The causes, not just the score · free · no email

Core Web Vitals checker

Core Web Vitals measure what real users feel on your page. Google's definition: "Core Web Vitals are the subset of Web Vitals that apply to all web pages, should be measured by all site owners, and will be surfaced across all Google tools." Most tools show the three scores and stop. We read your HTML source and flag the 10 patterns that most often cause the bad scores, so you know what to change.

These are lab signals, what your code says, not the real-user field data from Chrome. We are honest about the difference, and we show the field data too when it exists for your site.

Free Visibility Scan: whole site, up to 150 pages, no email. Full Report is a one-time $10.

What Core Web Vitals measure

Three numbers, measured on real Chrome users.

LCP

Largest Contentful Paint

How fast the biggest image or text block loads. Good: under 2.5 seconds at the 75th percentile.

INP

Interaction to Next Paint

How fast the page responds when someone clicks or taps. Good: under 200 milliseconds.

CLS

Cumulative Layout Shift

How much the page jumps around as it loads. Good: score under 0.1.

Google's ranking systems reward pages that provide a good page experience, and these metrics are how that experience is measured. Fail them at the 75th percentile and your site is treated as slow. The deeper fix guide: fixing page speed and LCP.

What we check

The 10 HTML-source patterns behind most CWV failures.

  • Lazy-load strategy (LCP): we identify the most likely LCP image on the page. If it is set to loading="lazy", the browser will not fetch it until it is in the viewport, which delays the metric. We flag this.
  • Image formats (LCP): WebP and AVIF cut image file size 25-50% versus JPEG at the same quality. We check whether your content images use modern formats.
  • Responsive images (CLS): images without declared width and height cause layout shift when they finally load. We flag missing dimensions.
  • Render-blocking scripts (LCP): synchronous script tags in the head pause the render until they download and run. We count them and flag the ones that should carry defer or async.
  • DOM size (INP): a page with more than 1,400 DOM nodes is harder for the browser to paint and update. We flag oversized DOMs.
  • Font delivery (CLS): fonts without font-display: swap cause invisible text, then a flash when the font arrives. That flash is layout shift. We check your @font-face declarations.
  • Third-party scripts (LCP / INP): we count scripts loaded from domains other than your own. Each one is a network request you do not control.
  • Heavy embeds (LCP): YouTube iframes, maps and similar embeds load hundreds of kilobytes by default. We flag them and note that a facade pattern loads them only on click.
  • Inline data URIs (LCP): large base64 images embedded in the HTML inflate the document and delay the first parse.
  • Preconnect hints: whether your page warms connections to critical third-party origins before the browser needs them.

When real Chrome field data (CrUX) is available for your site, the report shows LCP, INP and CLS at the 75th percentile for your actual users. If your site does not have enough traffic for CrUX to have data, we say so rather than guess. Every check runs inside the full 164-check audit; the umbrella view is the AI search optimization hub.

Questions

Straight answers.

Do you run Lighthouse?

No. We read your HTML source and flag what we find there. When the PageSpeed feed is available, we surface Google PageSpeed Insights results, which do run Lighthouse. We are clear about which numbers come from the API and which from our source scan.

Why does my site have a bad CWV score even though Lighthouse says it is okay?

Lighthouse runs a synthetic test in a lab. Real Chrome field data (CrUX) reflects actual devices and network conditions. They can disagree. Field data is the one Google uses for ranking.

What if my site is not in CrUX?

Most small sites are not. CrUX only includes origins with enough Chrome traffic. If your site is not in the dataset, we tell you that plainly. We do not invent a field score.

Can you fix the issues for me?

The $10 Full Report packages every finding, including the lazy-load attribute change, the missing dimension attributes and the render-blocking script changes, into a prioritized action plan and an AI-ready fix file your own tools can run with.

Does this replace a full page speed test?

No. For a complete performance profile use PageSpeed Insights, Lighthouse or WebPageTest. We catch the most common HTML-level CWV causes fast, so you know where to start.

Find the page choices pushing your vitals into the red.

Free scan of your whole site, up to 150 pages. The causes named, not just the score. No email.

164 graded checks · official source citations on every finding · no made-up scores