Speed starts with the cause.
One test result is not a diagnosis. Speeding up a website needs repeatable measurements, device context and a network and code review.
Measure before change
Key pages, devices, conditions and repeated results are recorded to separate persistent problems from temporary variance.
Core Web Vitals in brief
Google tracks three metrics on the devices of real visitors. LCP measures when the main content of the first screen appears and is good up to 2.5 seconds. INP measures how quickly the page responds to a tap, click or key press and is good up to 200 milliseconds. CLS measures how much the layout shifts while loading and is good up to 0.1.
A lab test helps find the cause, but when field data from real devices exists, it shows how the website actually performs.
Largest cost first
Heavy images, blocking fonts, unnecessary scripts and a slow initial response are handled by impact rather than fashion.
What gets checked
The review runs from the first request to the last movement on screen:
- the largest element on the first screen and how late the browser discovers it
- CSS and JavaScript that block the first render
- image size, format and dimensions (AVIF, WebP, width and height)
- fonts: subsets, loading strategy and a fallback that does not move text
- third-party scripts, caching, compression and server response time
Regression after repair
Content, layout, features and the same measurement path are checked again. Faster must not mean broken.
The full workflow, from current-state snapshot to launch, is described in the process overview.