Home ci 16. speed-insights: metrics

16. speed-insights: metrics

Last updated on Aug 05, 2025

Understanding Website Performance Metrics: A Comprehensive Overview

  1. Key Performance Metrics:

    • First Contentful Paint (FCP): Measures the time taken for the first content to be rendered on a page, crucial for initial load speed.
    • First Input Delay (FID): Assesses the delay before a page responds to user interactions, such as clicks.

  2. Percentile Calculations:

    • Metrics like P75, P90, and P95 indicate that 75%, 90%, and 95% of users experience faster load times than the specified value. For example, a P75 score of 1 second means 75% of users see FCP in under a second.

  3. Color-Coding System:

    • Scores are color-coded as Poor (0-49), Needs Improvement (50-89), and Good (90-100). Aiming for 'Good' scores is recommended for optimal user experience, though higher scores become increasingly challenging to achieve.

  4. Real Experience Score (RES) vs. Virtual Experience Score (VES):

    • RES: Based on real user data post-deployment, providing insights after the fact.
    • VES: Predictive, using metrics like Total Blocking Time (TBT), allowing anticipation of performance changes pre-deployment.

  5. Device Considerations:

    • Device type affects scores; Checkly uses "Desktop" for VES, potentially underweighting mobile users. Weighting strategies may be necessary for comprehensive analysis.

  6. Effort vs. Outcome:

    • Improving from 99 to 100 on RES is harder than moving from 90 to 94 due to diminishing returns as scores approach perfection.

  7. Impact on User Experience and Search Rankings:

    • Higher scores enhance user experience but may not directly boost search rankings unless they move into higher color categories, like green from orange.

  8. Metric Differences:

    • VES uses TBT instead of FID, offering a broader view of user experience by considering total interaction delays.

  9. Data Updates and Analysis:

    • RES is updated with recent user data, while VES relies on automated checks for predictions.

  10. Targeted Optimization:

    • By identifying user drop-off points and slow responses, developers can focus on specific areas to enhance overall performance.