their product. Highly recommended!”

“Poper has been a total lifesaver for our agency! As a digital marketing agency, we’re always juggling a million things at once. Poper has been a real game-changer in terms of streamlining our workflow and keeping track of all our clients’ campaigns. The ability to track all our clients’ websites from one place is a…”

Pricing
Free forever for one widget. Upgrade only when you need faster sync, multiple instances, or to drop the watermark.
Everything you need to ship the widget today.
Remove the watermark, faster sync, more widget instances.
Multi-site, multi-account, white-label.
All plans include unlimited page views, no contracts, cancel anytime. Annual billing saves 20%.
Carousels are one of the most-debated patterns in web UX. Nielsen Norman Group has published multiple studies finding that auto-rotating homepage carousels are largely ignored by visitors, that the first slide receives the overwhelming majority of any clicks, and that motion in the hero area can actively distract users from completing the task they came for. At the same time, manually-controlled image carousels for product photos, portfolios, and feature showcases consistently lift engagement and time-on-page when implemented correctly. This guide walks through what the research actually says, the technical requirements for an accessible and performant image carousel in 2026, and the four placements where carousels demonstrably help (and the one where they almost always hurt).
NN/g has been studying carousels since 2013, and their consistent finding is that auto-rotating homepage hero carousels underperform static heroes for conversion. Visitors interpret motion as advertising and tune it out (banner blindness applied to your own content). The first slide receives 50-89% of any clicks the carousel earns, and slides 4 and beyond often get under 1% combined. The pattern survives anyway because it lets stakeholders avoid choosing one message over another. The honest takeaway: if you are using a hero carousel because the homepage felt empty or because three teams each wanted their slide, ship a static hero with the strongest message instead. If you are using a carousel because visitors genuinely want to browse multiple images of the same thing (a product, a portfolio, a property listing), carousels are the right pattern and they work well.
Even in the placements where carousels work, three controls are required to make them usable: pause-on-hover (so visitors can read a slide without it disappearing), pause-on-focus (so keyboard users do not have content scroll out from under them), and visible prev/next controls (so the navigation pattern is discoverable without hover). Many popular carousel libraries default these to off. The reasoning is usually that auto-rotation 'shows more content' but the actual effect is that visitors miss the slide they wanted, get frustrated, and bounce. Poper ships all three controls on by default, with auto-rotation off by default. If you turn rotation on, the controls remain. This produces a carousel visitors navigate instead of one they ignore.
The W3C ARIA Authoring Practices Guide includes a specific carousel pattern that has been the de facto accessibility standard since 2018. The required attributes: role=region on the outer container, aria-roledescription=carousel, an accessible name via aria-label or aria-labelledby, slide labels with role=group and aria-roledescription=slide, prev/next buttons with descriptive aria-labels, and a live region (aria-live=polite) that announces position when slides change. If auto-rotation is on, a Pause/Play control is also required, and rotation must pause when any element inside the carousel receives keyboard focus. Most third-party carousel widgets implement maybe half of this. Carousels that fail accessibility audits commonly do so on this exact pattern. Poper implements the full ARIA Authoring Practices spec by default.
Cumulative Layout Shift (CLS) is one of three Core Web Vitals Google uses for ranking, with a Good threshold of 0.1, Needs Improvement at 0.1-0.25, and Poor above 0.25. Image carousels are notorious CLS offenders because slides load asynchronously and push surrounding content around when they decode. The fix is to reserve each slide's width and height in the markup before the image loads (either via explicit width/height attributes that match the image's intrinsic ratio, or via aspect-ratio CSS). Poper does this on every slide. CLS for the carousel area stays at 0, which means embedding the carousel does not move your overall page CLS at all. Many third-party carousels score 0.15-0.30 in isolation, which can push an otherwise-Good page into Needs Improvement.
Multiple A/B testing studies (the most-cited one from Erik Runyon at Notre Dame in 2013, replicated several times since) show that on homepage hero placements, replacing an auto-rotating carousel with a single static hero with one clear value prop and one CTA increases conversion by 30-200% depending on baseline. The mechanism is straightforward: visitors who land on a homepage have a brief decision window before they bounce. Motion in the hero competes with the CTA for attention and almost always loses. A static hero with one message and one CTA respects that decision window. Where carousels do win conversion battles: product detail pages (where browsing multiple images is the user's task), agency portfolio pages (where the visual is the value prop), and SaaS feature tour sections (where the user is already evaluating and wants to see the product). The honest pattern for choosing: if visitors are in browsing mode and want to see variety of one thing, use a carousel. If visitors are in deciding mode and need one clear next step, use a static hero.
An image carousel widget is a slideshow component that displays multiple images in a single rotating or swipeable container, used for product galleries, portfolios, and hero sections. Modern image carousel widgets include accessibility ARIA roles, lazy-loaded images, pause-on-hover behavior, and reserved image dimensions to prevent layout shift.
Last fact-checked: . We re-verify every quarter.
Can't find the answer you're looking for? Chat with our support team.
Contact SupportPause-on-hover, ARIA roles, zero CLS. Two-minute embed, free plan, honest UX defaults.
Free plan available forever