
SEO

An effective hero image reinforces the headline's value proposition, keeps the primary call to action as the highest-contrast element on screen, loads as the page's Largest Contentful Paint element in under 2.5 seconds, and holds its focal point at every breakpoint from 375 px to 1920 px wide. Format, compression, and crop decisions carry as much weight as art direction.
Your hero image is the first thing visitors see, and it is often the reason they leave. A slow-loading stock photo paired with a vague headline pushes potential customers away before they read a word about your product.
Most hero sections fail on the fit between their parts rather than on the quality of any single part. This guide covers where that fit breaks down, then the specifications and practices that fix it. If you are rebuilding the whole page rather than just the hero, our B2B web design guide covers the structure around it.
A hero image is the large visual at the top of a webpage that sets the tone and communicates your brand identity within seconds. Hero image best practices center on three things: a clear value proposition, balanced visual weight, and strong mobile responsiveness. When the hero works, visitors understand what you offer and where to click.
Hero images serve a different purpose on each page type:
The hero section is the full band at the top of a page that holds the headline, supporting subhead, primary call to action, and hero image or video. The hero image is one element inside it.
The distinction matters in practice because the parts compete. The headline, the visual, and the button all draw from the same pool of attention and the same vertical space, so a hero section of a website has to be optimized as a unit.
A hero image that wins on its own terms, by being large and striking, usually wins at the button's expense. Judge every hero image decision by what it does to the section, not by how the image looks in isolation.
A hero can look polished and still fail to convert. These are the five failure modes we see most often.
Your headline carries the value proposition. The image supports that message.
When the two feel disconnected, visitors struggle to understand what you are offering. This happens when teams pick a striking image first and write a headline around it. Flip that order. Write the headline, then find or commission an image that reinforces the claim it makes.
Visual weight refers to the element that draws the eye first, based on size, color, contrast, or position. If your hero image dominates the viewport while the CTA sits quietly in a corner, you are guiding attention in the wrong direction.
Good hero section design frames the action and keeps the button the highest-contrast element in the band. Squint at your page. If the button disappears before the image does, your visual hierarchy needs work.
Stock photos of smiling professionals in conference rooms do not communicate what makes your product distinct. Visitors filter out generic business imagery because they have seen it hundreds of times.
Custom visuals, product screenshots, and brand illustrations perform better because they feel specific. On a tight budget, stylized illustration usually beats stock photography. Our B2B website design services cover this work when the in-house option is not available.
An unoptimized hero image directly hurts your Largest Contentful Paint (LCP), the Core Web Vitals metric measuring how long the largest visible element takes to load. According to the HTTP Archive Web Almanac, 73% of mobile LCP elements are images, and on most pages that image is the hero.
The cost of getting this wrong is measurable. In Google and SOASTA research, the probability of a mobile visitor bouncing rises 32% when load time goes from one second to three. Google also uses LCP as a ranking factor, so a heavy hero costs you both the visitor and the search visibility that brought them.
Desktop-first heroes crop awkwardly, hide text, or lose their focal point on smaller screens. Mobile is where most of the performance penalty lands too: the Web Almanac found 16% of mobile sites lazy-load their LCP image, which delays the exact paint the metric measures.
Test your hero across device sizes before launch. If the focal point shifts or the CTA becomes hard to tap, you have work to do.
srcset and sizes so a phone downloads a 750 px asset rather than a 1920 px one.fetchpriority="high" on the hero image and never apply loading="lazy" to it, because it sits above the fold and lazy loading delays the Largest Contentful Paint.The headline defines your value proposition. Start there, then select or create an image that reinforces it. This keeps the visual in service of the message and avoids the trap of falling for an image that does not support what you are trying to say.
Abstract or heavily artistic imagery asks the visitor to interpret before they understand. Their attention is measured in seconds.
Three quick checks:
Clear beats creative when conversion is the goal.
Structure the hero so the eye flows from headline to image to CTA. Use whitespace, contrast, and positioning to make the button impossible to miss. The image frames the action; the button closes it.
For B2B SaaS, showing the actual product interface builds credibility and sets accurate expectations. Product screenshots tell visitors exactly what they will get.
If your product UI is not polished yet, custom illustration handles abstract concepts well. Either option beats generic stock photography because it feels specific to your offering.
Compress your images and use modern formats. Google reports that WebP images run 25% to 34% smaller than comparable JPEGs at equivalent quality, and AVIF typically compresses tighter still.
Since the hero is usually the LCP element, optimization here has outsized impact on the whole page. Webflow development services handle much of this automatically when the images are configured correctly, which is one reason we build on the platform.
Accessibility improves usability for everyone.
When you bake text into an image, screen readers cannot read it and search engines cannot index it.
On contrast, reach for a gradient scrim rather than a flat overlay. A flat overlay dulls the entire image to protect the one corner where the text sits, which costs you the image to save the copy. A gradient darkens only where it is needed.
Design for mobile constraints first, then scale up for desktop. This keeps the focal point, text size, and CTA effective on small screens.
Desktop-first heroes create problems that surface late and cost real money to unwind, because fixing them usually means recropping or recommissioning the asset rather than adjusting CSS.
| Use Case | Recommended Dimensions | Aspect Ratio |
|---|---|---|
| Full-screen desktop | 1920 × 1080 px | 16:9 |
| Short desktop band | 1920 × 820 px | 21:9 |
| Standard banner | 1600 × 900 px | 16:9 |
| Tablet | 1200 × 900 px | 4:3 |
| Mobile portrait | 750 × 1000 px | 3:4 |
| Retina displays | 2x the display size | Varies |
A 16:9 aspect ratio works for most full-width heroes. Width is the number that matters: 1920 px covers nearly every desktop viewport, and 2560 px covers retina density where the design calls for it.
The right website hero image size is the smallest one that still looks sharp at your widest breakpoint. Browsers crop height to the viewport anyway, so extra vertical pixels buy you file size and nothing else.
Serve mobile a separate 750 x 1000 px portrait asset rather than letting CSS crop the desktop file. A phone that downloads a 1920 px image to display it at 390 px wastes roughly 80% of the bytes it paid for, and that waste lands directly in mobile LCP.
Test the focal point across breakpoints. What sits centered on desktop can crop badly on a phone.
Two different problems need two different tools, and teams routinely reach for the wrong one.
Resolution switching applies when the mobile image is the same composition at a smaller size. Use srcset and sizes and let the browser pick:
Art direction applies when the mobile crop is a genuinely different composition, for example a wide desktop shot that becomes a tight portrait crop on a phone. srcset cannot do this, because it assumes every candidate shows the same thing. Use <picture> with a media condition:
One Webflow-specific catch. Images placed as image elements through the Designer get responsive variants generated automatically. Background images set in CSS do not. If your hero is a CSS background, you are serving one file to every device, which is the most common reason a Webflow hero passes on desktop and fails on mobile. Either move the hero to an image element, or handle the breakpoints yourself with custom code.
WebP and AVIF are the default formats for web images today. Compression can be aggressive, though watch gradients and flat color fields, where artifacts show first.
| Format | Use It For | Notes |
|---|---|---|
| AVIF | Photographic heroes, where your pipeline supports it | Best compression of the five; encoder support is the limiting factor |
| WebP | The default for anything photographic | Broad support, 25% to 34% smaller than JPEG at equivalent quality |
| JPEG | Fallback only | Still the safest universal format |
| SVG | Illustration-based heroes | Scales to any breakpoint from a file measured in kilobytes |
| PNG | Flat graphics with hard edges | Avoid for photographs; the file size is punishing |
Hero images load above the fold, so lazy loading never applies. Two attributes actually move the needle:
fetchpriority="high" on the hero <img> tells the browser to fetch it ahead of other images rather than treating it as one request among many.<link rel="preload" as="image"> in the head matters for CSS background heroes, which the browser cannot discover until it has parsed the stylesheet. Without the preload, the hero starts downloading late no matter how small the file is.The target is an LCP under 2.5 seconds at the 75th percentile, which is Google's threshold for a "good" score.
| Specification | Target | Why It Matters |
|---|---|---|
| Desktop file size | Under 200 KB | A practical budget for staying inside a 2.5 second LCP |
| Mobile file size | Under 100 KB | Mobile connections carry less headroom |
| Compression quality | 75 to 85 for WebP and AVIF | Above 85 the file grows with no visible gain; below 75 artifacts show on gradients |
| LCP budget | Under 2.5 seconds at the 75th percentile | Google's "good" Core Web Vitals threshold |
| Loading attributes | fetchpriority="high", never loading="lazy" | Add <link rel="preload" as="image"> for CSS background heroes the browser cannot discover early |
| Video hero poster | Under 150 KB | The poster is the LCP element until the video paints |
| Video hero file | Under 2 MB, muted, no autoplay with sound | Video is the heaviest hero format by a wide margin |
| Text contrast on image | 4.5:1 body text, 3:1 large text | WCAG 2.2 AA. Use a gradient scrim rather than a flat overlay |
Choose the format by what the reader needs to see, whether that is the product, the outcome, or the brand.
Best for simple, fast-loading pages with a straightforward message. Works when you have strong photography or a clear product shot. Watch text legibility over the image.
Creates a high-impact first impression. Best for brand-forward pages where atmosphere carries weight. Needs more careful optimization and responsive testing than smaller hero formats, and it is the case where <link rel="preload"> matters most.
Effective for showing a product in action or conveying motion. Video adds significant page weight, so use it sparingly. Always load a fallback poster image first, keep it muted, and never autoplay with sound.
Covers subtle CSS animation, parallax, and micro-interactions. Adds polish without the weight of video. Works well for SaaS brands that want a contemporary feel.
The most direct option for B2B SaaS. Shows exactly what the user gets, which builds credibility and sets accurate expectations. Pair it with a value proposition headline.
Three tests tell you whether a hero is working: recall, attention, and load time.
Show your hero to someone unfamiliar with your product for five seconds, then ask what they remember. If they cannot recall your value proposition, the hero is not working.
Heatmap and scroll depth analysis reveals where visitors look and click. If attention clusters on the image and skips the CTA, your visual weight is off. If visitors scroll straight past, the hero is not holding them.
Use Google PageSpeed Insights or Lighthouse to check whether the hero is your LCP element and how long it takes to paint. Under 2.5 seconds at the 75th percentile is the passing mark. Test on mobile first, since that is where the budget is tightest.
Test one element at a time: headline, image, or CTA. Small changes, such as swapping a stock photo for a product screenshot, often beat a full redesign, and isolating variables is the only way to know which change did the work.
Running a clean test needs enough traffic to reach significance and a single conversion goal to measure against. Our conversion rate optimization services cover the setup when the in-house bandwidth is not there.
At Curio Digital, we design conversion-focused hero sections for B2B SaaS companies as part of our UI/UX design and Webflow development work. We have built 250+ Webflow sites, and the hero is usually the fastest place to find a measurable gain on a page that is already getting traffic. If you're comparing partners, here's how we stack up against other B2B web design agencies.
A hero image is the large visual at the top of a webpage, above or beside the headline and primary call to action. Its job is to reinforce the value proposition the headline states, within the two or three seconds a first-time visitor spends deciding whether to stay.
On most pages, it is also the Largest Contentful Paint (LCP)element, which makes it a performance asset as well as a design one.
The hero section is the full band at the top of a page that contains the headline, supporting subhead, primary call to action, and hero image or video. The hero image is one element inside it. Designers optimize the section as a unit, because the headline, the visual, and the button compete for the same attention and the same vertical space.
Use 1920 px wide for full-viewport desktop heroes, or 2560 px if you are serving 2x retina density, at a 16:9 ratio. Standard in-container banners work at 1600 x 900 px. Hold the desktop file under 200 KB in AVIF or WebP. Width matters more than the exact ratio, because the browser crops height to the viewport anyway.
Serve mobile a separate 750 x 1000 px portrait asset rather than letting CSS crop the desktop file. A phone that downloads a 1920 px image to display it at 390 px wastes roughly 80% of the bytes it paid for, and that shows up directly in mobile LCP. Deliver the switch with srcset and sizes, or with the picture element when the mobile crop is a genuinely different composition.
Use AVIF where your pipeline supports it, WebP as the default, and JPEG only as a fallback. Google reports WebP files run 25% to 34% smaller than comparable JPEGs at equivalent quality, and AVIF typically compresses tighter still. For illustration-based heroes, SVG usually beats all three, because it scales to any breakpoint from a file measured in kilobytes. Reserve PNG for flat graphics with hard edges.
The hero image is usually the Largest Contentful Paint element, so its file size and load priority set your LCP score directly. Google's "good" threshold is under 2.5 seconds at the 75th percentile. Compress the file, serve a modern format, set fetchpriority to high, and never apply lazy loading to a hero, since deferring an above-the-fold image delays the exact paint LCP measures.
Describe the image's purpose on the page rather than cataloging what is visually in it. "Analytics dashboard showing pipeline by lead source" tells a screen reader user and a crawler something useful; "screenshot of software" tells them nothing. If the hero is purely decorative and the headline already carries the meaning, use an empty alt attribute so assistive technology skips it.
When you're ready to turn your website into a pipeline engine, here are 3 ways Curio can help:
Contact us
Every engagement starts with a conversation. We’ll learn about your business, your buyers, and where your website is losing them.