DIRECT ANSWER
Core Web Vitals describe loading, responsiveness and visual stability. They help you evaluate the experience of using a page. They do not measure whether the offer is clear, whether inquiries arrive or whether customers buy. Review performance alongside the actual customer path.
Read the three measures
Google's Web Vitals documentation identifies these measures and recommended good thresholds:
| Metric | What it measures | Good threshold |
|---|---|---|
| Largest Contentful Paint, LCP | Loading performance | 2.5 seconds or less |
| Interaction to Next Paint, INP | Responsiveness to interactions | 200 milliseconds or less |
| Cumulative Layout Shift, CLS | Visual stability | 0.1 or less |
For assessment, the documented target is the 75th percentile of page loads, separated by mobile and desktop, with all three meeting their thresholds. CLS is a score, not seconds. These are technical experience thresholds, not sales targets.
Check your values against the thresholds
Compare each value with its threshold in the table above: LCP 2.5 seconds or less, INP 200 milliseconds or less, CLS 0.1 or less.
Check what the report actually measured
PageSpeed Insights presents lab and real-world field information when available. Field data reflects eligible real-user experiences. Lab data comes from a controlled test and helps investigate issues. A report may lack sufficient field data for a particular page.
Record whether the displayed field information applies to the exact URL or a broader origin. Do not present site-wide information as if it came exclusively from the page you changed. Record the device category and collection window as well.
A single lab run is not the same thing as the 75th-percentile field assessment. Missing field data is not evidence that the page passed or failed.
Ask for a diagnosis tied to the page
If a metric needs work, ask the builder to identify the affected element or interaction and provide evidence. “The score is low” names a symptom. It does not specify the repair.
Review the pages people actually use: the service explanation, inquiry path and any important tool. A homepage-only test leaves the rest of the customer journey unexamined.
Use a before/after record with comparable test conditions. Keep field results and lab results labeled separately. A technical improvement may be worth making without claiming a measured increase in leads.
Keep design additions accountable
An image, animation, embed or calculator needs a clear purpose. Ask what the visitor gains and test the page after adding it. An interactive element that explains a useful calculation can belong on the page; its presence still needs performance and usability review.
When a third-party tool is involved, understand the distinction between a link, an embed and an integration. They create different implementation responsibilities.
Check the whole experience
After performance changes, confirm that the page remains readable, keyboard-usable and functional. Run the contact-form test and examine whether the service page answers the buying questions. Performance is one part of the technical layer; what it means for a website to work explains how it depends on the functional and aesthetic layers.
Bring the report, URL and device category to a review. Those details support a concrete discussion about what needs work. Request a website review.
Sources and method
Thresholds come from Google’s Web Vitals documentation. Report distinctions come from PageSpeed Insights documentation. The page-review sequence is SIS guidance. No conversion lift or ranking outcome is claimed.
Sources verified October 10, 2026: