A performance audit is the practice of methodically measuring how fast your site actually loads and identifying what is making it slow. Most site owners have an impression of their site’s speed (it feels fast, or it feels slow) without specific data. The audit replaces impressions with measurements.
This piece covers what a performance audit includes, how to do one, and how to act on the findings to actually make your site faster.
Why Performance Audits Matter
The case for performance work is well-established.
SEO Impact
Google uses page speed as a ranking factor. Sites that load faster rank better, all else equal. Core Web Vitals specifically measure user experience metrics tied to performance.
Slow sites get pushed down in rankings. The effect compounds over time.
Conversion Impact
Every extra second of load time costs conversions. Studies consistently show that conversion rates drop measurably as load times increase. For e-commerce, this translates directly to revenue.
User Experience
Visitors form impressions of sites in the first few seconds. Slow sites feel broken. Visitors leave for faster competitors.
Mobile Considerations
Mobile users have slower connections and less patience than desktop users. Performance matters more on mobile, and mobile traffic is the majority for most sites now.
Bandwidth & Cost
Slow sites with large pages cost more to serve. Bandwidth bills are higher. Performance optimization can reduce hosting costs alongside improving user experience.
What to Measure
Several metrics matter for performance.
Page Load Time
The total time from request to fully loaded page. Includes everything. Useful as a general measure but not always actionable.
Time to First Byte (TTFB)
How long the server takes to respond to a request. Earlier piece in this series covered TTFB in detail.
Good TTFB is under 200ms. Acceptable up to 600ms. Slower than that needs work.
First Contentful Paint (FCP)
When the first content appears on screen. Good FCP is under 1.8 seconds.
Largest Contentful Paint (LCP)
When the main content has appeared. This is one of the Core Web Vitals. Good LCP is under 2.5 seconds.
Cumulative Layout Shift (CLS)
How much the layout shifts during loading. Visual stability. Good CLS is under 0.1.
First Input Delay (FID) / Interaction to Next Paint (INP)
How quickly the page responds to user input. The metric is being updated. Good INP is under 200ms.
Time to Interactive (TTI)
When the page is fully usable.
Total Page Weight
How many bytes the page is. Smaller pages load faster. Pages over 1.5 MB are heavy.
Number of Requests
How many separate files the browser has to download. Fewer requests is better. Pages with over 100 requests are usually heavy.
Tools for Performance Audits
Several tools provide performance measurements.
Google PageSpeed Insights
Free tool that tests any URL. Provides Core Web Vitals data from real users (if available) plus lab-test metrics. Tied directly to Google’s ranking signals.
The recommendations are actionable and specific.
GTmetrix
More detailed performance reports with waterfall charts showing how every asset loads. Useful for identifying specific slow resources.
Free for occasional use. Paid for ongoing monitoring.
WebPageTest
Most detailed free tool. Tests from multiple global locations on various connection speeds. Provides detailed metrics and screenshots of the page load process.
The depth is overwhelming for new users but excellent for serious performance work.
Google Lighthouse
Built into Chrome DevTools. Audits performance plus accessibility, best practices, and SEO. Good for ongoing measurement during development.
Pingdom Speed Test
Simple interface, single global test location, useful for quick checks.
Chrome DevTools Performance Tab
For technical users, the Performance tab profiles actual page loads in detail. Shows exactly what is happening during load and where time goes.
How to Conduct a Performance Audit
The process should be methodical.
Test Multiple Pages
Do not just test the home page. Test high-traffic pages, conversion pages, and pages that might be particularly slow (heavy content pages, search results, dynamic listings).
Different pages have different performance characteristics. Auditing only one page misses problems elsewhere.
Test Multiple Times
Single tests can be misleading due to caching, network variation, or temporary issues. Run each test multiple times and look at consistent results.
Test on Multiple Devices
Mobile performance is often dramatically different from desktop. Test both. Ideally test on actual mobile devices, not just emulators.
Test from Multiple Locations
For sites with global audiences, test from different regions. Performance can vary significantly by visitor location.
Compare to Benchmarks
Numbers without context are not as useful. Compare your numbers to industry benchmarks. Is 2 seconds good for your situation? Bad? Depends on the industry.
Document Findings
For each issue identified, document:
What the issue is.
Where it appears.
How much it affects performance.
What the recommended fix is.
Estimated effort to implement.
The documentation makes the audit actionable.
Common Performance Issues
A few patterns of issues come up regularly.
Slow Server Response
The server takes too long to generate pages. Causes include slow hosting, database issues, lack of caching, heavy backend code.
Earlier piece in this series covered TTFB optimization in detail.
Large Images
Images are usually the biggest assets on pages. Unoptimized images can add megabytes to page weight.
Image optimization (covered in an earlier piece) addresses this.
Render-Blocking Resources
CSS and JavaScript that load before the page can display block rendering. Even small render-blocking files can delay the first paint significantly.
Inline critical CSS. Defer non-critical JavaScript. Load fonts asynchronously.
Too Many HTTP Requests
Each separate file requested takes time. Pages with hundreds of requests load slowly even if each individual file is small.
Combine files where possible. Reduce dependencies. Lazy-load below-the-fold content.
Heavy Third-Party Scripts
Analytics, tracking, chat widgets, ad networks. Each one adds weight and processing time.
Audit what is loaded. Remove what is not needed. Defer what can be deferred.
Lack of Caching
Without caching, the server does all the work on every request. With caching, results are reused. The difference is dramatic.
Server-side caching, browser caching, CDN caching. Each layer helps.
Database Performance
For dynamic sites, slow database queries can be the bottleneck. Earlier piece in this series covered database optimization.
No CDN
Sites without CDNs serve all files from one location. Visitors far from that location have slow loads. A CDN distributes content globally.
Heavy Plugins/Themes
For WordPress, some plugins and themes are heavier than they need to be. Audit what is installed. Replace heavy plugins with lighter alternatives where possible.
Unoptimized Code
Custom code that was not written with performance in mind can slow things down. Code review and optimization help.
Prioritizing Performance Fixes
Performance audits often reveal many issues. Prioritization matters.
Impact First
Some fixes produce dramatically larger improvements than others. Fixing slow hosting can cut load times in half. Optimizing one image might save 100 milliseconds. Focus on the high-impact fixes first.
Ease of Implementation
Some fixes are easy. Enable a CDN. Install a caching plugin. Compress images. Others are involved. Rebuild a theme. Migrate to better hosting.
Start with easy high-impact fixes.
Match Resources
Some fixes require technical expertise you may not have. If you cannot do them yourself, you need to hire help or skip them. Match the work to your resources.
Quick Wins vs Strategic Work
Quick wins boost morale and show progress. Strategic work produces larger improvements but takes longer.
A reasonable approach is to do quick wins first, then commit to strategic work.
Optimizing After the Audit
Implementation matters as much as auditing.
Address the Biggest Issues First
The findings should be in priority order. Work through them starting with the biggest impact.
Measure After Each Change
After implementing a fix, re-measure. Did performance actually improve? By how much?
Without measurement, you do not know if your work is producing results.
Watch for Regressions
After fixing performance, watch for regressions. New code, new images, new plugins can all reintroduce performance issues. Ongoing monitoring catches this.
Set Up Continuous Monitoring
Tools like SpeedCurve, Pingdom, or Calibre monitor performance continuously. They alert you when performance degrades.
This catches issues before they affect users significantly.
Performance Budget
For ongoing performance work, a performance budget helps.
What It Is
A performance budget is a set of constraints you commit to. Maximum page weight. Maximum load time. Maximum number of requests.
New code or features that would push the site beyond the budget get flagged.
How to Set It
Look at your current performance and pick targets. Slightly tighter than current is a good start. Tighten over time as you make improvements.
How to Enforce It
For development teams, automated tools can check performance budgets on every code change. For smaller operations, regular manual checks catch issues.
Common Performance Audit Mistakes
People stumble in predictable ways.
Auditing Once
A one-time audit produces value but the value depreciates. Sites change. New issues appear. Regular audits maintain performance.
Focusing on Wrong Metrics
Some metrics matter more than others. PageSpeed scores are useful but Core Web Vitals are what affects rankings. Real user experience metrics matter more than lab metrics.
Testing Only the Home Page
Different pages have different performance characteristics. Audit multiple page types.
Ignoring Mobile
Mobile performance is often the biggest opportunity. Mobile-first testing matters.
Optimizing Without Measuring
Making changes without measuring the impact wastes effort. Always measure before and after.
Premature Optimization
Some optimization is unnecessary. Optimizing things that are already fast wastes time. Focus on actual problems.
Chasing PageSpeed Score
A high PageSpeed score is satisfying but not always the right goal. Real user experience metrics matter more.
Closing Thoughts on Speed Reviews
Performance audits identify what is slowing your site down and what to do about it. The work is methodical and the findings are actionable.
For most sites, the right approach is a thorough initial audit followed by ongoing monitoring. The initial audit reveals where performance currently stands. The monitoring catches regressions before they affect users significantly.
The tools have matured to the point where basic performance auditing is accessible to non-technical users. PageSpeed Insights and GTmetrix produce understandable results. The recommendations are specific.
For more involved performance work, technical expertise matters. Database optimization, code-level improvements, hosting migrations all benefit from knowledgeable execution.
Sites that take performance seriously perform better in every measurable way. Search rankings, conversion rates, user satisfaction, mobile experience. The investment in performance work pays back across multiple dimensions.
If your site has never been performance audited, this is one of the higher-leverage activities available. The audit takes hours. The improvements take weeks. The benefits compound for years. Sites that audit and act on findings consistently outperform sites that ignore performance until problems become urgent. The choice between proactive performance work and reactive firefighting is not really a choice. Make performance audits part of your routine, and the speed advantages compound over time.
