Free · Fast · Privacy-first

Test Website Speed Before Launch

Launching a slow website costs you first impressions, search rankings, and conversion rate from day one, and recovering from a poor performance baseline after launch is significantly harder than preventing it before the DNS switch.

Pre-launch performance audit

🔒

Core Web Vitals pass/fail check

Actionable recommendations before go-live

No sign-up required

Cost
Free forever
Sign-up
Not required
Processing
In your browser
Privacy
Files stay local
FreeNo signupWhite-label

Add this Website Speed Test to your website

Drop the Website Speed Test into any page — blog post, product docs, intranet, school portal — with a single line of HTML. Your visitors get the full tool, processed entirely in their browser. No backend, no uploads, no signup.

  • Files stay 100% in the visitor's browser
  • Responsive — adapts to any container width
  • Free forever, no API key needed

Embed code

<iframe
  src="https://www.fixtools.io/web-tools/website-speed-test?embed=1"
  width="100%"
  height="780"
  frameborder="0"
  style="border:0;border-radius:16px;max-width:900px;"
  title="Website Speed Test by FixTools"
  loading="lazy"
  allow="clipboard-write"
></iframe>

Attribution-friendly: a small "Powered by FixTools" link appears in the embed footer.

Pre-Launch Performance Checklist: What to Test Before Your Website Goes Live

Launching a website with poor performance is significantly harder to recover from than preventing the performance issues before launch in the first place. Post-launch speed fixes require careful staged rollouts to avoid disrupting live traffic, and the period between launch and fix means real users and Googlebot encounter poor performance at the critical moment when first impressions, first conversions, and first crawl data are established. CrUX field data also begins accumulating from the moment your site goes live, which means a poor first month can influence Search Console reporting for weeks even after the underlying issues are fixed. A pre-launch performance audit takes 30 to 60 minutes and should be treated as a non-negotiable part of the launch checklist alongside DNS configuration, SSL setup, and content proofreading. The minimum pre-launch performance bar should be a Lighthouse mobile score of 75 or higher and all Core Web Vitals in the Good or Needs Improvement range.

A comprehensive pre-launch speed audit covers five layers in sequence. First, server configuration: verify TTFB is under 500ms, HTTP compression using gzip or Brotli is enabled, and browser caching headers are set with appropriate max-age values for static assets. Second, images: confirm all images are compressed, converted to WebP where supported by the visitor's browser, and have explicit width and height attributes that prevent CLS. Third, JavaScript: confirm render-blocking scripts are deferred or marked async, unnecessary scripts are removed entirely rather than just hidden, and total JavaScript weight is under 300KB compressed for the typical landing page. Fourth, CSS: confirm render-blocking stylesheets are either inlined for critical above-fold content or loaded with appropriate priority hints. Fifth, third-party integrations: every analytics, chat, and advertising script adds measurable load time, so confirm each one is genuinely justified and loaded with async or defer.

Test performance under realistic conditions before launch rather than on your development machine where the comparison is meaningless. Development environments with local servers and uncached builds produce artificially fast or artificially slow performance depending on the specific configuration, and the numbers cannot be relied upon as a predictor of production behaviour. Staging environments running on the actual production hosting infrastructure with the production CDN and caching configuration enabled are the most accurate pre-launch proxy you can build. If your staging environment does not have CDN and caching enabled to match production, you will discover performance issues that only appear in staging but not production, or worse, the inverse case where staging appears fast but production is slow because nobody noticed the caching layer was never enabled in the live environment.

The pre-launch audit is also the right moment to establish a documented performance baseline that you can defend against future regressions. Capture the Lighthouse report for each key page template, save screenshots of the Core Web Vitals values, and record the page weight and request count for each test. Store these artefacts in the project documentation alongside the launch sign-off, because they become the reference point against which any post-launch performance regression can be measured. A team that ships at 85 on mobile but drifts to 65 within a year cannot easily diagnose what changed without that baseline, while a team with documented pre-launch numbers can quickly identify which deployment, plugin, or third-party addition pushed the score down and revert it before the regression compounds further.

How to use this tool

💡

Enter your staging URL to run a complete pre-launch speed test. Fix any failing metrics before switching to your live domain.

How It Works

Step-by-step guide to test website speed before launch:

  1. 1

    Test on staging before launch

    Run the speed test on your staging URL, not your live production site, so any issues you discover can be fixed without disrupting real traffic. The staging environment should mirror production hosting, caching, and CDN configuration so the test results actually predict production behaviour rather than reflecting a development-only environment that bears no resemblance to what users will experience after the DNS switch goes live and traffic starts flowing.

  2. 2

    Test all key page templates

    Test the homepage, a product or service page, a blog post or article, and any page with heavy media, embeds, or third-party widgets. Each template has different performance characteristics because each has different content types, different image weights, different scripts, and different layout complexity. Testing only the homepage gives a misleading picture of overall site performance and risks launching with hidden issues on the templates that actually drive traffic and revenue.

  3. 3

    Ensure all Core Web Vitals pass

    Check LCP, CLS, and INP against the Good thresholds of 2.5 seconds, 0.1, and 200ms respectively on every tested template. Failing any single Core Web Vital at launch means starting with a Page Experience penalty applied from the very first moment Google crawls the site, which delays the eventual ranking benefit that good performance provides and creates an avoidable handicap your competitors do not have to overcome.

  4. 4

    Target 80+ on mobile before go-live

    Do not launch with a mobile Lighthouse score under 70 unless there is a specific business reason that outweighs the SEO and conversion cost of doing so. Address the top failing opportunities shown in the audit report until you reach at least 80 on mobile, and aim higher than that if the site competes in a performance-sensitive category such as e-commerce or news where rival sites already score in the 85-plus range. Delaying launch by a week to hit the target is almost always cheaper than launching at 60 and trying to recover later.

Real-world examples

Common situations where this approach makes a real difference:

Agency handover quality gate

A web agency includes a documented speed test report as part of their formal launch sign-off process, embedding it directly into the project handover checklist that the client must counter-sign before go-live. The client must approve a mobile Lighthouse score of at least 80 and passing Core Web Vitals on every key page template before the agency authorises the live DNS switch. This contractual quality gate prevents under-optimised sites from launching, protects the agency's reputation against future complaints about slow performance, and gives the client a clear baseline against which they can measure any future regression caused by their own post-launch changes.

WooCommerce redesign

A developer relaunching an e-commerce site on a new theme tests the staging environment and discovers a mobile score of 42 caused by the new theme loading 22 font variants at first paint. The font weight inventory is reviewed, the design team agrees that only two of those weights are actually used in the visual identity, and the remaining 20 are removed from the theme configuration. Mobile Lighthouse jumps to 76 within a single deployment cycle, the launch proceeds on schedule, and the team avoids the much harder post-launch task of explaining to the client why the new site loads more slowly than the old one it replaced.

Startup MVP launch

A startup founder preparing their first marketing website uses the pre-launch speed test to discover that their auto-playing hero video is causing a 7-second LCP on mobile, which would have given visitors a terrible first impression of a product that depends on quick judgement to acquire users. They replace the video with a static hero image paired with a play button that loads the video only on user click, and the page now launches with a 2.1-second LCP instead. The change takes one afternoon, costs nothing, and protects the launch campaign from being undermined by a performance problem that would have been invisible to the founder until users started bouncing.

When to use this guide

Use this as a mandatory step in your website launch checklist, before switching DNS to a new site or deploying a major redesign.

Pro tips

Get better results with these expert suggestions:

1

Test on the actual hosting environment before launch

Development machines and staging servers often differ significantly from production hosting in CPU speed, available memory, caching configuration, and CDN presence. Always run your pre-launch speed test on the actual production hosting environment, or on an identical staging environment that mirrors production in every relevant dimension, to get representative results rather than testing locally on a machine that bears no resemblance to what real visitors will encounter. The test is only as good as the environment it runs in, and a misleading test result is worse than no test at all.

2

Set a performance budget before development begins

A performance budget defines maximum acceptable values for page weight, number of requests, and load time before development starts, making performance a design constraint rather than an afterthought. Establishing these constraints upfront prevents the steady accumulation of just-one-more-plugin and just-one-more-script changes that quietly add seconds to load time over the course of a build. Reasonable budgets for most marketing sites are under 500KB total compressed page weight, under 50 HTTP requests on initial load, and LCP under 2 seconds on desktop with an equivalent target for mobile.

3

Include third-party scripts in your pre-launch test

Many performance issues only appear in production because third-party scripts such as analytics, chat widgets, heatmaps, and advertising tags are conditionally loaded and not present in early development testing. Run at least one pre-launch test with the full complement of production scripts enabled to see how they affect performance under realistic conditions. Third-party scripts commonly add 300ms to 800ms to mobile load time and can shift Core Web Vitals from Good to Poor on their own, so testing without them gives a falsely optimistic picture of launch readiness.

4

Document your baseline before launch for comparison

Record your pre-launch Lighthouse scores, Core Web Vitals values, and page weights as a documented baseline saved alongside the launch sign-off. After launch, re-test the same pages weekly for the first month so any significant performance regression is immediately visible against the documented baseline. This practice allows quick identification and remediation of post-launch changes that degraded performance, and it produces a permanent record that protects the original launch quality against drift over time as new content and plugins accumulate.

5

Test on staging before switching DNS

Run all performance tests on your staging URL before pointing your domain to the new site. This lets you fix issues in a safe environment without affecting live traffic or current search rankings.

6

Include all page templates, not just the homepage

Performance issues often appear on specific page types, product pages with image galleries, long-form content pages, or pages with embedded videos. Test one representative page of each template type.

7

Target a mobile performance score of 80+ before launching

A score under 70 on mobile at launch means you are starting with a Page Experience disadvantage that will take weeks to fix. Delay launch if needed to get mobile performance to at least 80.

FAQ

Frequently asked questions

Minimum reasonable targets before launch are a Lighthouse mobile performance score of 75 or higher, a desktop score of 85 or higher, LCP under 3 seconds on mobile, CLS under 0.1, and INP under 200ms across every major page template. For competitive industries or high-traffic sites where speed materially affects search rankings, aim higher with a mobile score of 85 or higher and LCP under 2.5 seconds. These thresholds ensure the site passes Google's Good threshold for Core Web Vitals from day one and gives the launch CrUX data a strong starting point rather than a performance deficit that takes weeks to recover from after fixes are deployed post-launch.
Run at minimum Google PageSpeed Insights for both mobile and desktop, WebPageTest from a location matching your primary audience's geography, and a manual check of Core Web Vitals using Chrome DevTools on a representative selection of pages. Verify that images are compressed and have correct width and height attributes, JavaScript is deferred or marked async appropriately, caching headers are set on static assets, HTTPS is correctly configured with no mixed-content warnings, and no render-blocking resources exist in the critical path before the LCP element. Run each test at least twice and use median values to suppress single-run variance that could mislead launch decisions.
Pre-launch fixes are simpler and lower-risk than post-launch fixes because no live traffic is disrupted, no SEO signals are established on poor-performing pages, and changes can be made without staged rollouts or canary deployments. Post-launch fixes require careful deployment to avoid impacting live traffic and can introduce regressions of their own that need their own fixes in a compounding cycle. Google also establishes baseline performance data early in a site's life through CrUX, so launching with good performance creates a better starting point for the field data that eventually feeds into Search Console and ranking signals over the first few months.
In rough order of frequency on real projects: unoptimised images without WebP conversion, render-blocking JavaScript that has not been deferred, missing server-side caching causing high TTFB, oversized JavaScript bundles with significant unused code, web fonts loading without font-display: swap and producing invisible text during the load window, missing CDN for geographically distributed audiences, no gzip or Brotli compression on text resources, and excessive third-party scripts loaded synchronously. These eight issues account for the substantial majority of pre-launch performance problems and most can be fixed in a single day of focused work.
Several options work for testing a site that has not yet been pointed at its production DNS. Use browser DevTools Network panel directly on the development URL for basic load analysis. Use ngrok or a similar tunnel service to expose your local server through a public URL that online tools can reach. Deploy to a private staging environment and run PageSpeed Insights against the staging URL, since PSI only needs the URL to be publicly reachable rather than the production domain. Run the Lighthouse CLI locally with a command such as npx lighthouse against the staging URL to produce a report that matches what PSI would generate. Each method has trade-offs but all give you actionable data before launch.
Fix critical issues, defined as anything that would score under 50 on mobile or that fails any Core Web Vital threshold, before launch without exception. Minor optimisations such as pushing a score from 80 to 90 can reasonably be done post-launch as part of a continuous improvement programme. The most important pre-launch fixes to prioritise are enabling page caching, compressing and converting images, deferring non-critical JavaScript, and ensuring TTFB is under 800ms. These are typically configuration and asset changes that do not require code refactors and can be done by anyone with hosting and CMS access in a single afternoon.
Add an explicit note in your test results documentation that the staging environment lacks CDN coverage, then manually add expected CDN improvement estimates based on the known performance difference between origin-only and CDN-served responses. In PageSpeed Insights, look at the Reduce Initial Server Response Time opportunity: if staging TTFB is 800ms and production CDN will serve from edge nodes at 100ms TTFB, adjust your expected production LCP downward by 700ms when interpreting the result. The cleaner option is to deploy CDN rules to staging so the test environment actually matches production behaviour.
Frame the conversation in terms of revenue and ranking risk rather than abstract technical scores, because stakeholders generally do not respond to Lighthouse numbers but do respond to commercial consequences. Quantify the expected impact: cite published research showing that a one-second LCP delay reduces mobile conversion by roughly 7 per cent, calculate what that means for the projected first-year revenue, and present the launch delay cost as a small fraction of the revenue at stake. Bring competitor benchmarks showing where the launch site sits relative to rivals already competing for the same search visibility. The case for a one or two week delay becomes self-evident when framed this way.
A complete pre-launch performance audit on a small marketing site with under twenty pages typically takes one to two hours including testing every key template, documenting baselines, and writing up the fix recommendations. A mid-sized site with twenty to one hundred pages typically takes half a day to a full day depending on how many distinct page templates exist and how many third-party integrations need verifying. Large enterprise sites with many templates and several integrations can take two to three days. Build the audit time into the launch schedule explicitly rather than treating it as a final-day check that gets squeezed when other tasks overrun.

Ready to get started?

Open the full Website Speed Test — free, no account needed, works on any device.

Open Website Speed Test →

Free · No account needed · Works on any device