Performance work is more reliable when it starts with a baseline rather than a list of plugins. Core Web Vitals describe user experience through loading, responsiveness, and visual stability. A fast lab test is useful for diagnosis, but field data helps show how real visitors experience a site across devices and networks.
1. Measure representative pages
Choose important page types: the home page, a service or product page, a search or archive page, and checkout if the site sells online. Review field data where available, and use a lab tool to reproduce bottlenecks. Record the test URL, device profile, network conditions, and date so later comparisons are meaningful.
2. Find the largest bottlenecks first
For a slow Largest Contentful Paint (LCP), identify the element being measured and inspect its image size, priority, server response, and render-blocking resources. For Interaction to Next Paint (INP), profile long JavaScript tasks and third-party scripts. For Cumulative Layout Shift (CLS), reserve space for images, embeds, and late-loading UI.
3. Make targeted changes
- Serve appropriately sized, modern image formats and avoid lazy-loading the primary above-the-fold image.
- Cache pages and assets where the content and logged-in state allow it; test cart, account, and personalized pages separately.
- Reduce unused JavaScript and defer non-essential third-party widgets.
- Load only the font weights and styles the design uses, and avoid shifting text while fonts load.
- Review hosting response time and database work before adding more optimization plugins.
4. Validate the result
Change one group of variables at a time, rerun the same tests, and check key user journeys for regressions. Lab scores can vary between runs; field metrics take time to reflect changes. Report the exact pages, measurement method, and date instead of promising a universal score.
For a broader site review, see our WordPress speed optimization service or performance and security work.