TL;DR
If you want to improve your Lighthouse score, the work is mostly mechanical. Lighthouse is a free Google tool that grades your site on performance, accessibility, best practices, and SEO from 0 to 100. Performance is the category most owners struggle with, and the one tied directly to revenue. Below 50 means you are losing visitors and search rankings.
The biggest gains come from compressing images, removing unused JavaScript, and fixing layout shifts. Those three changes alone can push a score from 40 to 75 or better. A free audit at PageSpeed Insights takes under a minute. Most fixes cost between $500 and $5,000, and the payback comes from better Google rankings and fewer visitors bouncing off the site.
Table of contents
- What is a Lighthouse score
- Why your score matters for revenue
- How to run your first Lighthouse audit
- Understanding your results
- The 8-step improvement plan
- Before and after: real score improvements
- When to fix it yourself vs. hire a developer
- Industry benchmarks
- FAQ
- Reflecting on what actually moves the score
What is a Lighthouse score
Lighthouse is a free tool built by Google that grades your website on four categories.
| Category | What it measures | Why it matters |
|---|---|---|
| Performance | How fast your pages load and respond | Slow pages lose visitors and rank lower on Google |
| Accessibility | Whether people with disabilities can use your site | Affects 15-20% of users and has legal implications (ADA) |
| Best Practices | Security, modern standards, error-free code | Outdated practices create vulnerabilities and trust issues |
| SEO | Basic search engine optimization | Missing basics mean Google cannot properly index your pages |
Each category gets a score from 0 to 100. Google color-codes them.
- 0-49 (Red): Poor. Needs immediate attention.
- 50-89 (Orange): Needs improvement. Money is being left on the table.
- 90-100 (Green): Good. Your site meets Google's standards.
Think of it like a car inspection. Green means everything passes. Orange means things need fixing before they become expensive problems. Red means the inspector politely suggests you walk home.
Performance is the category most websites struggle with, and the one that directly affects revenue. That is where I will spend most of this article.
Why your score matters for revenue
A Lighthouse score is not a vanity metric. It connects to three things business owners care about: search rankings, visitor behavior, and conversion rates.
Google uses performance data to rank websites. Core Web Vitals, the real-world speed metrics that feed into your Lighthouse score, became a Google ranking factor in 2021. Sites that score green get a ranking boost. Sites that score red get penalized. Two competing sites with similar content? The faster one wins.
Visitors leave slow websites. Research from Google and Deloitte shows that 53% of mobile visitors abandon a page if it takes longer than 3 seconds to load. A Lighthouse performance score below 50 almost always means a 4+ second mobile load. That is half your visitors gone before they see the product.
Speed affects conversion rates. Vodafone improved their Lighthouse score by 31 points and reported an 8% increase in sales. The same pattern showed up on the Imohub real estate portal: sub-0.5s queries across 120k+ properties indexed, top-three Google rankings, and a 70% infrastructure cost cut after a focused rebuild. On Cuez, the API went from 3 seconds to 300ms — a 10x improvement that unlocked the user growth they had been blocked on.
How to run your first Lighthouse audit
Running an audit takes less than a minute and costs nothing.
- Open pagespeed.web.dev in your browser.
- Type your website address and click "Analyze."
- Wait 15-30 seconds while the tool tests your site on both mobile and desktop.
You will see four colored circles with your scores. Below those, the tool lists specific issues and estimates how much each one affects your score.
Mobile results usually score lower than desktop because phones have less processing power. Google cares more about mobile scores because most web traffic comes from phones.
One important note. Lighthouse scores fluctuate between runs. You might get 62 on one test and 58 on the next. That is normal. Run the test three times and use the middle number as your baseline.
Understanding your results
The scores themselves are useful. The real value is in the diagnostics below them. Lighthouse tells you exactly what is slowing your site down and estimates how much time each fix would save.
Performance metrics that matter
Lighthouse calculates your performance score from six metrics. Three carry the most weight.
Largest Contentful Paint (LCP) measures how long it takes for the biggest visible element (usually a hero image or headline) to appear on screen. Target: under 2.5 seconds. This single metric accounts for 25% of your performance score.
Interaction to Next Paint (INP) measures how quickly your site responds when someone clicks a button or fills out a form. Target: under 200 milliseconds. If your site feels laggy when visitors try to interact, this is why.
Cumulative Layout Shift (CLS) measures how much the page content jumps around while loading. You have seen this. You start reading text, then an ad loads above it and pushes everything down. Target: under 0.1. Google penalizes this because it ruins the user experience. Read Google's official explainer for the full definition.
For a deeper explanation of all three metrics in plain language, I wrote a separate guide: Core Web Vitals for Business Owners.
The diagnostics list
Below the metrics, Lighthouse shows "Opportunities" and "Diagnostics." Each item lists the issue, estimated time savings, and a color-coded priority. Sort by estimated time savings. The item at the top is usually the single biggest thing you can do to improve your score.
The 8-step improvement plan
These steps are ordered by impact. Fixing step 1 alone can sometimes improve your score by 20-30 points.
Step 1: Compress and convert your images
Images are the heaviest files on most websites. A single uncompressed photo can be 3-5 MB. Your entire page should ideally be under 1.5 MB total.
What to do:
- Convert images from PNG or JPEG to WebP format. WebP files are 25-35% smaller with no visible quality difference.
- Resize images to the actual dimensions they display at. If an image appears at 800 pixels wide on your site, do not upload a 4000-pixel original.
- Use a compression tool like Squoosh (free, by Google) or TinyPNG to reduce file size further.
Expected impact: Compressing images typically improves your performance score by 15-25 points if your site currently uses unoptimized images.
Step 2: Remove unused JavaScript
JavaScript is the code that makes your website interactive: dropdown menus, animations, forms, analytics. Most websites load far more than they need. WordPress sites with 15+ plugins are the worst offenders.
What to do:
- Lighthouse lists unused JavaScript under "Reduce unused JavaScript." Share this list with your developer.
- Audit plugins, widgets, and third-party scripts. That abandoned A/B testing script from 2024 is still loading on every page, quietly.
Expected impact: 10-20 point improvement.
Step 3: Fix layout shifts
Layout shifts happen when elements move around as the page loads. Common causes: images without defined dimensions, ads that load late, and fonts that swap in after the page appears.
What to do:
- Add width and height attributes to every image so the browser reserves space before it loads.
- Load web fonts with
font-display: swapso text appears immediately. - Move ads or dynamic content into fixed-size containers.
Expected impact: 5-15 point improvement, plus a more polished feel for visitors.
Step 4: Enable browser caching
Caching tells a visitor's browser to save files locally so they do not need to re-download them on the next visit. Without caching, every page view re-downloads your logo, your stylesheets, and your JavaScript files.
What to do:
- Ask your developer to set cache headers for static files to at least 30 days.
- If you use a CDN like Cloudflare or Vercel, caching is often enabled by default but may need configuration.
Expected impact: 5-10 point improvement on repeat visits.
Step 5: Defer non-critical JavaScript and CSS
By default, browsers stop rendering your page while they download JavaScript and CSS files. Deferring tells the browser to load these files in the background without blocking the page from appearing.
What to do:
- Ask your developer to add
deferorasyncattributes to non-critical scripts. - Move analytics, chat widgets, and social media scripts to load after the main content appears.
Expected impact: 5-15 point improvement, especially on mobile.
Step 6: Use a content delivery network (CDN)
A CDN stores copies of your website on servers around the world. When someone in London visits your site, they get files from a London server instead of one in Virginia.
What to do:
- If you host on Vercel, Netlify, or Cloudflare Pages, you already have a CDN.
- If you host on traditional hosting (GoDaddy, HostGator), put Cloudflare (free plan available) in front of your site.
Expected impact: 5-10 point improvement, with bigger gains for visitors far from your server.
Step 7: Minimize third-party scripts
Every third-party tool on your website (Google Analytics, Facebook Pixel, live chat, heatmaps) loads its own JavaScript. Five tools can add 500KB-1MB of extra code.
What to do:
- List every third-party script on your site.
- Remove any you no longer actively use.
- For the ones you keep, ask your developer to load them asynchronously.
Expected impact: I have seen removing two unused scripts improve a score by 10 points.
Step 8: Optimize server response time
If your server takes more than 600 milliseconds to respond (a metric called Time to First Byte, or TTFB), everything else stacks on top of that delay.
What to do:
- Check your TTFB in the Lighthouse results under "Initial server response time was short."
- If TTFB is high, the fix depends on the cause: underpowered hosting, slow database queries, or missing server-side caching.
- Upgrading from shared hosting ($5/month plans) to a proper hosting platform often cuts TTFB in half.
Expected impact: 5-15 point improvement.
For a full checklist of speed improvements with cost estimates, see my website speed optimization guide. If the fixes sit at the framework or rendering level, my website development service covers the rebuild path, and my custom applications service handles the same work for product UIs.
Before and after: real score improvements
Here are typical results I see when working with clients on Lighthouse improvements.
| Client type | Before | After | Key fixes | Time spent |
|---|---|---|---|---|
| E-commerce (Shopify) | 32 | 78 | Image compression, removed 4 unused apps, deferred scripts | 2 days |
| Professional services (WordPress) | 41 | 85 | WebP conversion, plugin audit (removed 8 plugins), added caching | 3 days |
| SaaS marketing site (Next.js) | 58 | 94 | Code splitting, lazy-loaded images, optimized fonts | 1 day |
| Local business (Squarespace) | 45 | 72 | Compressed images, reduced custom code blocks, optimized video embeds | 1 day |
The first 30-40 points of improvement are always the easiest. Going from 35 to 75 means fixing obvious issues. Going from 75 to 95 requires rewriting parts of the codebase.
Platform matters too. Squarespace and Wix have a performance ceiling because you cannot control the underlying code. Getting above 80 on those platforms is difficult. Next.js or Astro sites can reach 95+ with proper optimization. The LAK Embalagens corporate site is a good example: a React + Next.js + TypeScript build that delivered a 45% bounce rate reduction and 3x Search Console impressions.
When to fix it yourself vs. hire a developer
Some Lighthouse improvements need zero technical skill. Others need a developer.
You can fix these yourself
- Compress images before uploading. Use Squoosh or TinyPNG.
- Remove unused plugins from WordPress, Shopify, or your CMS.
- Delete third-party scripts you no longer use (old analytics, abandoned chat widgets).
- Add image dimensions in your CMS if it exposes width/height fields.
- Enable CDN caching through your hosting provider's control panel.
You need a developer for these
- Code splitting and lazy loading (restructuring how JavaScript loads).
- Font optimization and critical CSS extraction.
- Server response time fixes (database optimization, server configuration).
- Framework-level optimizations (server-side rendering, image pipelines).
What it costs
| Fix category | DIY cost | Developer cost | Time |
|---|---|---|---|
| Image optimization | Free | $200-500 | 1-2 hours |
| Plugin/script audit | Free | $300-800 | 2-4 hours |
| Caching setup | Free (basic) | $500-1,500 | 2-8 hours |
| JavaScript optimization | N/A | $1,000-3,000 | 1-2 days |
| Full performance overhaul | N/A | $2,000-8,000 | 3-5 days |
If your score is below 50, hiring a developer almost always pays for itself within 3-6 months through better rankings and conversions.
Industry benchmarks
| Industry | Typical score | Competitive score |
|---|---|---|
| E-commerce | 35-55 | 75+ |
| SaaS / Tech | 50-70 | 85+ |
| Professional services | 40-60 | 80+ |
| Media / Publishing | 25-45 | 70+ |
| Local business | 40-65 | 80+ |
The goal is not perfection. Run Lighthouse on three competitor sites. If they all score 45-55 and you score 80, you have a meaningful advantage in Google rankings.
FAQ
Does Lighthouse score directly affect Google rankings?
Lighthouse scores themselves do not directly affect rankings. But Core Web Vitals, measured by the same underlying metrics, are a confirmed Google ranking factor. A poor Lighthouse performance score almost always means poor Core Web Vitals, which does hurt your rankings.
What is a good Lighthouse performance score?
A score of 90 or above is considered good by Google's standards. For most business websites, aiming for 75-85 on mobile is a realistic and competitive target. Anything below 50 needs immediate attention because the site loads slowly enough to lose significant traffic.
How often should I run a Lighthouse audit?
Run an audit after every major change to your website (new pages, plugin additions, redesigns) and at least once per quarter as a routine check. Third-party scripts update themselves, content grows, and scores drift downward over time without maintenance.
Can I get a perfect 100 score?
Technically yes, but it requires a very minimal website. Every analytics script, font file, third-party widget, and complex animation reduces your score. A 100 is possible for a simple landing page but impractical for most business websites. Aim for 90+ and spend the remaining effort on content and conversion optimization instead.
My desktop score is 90 but my mobile score is 55. Which matters more?
Mobile matters more. Google uses mobile-first indexing, which means your mobile performance determines your search rankings. A 90 desktop score with a 55 mobile score means Google sees your site as a 55. Focus optimization on mobile first.
Will improving my Lighthouse score also improve my conversion rate?
Usually yes, but not always one to one. Conversion depends on the offer, the copy, and the trust on the page. Speed removes friction, it does not generate desire. The pattern I see most often: a score climb from below 50 to above 75 reduces bounce rate first, and conversion follows over the next 30-60 days as more visitors stay long enough to read the offer.
Reflecting on what actually moves the score
After 16 years and 250+ projects, the surprise is how often the same five issues do most of the damage. Oversized images. Plugins nobody uses anymore. Tracking scripts that quietly multiplied. A rendering choice nobody revisited. Hosting that was right two years ago and is wrong now.
A Lighthouse score is a thermometer. It tells you the temperature, not the cause. The score moves when the cause moves. If you fix the five common issues above, the score takes care of itself.
The other thing worth saying: a 100 is not the goal. A site at 88 with a clear offer outperforms a site at 100 with no offer every time. Fix the obvious problems, get into the green, and put the remaining effort into the words on the page.
If your score is below 50 and you are not sure where to start, I review sites and tell owners which fixes will have the biggest impact, with a cost estimate before any work begins. The Cuez API rebuild, the Imohub portal, and the LAK Embalagens site all started with the same kind of audit. If your numbers are red, that is the place to begin. Get a quote in 60s on the contact page.