Core Web Vitals still central
Experts are doubling down: Core Web Vitals remain critical for SEO and UX—advice this week pushed image compression to WebP/AVIF, code splitting and caching as priority fixes andrewroche99. AI workflows for auto‑optimizing server times and images are being discussed as practical paths to improve vitals at scale SaadGPT.
Performance triage this week followed a concrete sequence: fix server response time (TTFB) first, then address LCP, INP and CLS, per the step‑by‑step workflow published on corewebvitals.io corewebvitals.io. Tests and vendor reports show next‑gen codecs deliver big wins — guides cite AVIF cutting image bytes by roughly 50–80% in broad tests [WebPict] webpict.com while Cloudinary case examples reported reductions of about 37–50% after AVIF adoption. sia.codes Edge and CDN services are already automating format choice and resizing in production: Cloudinary’s image pipeline auto‑serves WebP/AVIF next.cloudinary.dev, Cloudflare’s Polish and Images products do on‑the‑fly compression and format switching developers.cloudflare.com, ImageKit offers real‑time automatic conversions imagekit.io and PixelFiddler advertises request‑aware smart optimization at the edge. pixel-fiddler.com Front‑end fixes called out as high‑impact this week were code‑splitting to shrink initial JS payloads and aggressive CDN/browser caching to cut TTFB and LCP; web.dev lists code‑splitting, lazy loading and caching among the top Core Web Vitals levers. web.dev Practical tooling examples named in the thread include ShortPixel Adaptive Images for automatic WebP/AVIF delivery in WordPress wordpress.org, DropAvif’s claim of up to ~85% image size reduction on some assets dropavif.com, and community projects like LightPixel that auto‑convert to WebP/AVIF with reported 25–35% savings on average. github.com Conversation about AI centered on two threads: automated audit-and-fix workflows (AI agents that ingest performance traces and propose or deploy fixes) are being prototyped in industry writeups and platform features corewebvitals.io, while monitoring vendors now sell “AI‑powered” PageSpeed recommendations and scoring to turn diagnostics into actionable change lists. indexly.ai Practical target numbers surfaced repeatedly: LCP under 2.5s and INP under 200ms remain the passing thresholds referenced in current guidance corewebvitals.io, and several posts recommended aiming for TTFB well below 200ms to both improve Core Web Vitals and meet emerging AI‑crawler expectations. amicited.com
