Three years after Google formally incorporated Core Web Vitals into its ranking signals, the SEO industry has settled into two camps: those who treat CWV as a top-tier ranking priority and those who dismiss it as a minor signal not worth engineering resources. The truth, as usual, is more nuanced than either position.
What the Research Actually Shows
The direct ranking impact of Core Web Vitals is real but modest in most verticals. Studies consistently show that pages with “Good” CWV scores have a measurable advantage over pages with “Poor” scores — but the effect size is typically smaller than topical relevance, content quality, and backlink authority. In competitive niches, a 2–3% ranking boost from improved CWV is often less valuable than a single good editorial link.
The indirect effects are harder to measure but potentially larger. A page that loads in 1.2 seconds versus 4.5 seconds will have lower bounce rates, longer dwell times, and higher conversion rates — all of which are correlated with ranking improvement, even if the causal mechanism is debated. Better user experience feeds into better engagement metrics, and Google uses those engagement signals.
The Three Metrics in Practice
LCP (Largest Contentful Paint) is the metric that moves rankings most directly and is also the most fixable. The main causes of poor LCP are unoptimised hero images (not using WebP or AVIF, not preloaded), render-blocking resources, and slow server response times. For most sites, addressing those three issues gets you from “Poor” to “Good” without touching complex JavaScript.
CLS (Cumulative Layout Shift) has been the source of most engineering anguish since 2021. The practical fix is simpler than it seems: set explicit dimensions on images and embeds, don’t inject content above existing content, and be careful with web fonts causing layout shifts during load. CLS under 0.1 is achievable for most sites without significant rework.
INP (Interaction to Next Paint) replaced FID in March 2024 and is the metric where JavaScript-heavy sites struggle most. INP measures the full latency of user interactions — not just the first one. Sites running large amounts of third-party JavaScript (ad networks, chat widgets, analytics stacks) typically have INP issues that are difficult to resolve without removing or deferring those scripts. This is often a product or commercial negotiation more than an engineering problem.
Where to Spend Your Time
If your site is already in the “Good” range on all three metrics, CWV optimisation is not where to focus. The marginal ranking gain from moving from “Good 1.8s LCP” to “Exceptional 0.9s LCP” is negligible.
If your site has “Needs Improvement” or “Poor” scores, especially on LCP, fixing that is worth prioritising — not because of the direct ranking signal alone, but because a slow site loses conversions and engagement regardless of its ranking position.
The most overlooked CWV issue is mobile performance. Desktop scores and mobile scores diverge significantly for most sites, and Google uses mobile-first indexing. Check your CrUX data in Search Console filtered to mobile specifically before drawing conclusions from desktop lab tests.
The Realistic Priority Stack
For most sites, the CWV priority order should be: fix catastrophic failures (Poor scores) immediately, get to “Good” on mobile LCP, then shift resources elsewhere. CWV is infrastructure, not a growth lever. The sites that treat it as a primary SEO strategy are optimising in the wrong direction.
The sites that have compounding organic growth are building topical authority, earning editorial links, and publishing content that genuinely answers what their audience is searching for. Fast load times support that work. They don’t replace it.