Updated April 18, 2026

How to Improve Page Speed

Each additional second of load time costs you conversions. Here are the specific fixes, ordered by impact.

https://
FreeNo signup~1 minute

Page Speed: the conversion impact

Page speed affects conversions more than most teams realize. Research from Google and Deloitte consistently shows that each additional second of load time reduces conversions by 7-20%. A page that loads in 5 seconds instead of 2 is leaving 20-40% of potential conversions on the table.

But most "speed optimization" advice is generic: "optimize your images, minify CSS, use a CDN." You probably already know that. Here are the specific, high-impact techniques that actually move the metrics Google cares about — LCP, CLS, INP, and FCP — with the conversion data explaining why each matters.

The metrics that actually matter

LCP (Largest Contentful Paint) is the metric with the strongest conversion correlation. It measures when the largest visible element finishes rendering — usually your hero image or headline text. Google's threshold: under 2.5 seconds is "Good." Over 4 seconds is "Poor." Most landing pages we analyze have LCP between 3-6 seconds, which means most pages are leaving money on the table.

CLS (Cumulative Layout Shift) measures visual stability — elements jumping around as the page loads. It destroys trust. A page where buttons shift position as ads or images load feels broken, even if it's technically functional. Google's threshold: under 0.1 is "Good."

INP (Interaction to Next Paint) replaced FID in 2024. It measures how fast the page responds to user interactions — clicking a button, expanding an accordion, typing in a form. Sluggish interactions feel broken. Under 200ms is "Good."

Step-by-step guide

1

Fix Your LCP Element First

Identify your LCP element (usually the hero image or headline) using PageSpeed Insights or Chrome DevTools. If it's an image: convert to WebP or AVIF (30-50% smaller), set explicit width/height attributes, add fetchpriority='high', and preload it with <link rel='preload'>. If it's text: ensure fonts load with font-display: swap and preload your primary font file. Target: LCP under 2.5 seconds.

2

Eliminate layout shifts from images and embeds

Every image and iframe needs explicit width and height attributes so the browser can reserve space before loading. Use CSS aspect-ratio as a fallback. For web fonts: font-display: swap prevents invisible text. For dynamic content (ads, embeds): reserve space with min-height. Run Chrome DevTools Performance panel and look for the 'Layout Shift' entries to find the specific offenders.

3

Audit and defer third-party scripts

Analytics, chat widgets, marketing pixels, and A/B testing scripts are often the biggest performance killers. Each third-party script adds DNS lookups, connection time, and execution time. Audit every external script: does it need to load on initial page render? Move non-critical scripts to defer or async loading. Consider loading chat widgets only after user interaction. One large SaaS client we worked with cut LCP by 1.8 seconds just by deferring their chat widget.

4

Implement proper image loading strategy

Hero image: preloaded, WebP/AVIF, responsive srcset. Below-fold images: loading='lazy'. All images: explicit dimensions. Use next/image if you're on Next.js — it handles optimization, lazy loading, and responsive sizes automatically. For non-Next sites: use a CDN with automatic format negotiation (Cloudflare, Imgix, Cloudinary). This alone typically saves 40-60% of image bytes.

5

Reduce Main-Thread JS for Better INP

Heavy JavaScript blocks the main thread and makes interactions feel sluggish. Code-split your bundles so only critical code loads initially. Use dynamic imports for below-fold components. If you have a React/Next.js app, wrap interactive components with lazy() and Suspense. Move expensive computations to Web Workers. Target: total blocking time under 200ms.

6

Set up a CDN and optimize server response time

Your Time to First Byte (TTFB) is the floor — nothing renders until the server responds. Use a CDN (Cloudflare, Vercel Edge, AWS CloudFront) to serve pages from edge locations near your users. For server-rendered pages: implement caching headers aggressively. For static pages: pre-render at build time. Target: TTFB under 800ms, ideally under 200ms from the nearest edge.

Common questions

Which Core Web Vital has the biggest impact on conversions?

LCP (Largest Contentful Paint) shows the strongest correlation with conversion rates. A 1-second improvement in LCP has been shown to increase conversions by 8-15% in multiple studies. CLS is second — layout shifts erode trust and increase bounce rate.

Is mobile speed or desktop speed more important?

Mobile. Google uses mobile-first indexing, and mobile connections are slower by default. If your page is fast on desktop but slow on mobile, you're losing the majority of visitors. Always optimize for mobile first, then verify desktop.

Do I need a perfect 100 PageSpeed score?

No. Diminishing returns kick in hard above 80. Going from 40 to 70 has a massive conversion impact. Going from 90 to 100 is barely noticeable to users. Focus on getting into the 'Good' range for all Core Web Vitals rather than chasing a perfect score.

How do I know what's slowing my page down?

Run PageSpeed Insights (or use roast.page, which includes PageSpeed data). Look at the 'Opportunities' and 'Diagnostics' sections. Chrome DevTools Performance panel shows a waterfall of exactly what loads when. WebPageTest gives the most detailed breakdowns.

Will page speed improvements actually show in my conversion rate?

If your current LCP is above 3 seconds, almost certainly yes. The correlation between speed and conversions is well-documented across industries. The effect is strongest for ecommerce and mobile-heavy traffic. Improvements are usually visible within 2-4 weeks of data.

Related reading

Page Speed and Conversions: The Data

Hard data on how load time affects conversion rates.

Page Speed and Conversions

The complete guide to speed optimization.

Mobile Landing Page Optimization

Mobile-specific speed and conversion techniques.

Explore more

Page Speed Analyzer

Analyze your page speed with conversion context.

Mobile Page Analyzer

Mobile-specific page analysis.

Landing Page Analyzer

Full 8-dimension analysis including performance.

Resources

2026 Research Report

1,000+ pages scored. Median: 44/100. Full benchmark data by industry and dimension.

Free Tools

Headline analyzer, CTA checker, page speed audit, and more.

Statistics & Benchmarks

Conversion rates, score benchmarks, and data from 5,000+ page analyses.

Learn

Comprehensive guides on landing pages, CRO, copywriting, and design.

Industry Guides

Tailored analysis for SaaS, ecommerce, healthcare, and 50+ industries.

How-To Guides

Step-by-step playbooks for hero sections, CTAs, trust signals, and speed.

Tool Comparisons

See how roast.page compares to HubSpot, Hotjar, PageSpeed, and others.

Glossary

CRO and landing page terms explained — from A/B testing to visual hierarchy.

AI Prompts

Ready-to-use prompts for headlines, CTAs, pricing pages, and copy.

Alternatives

Compare the best landing page analyzers, CRO tools, and website graders.

Blog

Data-driven insights on landing page optimization and conversion rates.

See how your page scores

Free analysis. Specific fixes. About 1 minute.

https://