Google updated its Core Web Vitals assessment to use a composite scoring model instead of evaluating LCP, INP and CLS independently, changing how site performance translates to search ranking signals.
The old system was binary. Pass everything or pass nothing. The new system rewards overall performance, which is how users actually experience a page.
Google has updated how it evaluates Core Web Vitals, moving from independent pass/fail assessments of each metric to a single composite score. The change affects how Largest Contentful Paint, Interaction to Next Paint and Cumulative Layout Shift combine to form the page experience ranking signal.
Previously, Google assessed each Core Web Vital independently. A page needed to pass all three metrics at the 75th percentile of field data to earn a "good" Core Web Vitals assessment. Under the new model, the three metrics are weighted and combined into a composite score, meaning a strong performance in one area can partially compensate for weakness in another.
Core Web Vitals metrics now combined into a single composite assessment
Google has not published the exact weighting formula, but early analysis from web performance researchers suggests INP carries the heaviest weight, followed by LCP and then CLS. This makes intuitive sense. Interactivity (INP) directly affects whether a user can engage with your page. Load speed (LCP) affects first impressions. Layout stability (CLS) is annoying but rarely causes users to leave.
The practical impact is nuanced. Sites that previously failed Core Web Vitals because of one borderline metric (say, LCP at 2.6 seconds against a 2.5-second threshold) may now pass under the composite model if their INP and CLS scores are strong. Conversely, sites that scraped through with three barely-passing scores may find the composite less forgiving if one metric is significantly worse than the others.
Why it matters
Core Web Vitals is a confirmed Google ranking factor, though its weight relative to content quality and backlinks remains modest. The composite model does not change how important site speed is for rankings. It changes the optimisation calculus.
For Australian businesses, the practical implication is that performance optimisation becomes more strategic and less checklist-driven. Instead of trying to pass three independent thresholds, you can now make informed tradeoffs. A complex interactive application might accept slightly slower LCP in exchange for excellent INP. A content-heavy blog might prioritise LCP and CLS over INP.
This also affects how you interpret PageSpeed Insights and Search Console reports. The old red/amber/green per-metric view will be supplemented by an overall assessment that better reflects actual user experience.
What to do about it
The composite model is a better reflection of how people experience websites. Optimise for users, not for individual metrics.