When Core Web Vitals fail, investigate what prevents the important content from appearing or the page from responding before reaching for a smaller image or a network tweak. Cloudflare’s BEACON data points to LCP discovery and rendering, JavaScript execution and page presentation for INP, and the exact shifting element for CLS. Use those patterns to choose where to investigate; confirm the cause on your own pages and user segments.
What BEACON can—and cannot—tell you
Cloudflare describes BEACON (Browser Experience Across Cloudflare’s Observed Network) as an anonymized dataset of billions of real-world performance measurements across 10,000 of the largest websites on its network. It covers major browser engines, is updated daily in BigQuery, and follows the community-led RUM Archive standard. It reports the three current Core Web Vitals: Largest Contentful Paint (LCP), Interaction to Next Paint (INP), and Cumulative Layout Shift (CLS). Those characteristics make BEACON useful for understanding real-user patterns, but its results are not a diagnosis of your individual site.
The practical lesson is to use a failing metric to choose what to inspect, then use field data and browser performance tools on the affected pages to find the specific cause. A site-wide score alone can conceal a problem limited to a particular URL, browser, operating system, country, or element.
What should you fix first when Core Web Vitals fail?
Use the failing metric to prioritize the investigation rather than applying the same optimization to every page. Cloudflare’s Needs Improvement breakdowns show where time or visual instability can accumulate; they are diagnostic evidence, not promised fixes or values that every site should expect.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →#1 Best Overall
- EXPAND YOUR HORIZONS: 3440 x1440 UltraWide QHD (WQHD) resolution with 21:9 aspect ratio for efficient productivity
- CURVED IMMERSION: The 1500R radius curved VA panel allows for more immersion and better color accuracy. It can also help alleviate eye strain during long hours of working
- RICH COLORS FOR WORK AND PLAY: Ultra Wide-Color technology produces true-to-life images and a wider spectrum of colors with sRBG 123.24 percent , NTSC 99.25 percent color gamut area coverage
- WINDOWS HELLO WEBCAM WITH NOISE-CANCELING MIC: Comes with built-in 5MP webcam, noise canceling microphone, and speakers, perfect for remote working. The webcam is equipped with advanced sensors for Windows Hello facial recognition, which conveniently logs you into your Windows devices in less than 2 seconds
- ONE CABLE IS ALL YOU NEED: USB-C docking transfers high-speed data, high-resolution video signal, and power to your laptop (up to 65W of Power Delivery support) via a single USB-C cable. Play and work in high resolution while simultaneously charging your notebook
| Metric | What BEACON breaks out or exposes | First investigation |
|---|---|---|
| LCP | Document TTFB, load delay, load duration, and render delay | Check response and HTML delivery, when the LCP candidate is discovered, and what blocks its rendering before focusing on transfer size. |
| INP | Input delay, processing time, and presentation delay | Profile the slow interaction, its JavaScript work, and the layout or style work required to present the result. |
| CLS | Elements with negative impact and their CurrentRect and PreviousRect geometry in Debug View | Identify the element that moves and inspect its space reservation and the content or module that changes its dimensions. |
How to investigate an LCP failure
LCP measures when the page’s largest visible content element appears. Cloudflare breaks the interval into four parts. In BEACON’s Needs Improvement row, Cloudflare reports the following values:
| LCP sub-part | Needs Improvement value | What to inspect |
|---|---|---|
| Document TTFB | 1,015 ms | Server response and delivery of the document. |
| Load delay | 1,049 ms | How soon the browser can discover the LCP resource or candidate. |
| Load duration | 199 ms | How long the resource takes to transfer once loading begins. |
| Render delay | 437 ms | Work that keeps the candidate from being rendered after it is available. |
These are Cloudflare’s reported sub-part values for the dataset’s Needs Improvement row, not thresholds or a profile of every failing page. Cloudflare’s interpretation is that the larger opportunities for pages missing Good are usually earlier discovery of the LCP candidate and unblocking its rendering; downloading the image, font, or video often contributes least.
- Check document response and delivery. Investigate TTFB and whether the HTML arrives promptly. A browser cannot discover resources that the document has not revealed.
- Make the LCP candidate discoverable early. Check whether the browser encounters the actual largest visible element promptly, rather than discovering its resource only after other work.
- Remove work blocking its render. Inspect stylesheets, scripts, and other work that delays the candidate from appearing.
- Optimize transfer size after that. If load duration is the meaningful bottleneck for the affected users, then tune the image, font, or video transfer.
Cloudflare identifies Smart Hints as one way to address resource-discovery delays. Treat it as a potential aid to investigate, not a substitute for confirming which resource is late or why.
Rank #2
- Dell UltraSharp U2410 - LCD display - IPS- 24" - widescreen - 1920 x 1200 / 60 Hz
How to investigate a slow INP
INP reflects the responsiveness of interactions during a visit. Its sub-parts help distinguish time spent before processing, time spent handling the interaction, and time before the visual result is presented. In BEACON’s Needs Improvement row, Cloudflare reports input delay of 32 ms, processing time of 112 ms, and presentation delay of 111 ms.
For the slowest interactions, Cloudflare reports JavaScript execution as the longest period. It also says presentation time rises with complex CSS layout recalculations. That points toward profiling the interaction’s actual work rather than assuming the network is the main cause.
- Profile long tasks around the slow interaction and find expensive event handlers.
- Reduce or defer third-party JavaScript that competes with user-facing work.
- Split expensive handlers so they do not hold up the interaction.
- Simplify layout and style work when recalculation delays the next presentation.
Cloudflare identifies Zaraz as a way to reduce third-party JavaScript impact. Whether it helps depends on whether third-party scripts are contributing to the interaction delay you observe.
How to find the cause of layout shifts
A CLS score tells you that visible content moved; it does not by itself tell you which element moved or what caused the movement. Cloudflare Web Analytics Debug View identifies the top five elements with negative impact and exposes P50, P75, P90, and P99 values. For CLS, it records CurrentRect and PreviousRect layout information for the largest shifted area.
Use that evidence to inspect the element’s geometry and surrounding content. Check whether media or injected modules have reserved dimensions, and whether content inserted during loading changes the space other elements occupy. Prioritize elements that repeatedly shift in field data rather than guessing from a single local reproduction.
How to read field data without hiding the problem
Use P75 for the standard rating
Use the 75th percentile (P75) for the standard Good, Needs Improvement, or Poor decision. Then inspect P90 and P99 to see whether a worse tail experience is affecting a subset of visits. A median can look healthy while a meaningful group of users has a poor experience.
Rank #4
- Exclusively compatible with N scale locomotives, this test stand works with most standard N gauge models, making it a flexible and practical addition to any hobbyist’s collection.
- With a straightforward structure and simple connection design, the rolling test stand can be assembled and put into use quickly without complicated tools or procedures.
- Manufactured from premium metal materials, the test bench offers excellent stability and wear resistance, ensuring reliable performance even with frequent daily use.
- This dedicated test treadmill lets you safely debug, test speed and inspect locomotive performance, greatly upgrading the fun and professionalism of your model train hobby.
- This is not a toy. Not intended for use by children under 14.
Segment before choosing a fix
Compare results by URL, browser, operating system, country, and element. These dimensions can reveal a browser-specific rendering issue, a region-specific delivery problem, or a small set of pages responsible for the site-wide result. When comparing findings, keep five things together: the metric, percentile, sub-part, user segment, and remediation lever.
Apply the published LCP thresholds precisely
Google’s current Search Central documentation sets the LCP thresholds used for the Good, Needs Improvement, and Poor categories. Its Search Console report uses actual user data for LCP, INP, and CLS.
| LCP rating | Google threshold |
|---|---|
| Good | ≤ 2.5 seconds |
| Needs Improvement | > 2.5 to 4 seconds |
| Poor | > 4 seconds |
These values are LCP thresholds; do not apply them to INP or CLS. Use the corresponding metric’s own rating criteria when evaluating those results.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteBest Value
What Cloudflare’s collection details mean for interpretation
Cloudflare Web Analytics collects measurements with a lightweight JavaScript beacon. For the Vitals Explorer, Cloudflare says it does not use client-side cookies or localStorage, and that the source IP received by its RUM service is discarded at the nearest data center.
Core Web Vitals are reported when the page first becomes hidden after the page-load event. Data is sent to the site’s /cdn-cgi/rum endpoint when the site is proxied through Cloudflare, or to Cloudflare’s RUM endpoint otherwise. This collection context helps explain what the field data represents; use the page- and segment-level details to decide what to change.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
