Website Speed Audit Checklist for Small Business Owners
Slow pages lose customers. This checklist walks small business owners and decision-makers through a practical, prioritized website speed audit you can run now or hand to a vendor. It focuses on measurable checks, easy-to-understand pass/fail gates, and clear decision criteria so you can decide whether to fix performance in-house or hire Grover Web Design (GWD).
How to use this checklist
- Start with the quick triage (first section) to see whether speed is a likely revenue problem.
- Collect the pre-audit details so a developer or agency can act fast.
- Run the lab and field tests listed under Measurements. Use the checklist items to triage issues by severity.
- Prioritize fixes with the scoring rubric and the qualitative effort guide in the table.
- If fixes touch hosting, server-level configuration, complex JavaScript, or business-critical pages, use the decision criteria to consider hiring a specialist.
Pre-audit info to collect (5 minutes)
Before running tools, gather these facts. They save time and prevent false conclusions.
- Primary goals and conversion pages: Which pages drive calls, sales, or leads (homepage, product pages, booking forms)?
- Traffic & peak times: Typical daily visitors and known traffic spikes (promotions, seasonality).
- Hosting details & access: Host provider, control panel (cPanel, Plesk, managed WP), and SFTP/SSH credentials for the team performing the audit.
- CMS & plugins: WordPress, theme name, and a list of major plugins (eCommerce, page builders, analytics, tag managers).
- Business constraints: Any HIPAA, PCI, or regulatory requirements; maintenance windows; and preferred times for deployments.
Measurements: the tests you should run
Combine field data (real users) and lab tests (repeatable debugging). Recommended free tools:
- Google PageSpeed Insights (field metrics + Lighthouse lab results)
- WebPageTest.org (waterfall, filmstrip, advanced throttling)
- Lighthouse (in Chrome DevTools for local debugging)
- Chrome User Experience Report (CrUX) via PageSpeed Insights for real-user Core Web Vitals
Quick triage (2–10 minutes)
- Open PageSpeed Insights for a high-value page (homepage or top product). Look at Field Data (if available) and the Core Web Vitals metrics: LCP, FID/INP, CLS.
- Run WebPageTest from a nearby test location using mobile throttling (4G) and a desktop run. Export the waterfall and HAR for later review.
- Check server response time (Time to First Byte) in the waterfall. If TTFB > 800ms consistently, suspect hosting or backend bottlenecks.
The checklist: specific checks and pass/fail guidance
1) Core Web Vitals
- Largest Contentful Paint (LCP): Check the LCP element and the time. Gate: LCP should be under ~2.5–3s for best UX; over ~4s is a severe conversion risk. If LCP is slow, identify whether the element is a hero image, large background, or slow-loading block from the server.
- Interaction to Next Paint / INP (or FID historically): Check for long main-thread tasks caused by heavy JavaScript. Gate: INP should be in a responsive range — long tasks > 50–100ms indicate problems.
- Cumulative Layout Shift (CLS): Identify unstable elements (ads, iframes, images without dimensions). Gate: CLS should be low; scores > 0.1 need fixes.
2) First-byte & server health
- TTFB: Measured in WebPageTest waterfall. Gate: Consistently > 800ms suggests moving to a better host, adding PHP workers, or investigating slow backend queries.
- Check server CPU/Memory spikes around slow requests. If the site is on shared hosting and you see throttling, consider managed WordPress hosting or a VPS.
3) Hosting, CDN, and DNS
- Confirm a CDN is configured for static assets (images, JS/CSS). Gate: No CDN for a geographically-distributed audience is a missed opportunity.
- DNS: Use a fast DNS provider; check lookup times in WebPageTest. Slow DNS adds latency for first-time visitors.
4) Images and media
- Image format and size: Ensure modern formats (AVIF/WebP where supported) and appropriately sized images for each breakpoint.
- Responsive images: Use srcset or picture elements so mobile devices receive smaller files.
- Lazy-loading: Enable for offscreen images, but exclude the hero image that contributes to LCP.
- Gate: Images that account for >40% of page weight should be optimized first.
5) JavaScript and CSS
- Audit unused CSS/JS: Use Lighthouse to find unused code. Long main-thread tasks often come from heavy JS bundles.
- Defer or async non-critical JS. Inline critical CSS for the above-the-fold content and defer the rest.
- Third-party scripts: Identify tag managers, ad pixels, chat widgets, and social embeds. Gate: Any external script that blocks main-thread execution on page load should be deferred or loaded conditionally.
6) Caching and cache invalidation
- Verify server and browser caching headers (Cache-Control, Expires). Static assets should have long max-age with proper cache-busting.
- Page caching for dynamic sites: Use page cache for logged-out visitors; exclude personalized pages.
- Confirm proper cache purge workflows (deploy, CMS update).
7) Database and backend
- Inspect slow queries or long database response times. Often seen on search, faceted navigation, or bloated admin dashboards.
- Check for excessive REST API or AJAX calls on initial page load.
8) Security, updates, and overhead
- Up-to-date CMS, theme, and plugin versions reduce unexpected performance regressions and security-related slowdowns.
- Remove unused plugins and lightweight alternatives for heavy functionality (e.g., replace bulky page builders if necessary).
9) Mobile UX & practical tradeoffs
- Test perceived performance: speed isn’t only raw time. Use skeleton loaders, preconnect, and prioritized content to make pages feel faster.
- Gate: If mobile clicks or conversions lag desktop significantly, prioritize mobile LCP and interaction fixes first.
Quick priority matrix (severity vs effort)
| Issue | Severity (Low/Med/High) | Typical Effort | Why fix first? |
|---|---|---|---|
| Large unoptimized hero images | High | Low | Direct LCP improvement on top-converting pages |
| Excessive JS on load (third-party widgets) | High | Medium | Improves interactivity and long tasks |
| Slow TTFB from hosting | High | Medium–High | Improves all requests and user experience |
| Missing caching headers / no CDN | Medium | Low–Medium | Reduces bandwidth and improves repeat view speed |
| Minor layout shifts (ads, embeds) | Medium | Low | Improves perceived quality and CLS |
Scoring rubric and prioritization plan
Use this simple scorecard to prioritize work across pages. Score each high-value page (homepage, product, checkout) on a 0–5 scale for each dimension (0 worst, 5 best): LCP, INP, CLS, TTFB, and Page Weight. Add the scores; lower totals mean higher priority.
Example action plan by total score:
- 0–10: Critical — immediate intervention, consider temporary mitigations (serve reduced hero images, disable heavy widgets) and schedule full remediation.
- 11–15: High — prioritize fixes in the next sprint; combine image and JS optimizations first.
- 16–20: Medium — gradual improvements and monitoring; set up continuous performance budget checks.
When to DIY vs hire GWD
Choose DIY when:
- Your site uses a standard managed WordPress setup and you can deploy safely (backups, staging environment).
- The issues are clearly isolated and low-effort (image compression, enabling a CDN, tweaking cache headers).
- You have internal technical resources comfortable with staging, SFTP, and plugin configuration.
Hire GWD when:
- Problems involve hosting architecture, server tuning, or persistent TTFB issues that require migrations or backend profiling.
- Performance fixes intersect with custom functionality (commerce flows, user portals) where regressions risk revenue or data integrity.
- You want a measured performance budget, continuous monitoring, and a documented remediation plan we can own as part of ongoing maintenance.
If you engage GWD, we typically start with a focused audit of high-value pages, deliver a prioritized remediation list, and then implement fixes in a controlled staging workflow. Learn more about our services here: Grover Web Design services. For custom development or complex integrations, see Custom Web Development, and for broader visibility work, see our SEO Services.
Practical next steps (30/90/180 day roadmap)
- 30 days: Quick triage fixes — optimize top hero images, enable a CDN, add basic caching, and defer non-critical scripts.
- 90 days: Address larger structural issues — move to a managed host if needed, refactor heavy JavaScript, and fix database inefficiencies.
- 180 days: Establish monitoring, a performance budget, and incorporate performance checks into release QA to prevent regressions.
Related reading and internal links
Pair speed work with general SEO and launch best practices. Start with our technical SEO guide: Website SEO Fundamentals: From Clean Code to Fast Loading Times. If you are launching a new site, use our Web Design & Custom Development service page as a checklist for technical requirements and handoff expectations.
FAQ
How do I know if speed is costing my business money?
Look for drops in conversion rate or increased bounce rate on pages that show poor LCP or INP. Use PageSpeed Insights field data and analytics to correlate performance drops with business metrics. If high-traffic pages have LCP > 3s or severe interaction delays, speed likely affects conversions.
Will installing a performance plugin solve everything?
Plugins (image optimizers, cache plugins, CDNs) often help low- to medium-complexity sites. But they can also conflict, especially with custom themes or complex JavaScript. Use plugins as part of a staged approach and test on staging before production.
How often should I run a speed audit?
Quarterly for stable sites and after any major release. After migrations or major content changes, run a quick audit and check the staging environment before pushing live.
Do I need to migrate hosting to speed up my site?
Not always. Many speed issues are fixable with caching, a CDN, image optimization, and JS deferral. However, if TTFB is consistently high and the host restricts resources, a hosting move may be the most reliable solution.
What are realistic expectations after fixes?
Expect meaningful improvements to LCP and interaction times when you prioritize high-impact items: hero images, server response, and third-party script management. Full optimization is iterative: fix the biggest wins first and monitor for regressions.
We can’t solve your problem if you don’t tell us about it!
