PERFORMANCE & CORE WEB VITALS GUIDES
Core Web Vitals, LCP, CLS, INP. Real numbers from real builds, not benchmarks. What actually moves the needle for site speed in 2026.
LCP, INP, and CLS: how to fix each Core Web Vital
Fix LCP, INP, and CLS as separate jobs. Concrete levers, common mistakes, and the order that clears field data.
The Core Web Vitals checklist I actually run
Twelve concrete Core Web Vitals checks for launch. Field data first, then LCP, INP, CLS, cache, and third parties.
Why is your website slow? The real causes, and how to fix each
Most slow sites share six causes. Measure field data first, separate server from front end, then fix the Largest Contentful Paint element before touching anything else.
Core Web Vitals and performance in 2026: a working guide
Performance as a structural posture in 2026. The four metrics that rank, the image pipeline that fixes most issues, and the CWV checklist I actually run.
Core Web Vitals for WordPress: a 2026 field guide
WordPress can pass Core Web Vitals, but the defaults work against you. Here are the WordPress-specific causes of bad scores and the fixes that move the needle.
CSP for static sites: the directives nobody documents
The CSP directives that actually matter for static sites in 2026. The FAL + Supabase + Pagefind playbook from real production incidents.
The image pipeline that hits Lighthouse 100 by default
The FAL plus Sharp image pipeline I run on every Seahawk site to hit Lighthouse 100 by default. With concrete code and file-size numbers.
From the blog
- WebP vs JPEG in 2026: Pick the Right Image Format
- Vercel Bandwidth Bills: How to Stop the Surprise Charges
- How bot farms spiked my Vercel bill, and the firewall fix that took 50 days to find
- Best CDN in 2026: reach, cost, or edge control
- Is Astro good for SEO? What running it at scale taught me
- Core Web Vitals on Large Sites: How to Hit Sub-1.5s LCP
Common questions
- Why is my website slow?
- Most slow sites trace to a few causes: unoptimised images, render-blocking JavaScript and CSS, no caching, a slow server response (TTFB), and too many third-party scripts. Fix the largest contentful element first, then cut unused JavaScript. Measure with PageSpeed Insights field data before and after each change.
- What counts as good Core Web Vitals in 2026?
- Aim for LCP under 2.5 seconds, INP under 200 milliseconds, and CLS under 0.1, measured at the 75th percentile of real users. Google reads field data from the Chrome User Experience Report, not your lab score, so a fast Lighthouse run is not enough on its own.
- How do I pass Core Web Vitals?
- Pass all three thresholds at the 75th percentile of field data. In practice: serve a fast, correctly sized LCP image, defer non-critical JavaScript to lower INP, and reserve space for images and embeds to keep CLS low. Re-check field data after about 28 days, since it lags real changes.
- Does Core Web Vitals affect SEO rankings?
- Yes, but as a tiebreaker, not a primary factor. Core Web Vitals is part of Google page experience signals. Content relevance still decides most rankings, but between two similar pages the faster one wins, and a failing score can hold back an otherwise strong page.
- Is a good Lighthouse score enough?
- No. Lighthouse is a lab test on one simulated device; Google ranks on field data from real visitors. A site can score 100 in Lighthouse and still fail Core Web Vitals if real users on slower devices and networks see worse numbers. Always confirm with field data.