Research brief: Page Speed as a Moat — why CWV separates the agencies from the freelancers (piece 9 of 15)
Created 2026-05-22
Status: Research material — not a finished article. Compiled May 2026.
TL;DR
- Core Web Vitals are necessary but not sufficient in 2026. Google's official "Good" thresholds remain LCP ≤ 2.5s, INP ≤ 200ms, CLS ≤ 0.1 (developers.google.com, last updated Dec 10, 2025). The widely-repeated SEO-blog claim that March 2026 tightened LCP to 2.0s and "elevated" INP is unconfirmed by Google and contradicted by current documentation — see REFUTED: "March 2026 LCP tightened to 2.0s / INP elevated" — no such Google announcement exists.
- More than half of the mobile web still fails CWV. Web Almanac 2025: 48% mobile / 56% desktop origins pass all three on July 2025 CrUX. LCP is the bottleneck (62% mobile good vs 77% INP, 81% CLS).
- The platform spread is the moat. Duda 83.63%, Shopify 75.22%, Squarespace 67.66%, WordPress 43.44%. Elementor sub-segments perform worse (25-35%). Agencies who deliver durable CWV on WordPress — or skip it for Astro/Next.js/Hugo — are doing structural work that freelancers using stock themes cannot replicate.
Honest caveats
- The "March 2026 core update tightened LCP to 2.0s" claim circulating across SEO blogs traces to a single Digital Applied post citing a Search Central blog entry that does not exist. Google docs as of Dec 10, 2025 still state 2.5s. Search Engine Roundtable confirmed (April 8, 2026) Google's only public communication was a Search Status Dashboard note. Treat as unconfirmed rumour.
- CWV is Chrome-only field data. Safari only added LCP+INP in Safari 26.2 (Dec 2025). For iPhone-heavy demographics, CrUX may not represent your audience accurately.
- Low-traffic SMB sites often show no URL-level CrUX data — Google falls back to origin-level. The moat may be invisible at very small scale.
- Per Dan Taylor (Search Engine Land, Jan 13 2026, n=107,352): CWV is a gate, not a growth lever. Severe failure suppresses AI Overview citation; "Good" → "great" does not lift it. Correlations: LCP r=−0.12 to −0.18, CLS r=−0.05 to −0.09.
Related
- reference 2026 CWV thresholds (confirmed unchanged): LCP ≤ 2.5s, INP ≤ 200ms, CLS ≤ 0.1 at p75 field data
- reference REFUTED: "March 2026 LCP tightened to 2.0s / INP elevated" — no such Google announcement exists
- reference Web Almanac 2025: 48% mobile / 56% desktop origins pass all 3 CWV; LCP is the bottleneck
- reference Web Almanac 2025: median mobile page weight crossed 2,559 KB in July 2025 (+7.8% YoY)
- reference Vodafone A/B test (2021): 31% LCP improvement → 8% more sales, 15% better lead-rate, 11% better cart-rate
- reference Deloitte/Google "Milliseconds Make Millions" (2020): 0.1s improvement → 8.4% retail conversion lift, 10.1% travel
- reference Pinterest PWA rebuild (2017, foundational): 60% faster → +40% conversions, +40% time spent, +44% ad revenue
- reference Walmart (2012, foundational): 100ms improvement = 1% incremental revenue; 1s improvement = 2% conversion increase
- reference BBC: lost 10% of users for every additional second the site took to load
- reference Chrome Aurora team: Google Tag Manager with 18 tags increases Total Blocking Time ~20×
- reference Astro vs Next.js (eastondev benchmark, Dec 2025): ~40% faster, ~90% less JS for static content
- reference Dan Taylor (SE Land, Jan 13 2026, n=107,352): CWV is a gate for AI citation, not a growth lever
- reference Google WebP Compression Study: WebP is 25-34% smaller than JPEG at equivalent SSIM (avg ~30%)
- reference Reference: the 8 most common LCP killers — ranked by observed frequency
- reference Reference: the 8 most common INP killers — ranked by observed frequency
- reference Reference: the 5 most common CLS killers — and the fixes
- reference Reference: 2026 performance budget — SMB marketing site (Candid default targets)
- rule RULE: Performance budgets enforced in CI. Build fails on budget violation. Documented budgets without CI enforcement do not work.
- rule RULE: Defer or delay every third-party script until user interaction, except those required for first paint
- rule RULE: The LCP image always has explicit width/height + fetchpriority="high" + no lazy-loading
- reference Research brief: What makes a marketing site do something (piece on brochure vs platform)
- reference Research brief: Built to Last — why most SMB sites rebuild every 3-4 years (piece 5 of 15)
- reference June 2025 platform CWV pass rates (SEJ): Duda 83.63%, Shopify 75.22%, Wix 70.76%, Squarespace 67.66%, Drupal 59.07%, WordPress 43.44%
Referenced by (4)
- reference Research brief: The Case Against Page Builders (piece 10 of 15) relates-to
- reference CANDID REFERENCE: how the 15-brief foundation roadmap connects — the throughline from strategic frame to editorial layer depends-on
- reference Research brief: Candid Creative 2026 Build-Standards — web stack decision framework for SMB marketing sites & lightweight apps (piece 16) relates-to
- reference Research brief: WordPress + Page Builders vs Modern Custom Stacks — sourced performance comparison (piece 18) depends-on