> ## Documentation Index
> Fetch the complete documentation index at: https://docs.daftplug.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Measuring performance

> How the score is calculated, what the metrics mean, and why it disagrees with PageSpeed Insights.

**Lightify > Overview**

The overview page reports a performance score, your Core Web Vitals over time, and which pages and assets have been optimized.

## Where the number comes from

Lightify injects a small collector into the **cached copy of your homepage** and records what real visitors experienced. Those measurements are averaged into a score.

Two consequences follow from that, and both matter.

<Columns cols={2}>
  <Card title="It reflects reality" icon="users">
    Real devices, real networks, real distances from your server. Not a lab test on a simulated phone.
  </Card>

  <Card title="It needs traffic" icon="hourglass">
    A quiet site takes time to collect enough samples. A new install shows nothing at first, and that is correct rather than broken.
  </Card>
</Columns>

<Note>
  The collector is injected into the cache file, not into the live render, so the first uncached visitor is never measured. The score describes the experience of people who got a cached page, which is what you are trying to optimize.
</Note>

## The metrics

| Metric | Weight | Good        | Poor       | What it measures                                    |
| ------ | ------ | ----------- | ---------- | --------------------------------------------------- |
| INP    | 30%    | Under 200ms | Over 500ms | How quickly the page responds to a tap or click     |
| LCP    | 25%    | Under 2.5s  | Over 4s    | When the largest visible element finishes rendering |
| CLS    | 25%    | Under 0.1   | Over 0.25  | How much content moves around while loading         |
| FCP    | 10%    | Under 1.8s  | Over 3s    | When anything first appears                         |
| TTFB   | 10%    | Under 800ms | Over 1.8s  | How long your server takes to respond               |

The weighting follows Lighthouse, with INP standing in for total blocking time and TTFB for speed index.

### What to do about each

<AccordionGroup>
  <Accordion title="INP is poor">
    Too much JavaScript running on the main thread. [Delay JavaScript execution](/lightify/settings/asset-optimizations#javascript-optimizations) is the direct fix, especially for third-party tags. Deferring helps too.

    This is 30% of your score and the metric most often ignored, because it only shows up when someone actually interacts.
  </Accordion>

  <Accordion title="LCP is poor">
    Your largest above-the-fold element arrives late. Usually an image.

    Turn on [preload critical images](/lightify/settings/media-optimizations#image-optimizations), make sure your hero is not lazy-loaded, and convert it to WebP. If your LCP element is text, the cause is usually a web font blocking render, so try [use system fonts first](/lightify/settings/asset-optimizations#font-optimizations).
  </Accordion>

  <Accordion title="CLS is poor">
    Content moves as the page loads. [Properly size images](/lightify/settings/media-optimizations#image-optimizations) fixes the most common cause. [Prevent layout shifts](/lightify/settings/asset-optimizations#html-optimizations) addresses the rest by not showing anything until it is settled.

    Also look for ads, embeds, and cookie banners injected above existing content, which no plugin setting can fix.
  </Accordion>

  <Accordion title="TTFB is poor">
    Your server is slow to respond. Page caching is the main answer, so first confirm it is [actually active](/lightify/getting-started/installation#confirm-page-caching-is-active).

    If it is active and TTFB is still poor, the cause is upstream of the plugin: shared hosting under load, a distant server, or a slow database on uncached requests. An [object cache](/lightify/settings/cache#object-cache) helps the requests page caching cannot.
  </Accordion>
</AccordionGroup>

## Why this disagrees with PageSpeed Insights

They measure different things, and both are right.

|             | Lightify             | PageSpeed Insights lab score |
| ----------- | -------------------- | ---------------------------- |
| Source      | Your real visitors   | One simulated run            |
| Device      | Whatever they used   | A throttled mid-range phone  |
| Network     | Theirs               | Simulated slow 4G            |
| Page        | Your cached homepage | The URL you enter            |
| Cache state | Always cached        | Often the uncached first hit |

A site can score 95 in Lightify and 60 in a Lighthouse lab test without either being wrong. Lighthouse deliberately simulates a slow device on a slow network; your visitors may be on fast phones near your server.

<Tip>
  Use both. Lightify tells you what your visitors experienced. Lighthouse tells you what a specific page does on a bad device, which is better for finding a specific problem to fix. If they disagree wildly, the lab test is usually showing you the uncached experience.
</Tip>

PageSpeed Insights also shows real-user data from the Chrome UX Report when your site has enough traffic. That is the number closest to Lightify's, and the two should broadly agree.

## Optimized pages and assets

The overview also lists what Lightify has processed: which pages are cached, and which CSS, JavaScript, image, and font files have been optimized.

Use it to confirm work is actually happening. If a template you expected is missing, it is either excluded, never visited since the last purge, or not cacheable. Preloading after a purge fills the gaps.

## Getting alerted

Rather than checking this page, turn on [score alerts](/lightify/settings/notifications) and get an email when the grade changes in either direction.