Insights:What Are Core Web Vitals? Explained Visually
Core Web Vitals (CWV) are a set of metrics Google uses to measure page experience. They focus on just three things:
- LCP: how quickly the main content loads;
- INP: how promptly the page responds after user interaction;
- CLS: whether content shifts unexpectedly during page loading.
All three metrics have clear thresholds, and they are based on real-user data rather than scores from a one-off Lighthouse lab test.
First, remember the passing threshold and methodology
- LCP ≤ 2.5 seconds;
- INP ≤ 200 milliseconds;
- CLS ≤ 0.1.
The methodology matters just as much: take the 75th percentile of real-user data over a rolling 28-day window. This means at least 75% of visits must meet the level—it is not about the average being sufficient. Averages can be skewed by a handful of extremely fast or slow visits.
LCP: When does the hero element appear above the fold
LCP measures the time when the largest content element within the viewport completes rendering. It is typically a hero image, a main heading, or a large block of text.
LCP can be broken down into four stages:
- The server returns the first byte (TTFB);
- The browser discovers and begins loading resources;
- The resource download completes;
- The element finishes rendering.
So optimizing LCP is not just about compressing images. If the server takes 2 seconds to return the first byte, no amount of image compression will save it.
Prioritized actions:
- Enable page caching to bring TTFB down first;
- Identify the actual LCP element instead of guessing when optimizing;
- Do not lazy-load the main image in the viewport;
- Preload the main image, or at least ensure it is discovered as early as possible;
- Do not let web fonts block text rendering in the initial view.
INP: How Long Until Response After a Click
INP measures the delay between a user interacting with the page and the next visual update. Clicks, typing, and selections all count.
Poor INP is rarely caused by "slow code"; more often, the main thread is blocked by long-running tasks such as large JavaScript executions, complex event handling, synchronous layouts, and third-party scripts.
Priority actions:
- Use the Performance panel to identify long tasks;
- Split large tasks so the main thread can respond to user input;
- Apply debouncing to input fields, but do not use it to mask slow logic;
- Defer third-party scripts wherever possible, and remove them entirely if you can;
- Avoid large-scale synchronous computation on the main thread.
CLS: Why Pages Suddenly Shift
CLS measures layout shift. Missing dimensions on images, late-inserted ads, font swaps, and dynamic content pushing body text downward can all cause CLS issues.
Priority actions:
- Set explicit width and height (or aspect ratio) on image and video containers;
- Reserve fixed space for ads and embedded content;
- When using
font-display: swap, match fallback font metrics as closely as possible; - Avoid inserting notification banners at the top of the fold that displace content;
- For dynamically loaded content, use skeleton placeholders instead of expanding the layout afterward.
Start with Field Data, Then Review Lab Data
The Core Web Vitals report in Search Console reflects field data from real users, while Lighthouse and WebPageTest provide lab data for reproduction and debugging.
The correct sequence is:
- First, check in Search Console which metrics and which group of pages fall short;
- Then reproduce issues using lab tools;
- Publish changes and wait for field data to roll up;
- Do not assume compliance just because the Lighthouse score turns green.
Optimization Sequence: Server-Side First, Then Above-the-Fold, Finally Interactivity
Step 1: TTFB. Page caching, slow database queries, CDN configuration, and backend blocking all affect TTFB. If TTFB doesn't meet thresholds, LCP will struggle too.
Step 2: LCP Element. Identify the largest above-the-fold element and address image format, dimensions, preloading, and render-blocking resources.
Step 3: CLS. Reserve adequate space for images, ads, and embedded content to avoid layout shifts caused by fonts and dynamic content.
Step 4: INP. Break up long tasks, reduce third-party scripts, and optimize event handlers.
Common Misconceptions
Misconception 1: A perfect Lighthouse score equals CWV compliance. Lab scores represent a single simulation and do not reflect the 75th-percentile experience of real users.
Misconception 2: Lazy-loading the hero image improves LCP. Lazy-loading the primary above-the-fold image typically makes LCP worse.
Misconception 3: Compressing images is all you need for performance. Images are only part of LCP; TTFB, render-blocking resources, and fonts all matter too.
Misconception 4: CLS is only about missing width and height on images. Ads, fonts, and dynamically inserted content are equally common culprits.
Pre-Launch Self-Checklist
- Reviewed Search Console field data, not just Lighthouse scores;
- Confirmed the LCP element for every page type;
- The above-the-fold hero image is not lazy-loaded, and its size and format are optimized;
- TTFB is backed by caching and CDN;
- Images, ads, and embedded content all have reserved space;
- Third-party scripts have been audited; non-essential ones are not loaded;
- Monitoring is in place after changes, so you can observe field data variations.
Conclusion
The essence of CWV is user experience, not three scores that can be gamed. Fix TTFB first, then the first-screen load, and finally interactivity; changing only one variable at a time lets you identify which step actually made a difference.
Related Reading: