Jordon Baade •

The Performance Spectrum — What Visual Effects Cost Your Lighthouse Score

You want a website that looks incredible. You also want one that loads fast and ranks well. The good news: you do not have to pick one or the other. The bad news: most people pick the wrong tool and end up with a slow site that looks only marginally better than what they could have built for free.

This is Part 1 of a three-part series where we break down every major category of visual effects on the web, what they actually cost your performance, and how to pick the right one for the job. By the end, you will know exactly which effects are basically free and which ones trade real Lighthouse points for wow factor.

The Lighthouse metrics that actually matter

Before we get into the effects themselves, you need to understand three numbers. You do not need to memorize them or become a performance engineer. You just need to know what they mean for the people visiting your site.

LCP (Largest Contentful Paint) measures how fast the main content of your page appears. Think of it as the moment your visitor stops staring at a blank screen and starts seeing something useful. Google wants this under 2.5 seconds. If your hero image or headline takes four seconds to show up, people leave before they ever read a word.

TBT (Total Blocking Time) measures how long JavaScript locks up the browser. While JS is running, your visitor cannot click buttons, scroll smoothly, or interact with anything. It is the difference between a site that feels snappy and one that feels like it is ignoring you. Every visual effect that requires JavaScript adds to this number.

CLS (Cumulative Layout Shift) measures how much stuff jumps around while the page loads. You have experienced this: you go to tap a button and the page shifts, so you tap an ad instead. CLS catches that. Poorly loaded animations and late-rendering canvases are common culprits.

These three metrics are what Google uses to judge your page experience, and they directly affect your search rankings. When we talk about the "cost" of a visual effect below, we are talking about what it does to these numbers.

The spectrum

Not all visual effects are created equal. Some are genuinely free. Others cost you real performance. Here is the full breakdown, from lightest to heaviest.

Free tier: CSS animations

CSS transforms, opacity fades, and keyframe animations are the closest thing to a free lunch in web development. They require zero JavaScript, add zero bytes to your bundle, and run on the browser's GPU compositor thread — completely separate from the main thread where your page logic lives.

That means they cannot block interactions, cannot delay your LCP, and will not cause layout shifts if you stick to transforms and opacity. Your Lighthouse score will not move at all. A CSS-animated hero section scores identically to a static one.

The limitation is complexity. You are not going to build a 3D product configurator with CSS keyframes. But for hover effects, entrance animations, floating elements, gradient shifts, and subtle background motion, CSS handles it beautifully. Most sites do not need anything more than this, and the ones that do often need it on only one or two pages. See what is possible with pure CSS in our CSS-only showcase.

Cheap tier: Vanilla JS and Canvas

One step up from pure CSS, you have lightweight JavaScript effects: Canvas 2D particle systems, scroll-triggered reveals with Alpine.js, and simple parallax layers. The JS footprint here is tiny — roughly 2 to 10KB — and everything runs on requestAnimationFrame, which means it plays nicely with the browser's rendering cycle.

The Lighthouse cost is minimal. You will see a small bump in TBT for script evaluation, maybe 5 to 15 milliseconds on a mobile device. That is well within Google's "good" threshold. CLS stays clean as long as you size your canvas elements properly upfront.

This tier is where most marketing sites should live. You get genuinely impressive visuals — particle fields, parallax depth, animated skylines — without any meaningful hit to your score. Take a look at our Parallax Storytelling, Canvas Particles, and City Skyline showcases to see what this tier can do.

Mid tier: GSAP ScrollTrigger

GSAP is the industry-standard animation library, and ScrollTrigger is its scroll-driven animation plugin. Together they weigh about 45KB gzipped. That is not trivial, but it is not alarming either. What you get in return is serious: pinned scroll sections, scrubbed timelines that animate in sync with the scrollbar, horizontal scroll journeys, and choreographed multi-element sequences.

The Lighthouse cost lands in the 20 to 50 millisecond TBT range on mobile for script evaluation. That is noticeable in your score — you might drop a few points — but it is entirely manageable. The key is loading GSAP only on pages that need it, not globally across your entire site.

This is the sweet spot for landing pages, product launches, and storytelling experiences where scroll interaction is the whole point. If your page is designed to be scrolled through like a narrative, GSAP earns its weight. See it in action at our GSAP ScrollTrigger and Horizontal Journey showcases.

Heavy tier: Three.js and WebGL

This is the top shelf. Three.js alone is 180KB or more gzipped. Raw WebGL with custom GLSL shaders can be lighter on download but heavier on GPU. Either way, you are running continuous render loops that demand real resources from your visitor's device.

The Lighthouse cost is significant: 50 to 150 milliseconds of TBT on mobile just for script evaluation, and LCP often gets delayed because your canvas does not render meaningful content until after JavaScript has fully loaded, parsed, and executed. On a mid-range phone, that can mean an extra one to two seconds before your hero content appears.

None of that means you should never use it. It means you should use it deliberately. A 3D globe showing your global presence, a morphing geometry that reacts to the cursor, a product configurator that lets customers see exactly what they are buying — these create experiences that no CSS animation can match. The cost is real, but so is the impact. Explore what this tier looks like in our Globe, Morphing Geometry, Product Configurator, Terrain Flythrough, and WebGL Shaders showcases.

The tradeoff is not binary

This is where most conversations about performance go wrong. People frame it as "fast or pretty" — pick one. That is not how it works.

A marketing hero section can absolutely afford a Three.js scene. That page's job is to make a first impression, and first impressions are worth a few Lighthouse points. An e-commerce product listing page, on the other hand, needs every millisecond it can get because speed directly converts to revenue there.

The right question is never "should I use visual effects." It is "which effects match this page's job." A homepage hero and a checkout page have completely different performance budgets. Treat them that way. Use the free and cheap tiers everywhere, and spend your heavy budget only where the payoff justifies it.

What is next

This article gave you the map. The next two parts zoom in on the territory.

In Part 2: The Lightweight Wins, we break down the CSS and vanilla JS tiers in detail — specific techniques, when to use each one, and how to get maximum visual impact for zero (or near-zero) performance cost.

In Part 3: When Heavy Is Worth It, we tackle Three.js and WebGL head-on — the real-world performance numbers, the mitigation strategies that keep your scores respectable, and the specific use cases where the tradeoff genuinely pays for itself.

The bottom line: you do not have to choose between a site that looks stunning and one that performs well. You just have to pick the right tools for the right pages. Now you know what those tools cost.