The original title is "How We Turned a High-Traffic News Site's Core Web Vitals from Red to Green"

Short summary

A detailed case study of fixing all three Core Web Vitals (LCP, INP, CLS) on finance.kapital.kz, a high-traffic Kazakh news portal built on Next.js. Key fixes included optimizing hero images for LCP, deferring Google Ads scripts and removing heavy libraries from the main bundle for INP, and reserving ad slot space plus correcting infinite-scroll height estimates for CLS. The team used BigQuery real-user monitoring field data rather than synthetic Lighthouse runs to target the right problems.

• •LCP fixed by optimizing hero image loading on article pages
• •INP improved by deferring ad scripts and lazy-loading heavy libraries out of the main bundle
• •CLS eliminated by reserving ad slot dimensions and fixing infinite-scroll height estimation
• •Field data via BigQuery used instead of lab-only Lighthouse scores to identify real user issues

Generated with AI, which can make mistakes.