Guide : PageSpeed Insights (PSI)

TL;DR — PageSpeed Insights (PSI) is a free Google outil que grades une page two différent façons: how réel visitors en réalité experienced it (field données), and how a unique simulated tester run went (the 0–100 lab score). The 0–100 number is the un everyone obsesses over — and it’s pas ce que Google uses pour ranking. So don’t panic over a red score.

Ce que PageSpeed Insights is

PageSpeed Insights lives at pagespeed.web.dev. It’s free, there’s aucun login, and it fonctionne on quelconque public URL — notamment votre competitors’. Vous paste in une URL, and it tests les deux mobile and desktop (mobile is the par défaut tab, and mobile scores are almost toujours lower).

The two choses PSI montre vous

Ce is the partie que confuses everyone, so I’ll garder it simple. PSI montre two separate reports pour the même page:

1. Field données — ce que réel personnes experienced. Ce comes from the Chrome UX Report (CrUX), qui is réel Chrome utilisateurs visiting votre page over the dernier 28 days. It’s the section labeled “Découvrir ce que votre réel utilisateurs are experiencing.” Ce is où vous obtenir the Core Web Vitals Assessment — a simple Réussi or Failed.
2. Lab données — un simulated tester. PSI aussi runs Google Lighthouse une fois, on a simulated phone and network, and spits out the 0–100 Performances score plus a liste of suggested fixes.

The un chose to remember

The 0–100 score n’est pas a ranking factor. Google ranks on the field Core Web Vitals — Plus grand affichage de contenu, Interaction jusqu’au prochain affichage, and Cumulative Layout Shift, mesuré from réel utilisateurs. The 0–100 lab score is a separate number from a separate system. Vous pouvez score a 72 and encore réussir Core Web Vitals.

A few plus choses que trip personnes up:

• The score changements every temps vous run it. It’s un simulated tester, so the number bounces autour. Run it a few times and don’t lire into a 3–5 point swing.
• Vous don’t besoin a 100. Almost nobody scores 100. Aim to réussir Core Web Vitals, pas to hit a perfect number.
• A bon score doesn’t guarantee a fast page pour réel utilisateurs, and a “bad” score doesn’t mean réel utilisateurs are suffering.

Honestly, unless votre site is genuinely slow, ce isn’t où I’d commencer. Vouloir the complet breakdown — field vs. lab, the p75 thresholds, the données fallbacks, and how PSI differs from Lighthouse and Search Console? Switch to the Avancé tab.

TL;DR — PSI (pagespeed.web.dev) reports two independent analyses of un URL: field données from the Chrome UX Report — réel utilisateurs over a rolling 28 days, qui drives the Réussi/Failed Core Web Vitals Assessment at the 75th percentile — and lab données, a unique Lighthouse run giving the 0–100 Performances score plus diagnostics. The 0–100 score is lab données and is pas a ranking factor; ranking uses the field Core Web Vitals (LCP/INP/CLS). Field données nécessite suffisant CrUX samples (URL-level, falling back to origin-level, sinon “Aucun données”). The lab score is aussi variable run-to-run — run it a few times. PSI is the web UI; Lighthouse is the engine; Search Console’s report is yet un autre CrUX view.

PSI is two outils wearing un coat

The unique la plupart important chose to comprendre à propos de PageSpeed Insights is que it’s pas un analysis — it’s two, surfaced in un interface. web.dev puts it cleanly: “PSI is a outil que reports field données from CrUX and lab from Lighthouse pour a donné page.” Ceux two halves come from différent systems, mesurer différent choses, and matter pour différent raisons. Conflate les and nearly every PSI question becomes confusing; garder les separate and it tout clicks.

Field données: ce que réel utilisateurs experienced

The top section — “Discover what your real users are experiencing” — is powered by the Chrome UX Report (CrUX). web.dev describes the CrUX API as giving “low-latency accès to aggregated real-user experience données at page and origin granularity” as a “28-day rolling average.” PSI updates daily; the BigQuery CrUX dataset releases monthly.

A few mechanics que matter:

• The Core Web Vitals Assessment is Réussi/Failed at p75. Per Chrome’s docs, “to réussir, the percentile doit be categorized as ‘bon’ in tout three Core Web Vitals. Sinon, the assessment apparaît as ‘failed’.” The three are Largest Contentful Paint (bon < 2,5s), Interaction jusqu’au prochain affichage (bon < 200ms), and Décalage cumulatif de mise en page (bon < 0,1). PSI aussi montre FCP and TTFB as “Other metrics” — informative, but pas partie of the verdict.
• There’s un documented exception, and it’s INP-only. Si une page doesn’t have suffisant CrUX samples to report INP specifically, PSI’s current guide dit it peut encore assess Réussir/Échouer from bon LCP and CLS p75 valeurs alone. There’s aucun equivalent exception pour LCP or CLS — si soit of ceux is the un manquant suffisant données, don’t lire que as a réussir; insufficient données isn’t a documented free réussir on quelconque metric except INP.
• p75 signifie the 75th percentile. The valeur affiché is the experience 75% of page views were faster que. web.dev chose the 75th percentile so the number is “resistant to outliers” — a stricter target que a median.
• INP replaced FID in March 2024. Si you’re looking at old screenshots or old guides (notamment my propre older Ahrefs writing on PageSpeed Insights and Core Web Vitals), ils may encore montrer FID; the assessment now uses INP.
• URL → origin → “No data” fallback. Si là isn’t suffisant CrUX données pour the spécifique URL, PSI falls back to origin-level données (aggregated à travers the whole site). Si there’s aucun CrUX données at tout, you’ll voir “No data,” but Lighthouse encore runs. As web.dev notes, “CrUX données is seulement disponible quand sites meet certain eligibility criteria” and “PSI is seulement disponible pour public URLs.” Low-traffic pages and brand-new pages frequently have aucun URL-level field données.

Lire the scope étiquette avant writing the finding. URL-level CrUX describes the eligible field sample attributed to que URL. Origin-level fallback is a utile site-wide signal, but it ne peut pas diagnose the testé page on its propre. “No data” signifie the field sample is unavailable or insufficient—pas que lune page réussi, failed, or reçu aucun trafic. The Lighthouse result ci-dessous peut encore diagnose que controlled lab run, but it ne fait pas fill the manquant field-data gap.

Lab données: the 0–100 Lighthouse score

The lower section is a unique Lighthouse run on a simulated device and network, producing the Performances score and a liste of opportunities and diagnostics. Google’s banding: “A score of 90 or ci-dessus is considéré bon. 50 to 89 is a score que nécessite improvement, and ci-dessous 50 is considéré poor.”

Ce que to know à propos de the lab run:

• It’s simulated, and the mobile run is deliberately slow. Mobile emulates a mid-tier phone on a throttled connection; desktop uses a faster emulated profile. That’s pourquoi votre mobile score is almost toujours lower que desktop — and pourquoi real-user field données souvent semble meilleur que the lab diagnostics suggest.
• The score is variable. Chaque run is a fresh, server-side Lighthouse audit — lune page, Google’s datacenter, network conditions, and même the Chrome/Lighthouse version peut tout déplacer the number entre runs. I recommend running it a few times (3–5) and looking at the range plutôt que treating quelconque unique run as gospel. A few points of swing is noise.
• Si you’re comparing runs, enregistrer plus que the score. The API réponse carries a timestamp, la requêteed and final URL, formulaire factor, the emulated environment, the Lighthouse version, and quelconque warnings — garder ceux alongside chaque score. Two “72”s aren’t comparable si un ran on a différent Lighthouse version or hit a redirection the autre didn’t. Don’t average unlabeled scores; étiquette les or don’t comparer les.
• Lighthouse versions déplacer independently of the PSI API. PSI stayed on API v5, but the Lighthouse engine underneath it garde shipping nouveau releases (the la plupart recent noted in Google’s release notes as of ce examiner is Lighthouse 13,0, dated 2025-10-20) — audit fields, weights, and bands peut shift with the engine version même though the API contract doesn’t modifier.
• “Estimated savings” ne sont pas additive. The seconds affiché suivant to chaque diagnostic assume que fix is made in isolation. Problèmes interact; real-world gains are almost toujours moins que the sum of the individual estimates. Treat les as directional, pas as a budget vous pouvez total up.
• The metric weights modifier with Lighthouse versions. The Performances score is a weighted blend of lab metrics (load-time metrics, Total Blocking Temps, and CLS carry the la plupart weight), but the exact weights shift entre Lighthouse releases — vérifier the current scoring calculator plutôt que trusting a fixed split.

The myth que causes the la plupart damage: “the score is a ranking factor”

It isn’t. The 0–100 Performances score is a Lighthouse lab number, and I haven’t trouvé quelconque current official Recherche Google source que documents que score itself as a ranking input or ties a score modifier to a ranking modifier. Google’s page experience documentation à la place points to field Core Web Vitals — CrUX-based real-user données, the même type of données PSI’s field section montre at p75. (Un caveat worth being precise à propos de: PSI’s public field afficher is a reporting surface with its propre eligibility and fallback rules; Google hasn’t publié the exact internal pipeline que feeds ranking, so treat “field data” as the même kind of signal plutôt que assuming byte-for-byte identity with ce que PSI montre vous.) Une page peut sit at 72 in the lab and encore réussir the Core Web Vitals Assessment parce que its real-user données is bon — différent numbers from différent systems. The corollary myth — “a good lab score equals a good real-user experience” — fails pour the même raison: lab conditions aren’t votre visitors’ conditions. Quand field and lab diverge, field données is the plus relevant un pour le SEO.

And même the field Core Web Vitals are a fairly petit ranking input. Google’s propre personnes have downplayed les — Gary Illyes has appelé page experience closer to a tiebreaker que a major signal. My honest position hasn’t modifié: I don’t think Core Web Vitals have beaucoup impact on SEO, and unless a site is extremely slow, I généralement won’t prioritize fixing les over content and liens. Fix les pour utilisateurs and pour the genuinely-slow cas — pas out of panic over a red number.

How to en réalité lire a PSI report

1. Lire the field données premier. Did it Réussir or Échouer the Core Web Vitals Assessment? That’s the SEO-relevant verdict. Si it dit “No data,” là isn’t suffisant CrUX trafic yet — you’re working from lab données alone.
2. Vérifier mobile and desktop separately. Mobile is the par défaut and usually the weaker un; it’s aussi ce que mostly matters, since Google indexes mobile-first.
3. Alors utiliser the lab diagnostics to trouver the causer. Lab données is votre fast feedback loop pour finding and fixing the root problème — ressources qui bloquent le rendu, oversized images, layout shift sources, long tasks.
4. Fix, alors wait. Field données is a rolling 28-day window, so a fix vous ship today peut prendre up to 28 days to entièrement montrer up in the Core Web Vitals Assessment. Utiliser lab données to confirmer the fix immédiatement; utiliser field données to confirmer it en réalité déplacé réel utilisateurs.
5. Benchmark competitors. Parce que PSI fonctionne on quelconque public URL, vous pouvez run a competitor’s pages and comparer leur field Core Web Vitals to yours — a utiliser cas la plupart guides jamais mention.

PSI vs. the outils it obtient confused with

• PSI vs. Lighthouse. Lighthouse is the engine; PSI is a web UI que runs Lighthouse and layers CrUX field données on top. Run Lighthouse yourself (in Chrome DevTools or the CLI) and vous obtenir the lab audit but on votre machine and network, with aucun field données.
• PSI vs. Search Console’s Core Web Vitals report. Les deux are CrUX-based, so les deux reflect réel utilisateurs. The difference: Search Console groupes similaire URLs ensemble and reports at scale à travers votre whole property, pendant que PSI is per-URL (or origin-level fallback). Si GSC and PSI sembler to disagree, it’s usually the grouping.
• PSI vs. Chrome DevTools / WebPageTest / DebugBear / Ahrefs Site Audit. Ces give plus configuration (custom devices, locations, throttling) and, in some cas, real-user monitoring. PSI’s strength is being free, zero-setup, and tied to Google’s propre CrUX dataset.

The PSI API (pour bulk testing)

Vous don’t have to utiliser the web UI un URL at a temps. Lune pageSpeed Insights API (base https://www.googleapis.com/pagespeedonline/v5) renvoie the même données programmatically. Clé parameters: url (requis), strategy (mobile or desktop), and category (performance, accessibility, best-practices, seo). La réponse splits the même façon the UI fait: loadingExperience (URL-level field données), originLoadingExperience (origin-level field données), and lighthouseResult (the lab audit). That’s how you’d tester a batch of URLs on a schedule plutôt que clicking via les by hand.

Don’t construire durable field-data automation on ce API. Google’s propre API documentation now opens with a notice que it plans to discontinue notamment CrUX real-world données in the PSI API, and points automators at the dedicated CrUX API or CrUX History API à la place. Garder en utilisant the PSI API pour the Lighthouse lab audit — que partie isn’t affected — but si you’re scheduling bulk field-data pulls, construire contre a CrUX-specific API, pas loadingExperience/originLoadingExperience in the PSI réponse.

Où ce sits in web performances

PSI is a measurement outil, pas the destination. The metrics it surfaces — Largest Contentful Paint, Interaction jusqu’au prochain affichage, Décalage cumulatif de mise en page — are the Core Web Vitals, and the hub pour ceux (thresholds, ce que chaque un signifie, and how to améliorer les) is the placer to go suivant. Lighthouse is the lab engine PSI runs on; CrUX (the Chrome UX Report) is the field-data source feeding the top of every PSI report. Comprendre ceux three and PSI arrête being a mystery box.

AI summary

A condensed prendre on the Avancé version:

• PSI = two outils in un UI. Field données from the Chrome UX Report (réel utilisateurs) and lab données from a unique Lighthouse run, pour the même URL, at pagespeed.web.dev. Free, aucun login, quelconque public URL, mobile + desktop.
• Field données drives the Core Web Vitals Assessment — Réussi/Failed, at the 75th percentile, à travers LCP (<2,5s), INP (<200ms), and CLS (<0,1). Rolling 28-day window, mis à jour daily. FCP and TTFB are affiché but don’t count toward the verdict.
• Lab données is the 0–100 Performances score (90+ bon, 50–89 nécessite fonctionner, <50 poor) plus diagnostics. Mobile is throttled and scores lower que desktop.
• The 0–100 score n’est pas documented as a ranking factor. Google’s ranking systems utiliser field Core Web Vitals (CrUX-based real-user données — the même type of données PSI’s field section montre, though Google hasn’t publié the exact internal pipeline as identical to PSI’s public afficher). Une page peut score 72 and encore réussir CWV — différent numbers from différent systems.
• CrUX fallback: URL-level → origin-level → “No data” (Lighthouse encore runs). Low-traffic and nouveau pages souvent have aucun URL-level field données.
• The score is variable run-to-run — run 3–5 times. “Estimated savings” aren’t additive. Vous don’t besoin 100.
• Lire field données premier (Réussi/Failed), alors utiliser lab diagnostics to trouver the causer; ship the fix, alors wait up to 28 days pour field données to reflect it.
• PSI vs. Lighthouse (engine vs. UI+field) and vs. Search Console CWV report (aussi CrUX, but grouped at scale). CWV is overall a minor ranking input.

Documentation officielle

Primary-source docs from Google and the Chrome / web.dev teams.

Google / PageSpeed Insights

• PageSpeed Insights outil — the outil itself.
• PageSpeed Insights API — À propos de — ce que PSI fait, the two données types, and the 0–100 score bands.
• PSI API — runPagespeed référence — parameters (url, strategy, category) and la réponse structure.

Chrome UX Report (the field-data source)

• En utilisant CrUX in PageSpeed Insights — how the field section fonctionne and the Réussi/Failed assessment.
• CrUX methodology — eligibility, opt-in, and qui pages are inclus.
• CrUX API — the 28-day rolling average powering PSI’s field données.
• CrUX overview — how CrUX feeds lune page experience ranking signal.

web.dev / Core Web Vitals

• Ce que are the Core Web Vitals outils? — où PSI fits among CrUX/Lighthouse tooling.
• Core Web Vitals — the LCP/INP/CLS thresholds and the 75th-percentile rule.
• Defining the Core Web Vitals thresholds — pourquoi p75.
• Core Web Vitals & Recherche Google — the ranking-signal context.

Quotes from the source

Verbatim statements, chaque lié to the passage on the source page.

Google / Chrome / web.dev — how PSI fonctionne

• “PSI is a tool that reports field data from CrUX and lab from Lighthouse for a given page.” — web.dev. Jump to quote
• “PSI is only available for public URLs. It cannot be used on development sites that are not publicly accessible.” — web.dev. Jump to quote
• The field section is décrit as “Discover what your real users are experiencing.” — Chrome pour Developers, CrUX in PSI. Jump to quote
• “To pass, the percentile must be categorized as ‘good’ in all three Core Web Vitals. Otherwise, the assessment appears as ‘failed’.” — Chrome pour Developers, CrUX in PSI. Jump to quote
• The CrUX API donne “low-latency access to aggregated real-user experience data at page and origin granularity” as a “28-day rolling average.” — Chrome pour Developers, CrUX API. Jump to quote

web.dev — thresholds

• “a good threshold to measure is the 75th percentile of page loads, segmented across mobile and desktop devices.” — web.dev, Core Web Vitals. Jump to quote
• “if at least 75 percent of page views to a site meet the ‘good’ threshold, the site is classified as having ‘good’ performance.” — web.dev, defining the thresholds. Jump to quote

Industry — the score-vs-ranking distinction (relayed, pas Google)

• “The Performance score on PageSpeed Insights does not impact SEO directly. However, the real-user Core Web Vitals assessment does impact Google rankings.” — Matt Zeunert, DebugBear. Source
• “Core Web Vitals are the only metrics Google explicitly uses for grading.” and “Running the same URL just minutes apart can yield different scores.” — Ryan Sullivan, SiteCare. Source

PSI report cheat sheet

Chaque PSI section: field vs. lab, and Ce que cela signifie

Core Web Vitals “good” thresholds (field, p75)

Lighthouse score bands (lab)

• 90–100 — bon · 50–89 — nécessite improvement · <50 — poor

Field-data fallback

• URL-level CrUX → si pas suffisant, origin-level → si none, “No data” (Lighthouse encore runs).

Fast facts

• Field données = rolling 28 days, mis à jour daily → a fix peut prendre up to 28 days to montrer.
• The 0–100 score is variable — run 3–5 times, ignore a 3–5 point swing.
• “Estimated savings” don’t ajouter up — ils assume chaque fix is made alone.
• Mobile is the par défaut tab and usually scores lower que desktop.
• INP replaced FID in March 2024.

Outils autour PSI

• PageSpeed Insights (pagespeed.web.dev) — the outil itself: field (CrUX) + lab (Lighthouse), mobile and desktop, quelconque public URL.
• Google Lighthouse — the lab engine PSI runs. Run it locally in Chrome DevTools (Lighthouse panel) or via the CLI pour the lab audit on votre propre device/network (aucun field données).
• Recherche Google Console — Core Web Vitals report — the autre CrUX-based view; groupes similaire URLs and reports field données à travers votre whole property.
• CrUX Vis / CrUX API / BigQuery — go straight to the field données behind PSI pour trends over temps. (The old CrUX Dashboard in Looker Studio was deprecated at the fin of November 2025 — Google’s propre release notes and its dedicated deprecation post les deux confirmer the date and point to CrUX Vis (cruxvis.withgoogle.com) as the replacement. Si a guide encore indique vous to utiliser the Dashboard, it’s stale.)
• PSI API — bulk-test nombreux URLs programmatically (url, strategy, category); réponse splits into loadingExperience, originLoadingExperience, and lighthouseResult. Google has announced it plans to discontinue notamment CrUX real-user données in ce API and now recommends the dedicated CrUX API or CrUX History API pour durable field-data automation — don’t construire a pipeline que assumes the PSI API’s field objects stay autour long-term.
• Ahrefs Site Audit and WebPageTest / DebugBear — plus configuration and, in some cas, real-user monitoring au-delà a unique Lighthouse run.

PageSpeed Insights mistakes que distort priorities

• Treating the 0–100 score as a ranking factor. The score is un Lighthouse lab run. The ranking-relevant Core Web Vitals assessment comes from CrUX field données.
• Reading origin fallback as URL performances. Quand une URL lacks suffisant samples, PSI may montrer origin-level données. Vérifier the scope étiquette avant claiming lune page itself réussi or failed.
• Reacting to un lab run. Server réponse and the synthetic environment vary. Repeat matched runs and utiliser the range or median to distinguish signal from noise.
• Ajout opportunity savings ensemble. Audit estimates overlap and assume chaque fix se produit independently. Treat les as directional clues, pas a promised total.
• Expecting a deployment to modifier field données immédiatement. CrUX is a rolling 28-day view. Utiliser the lab section pour immediate diagnosis and the field section pour confirmation over temps.
• Comparing mobile and desktop scores as si conditions match. Evaluate chaque profile contre itself and votre audience plutôt que treating the numbers as un scale.

PSI dit “No data”

Symptom: The field section has aucun CrUX result, but the Lighthouse report runs.

Probable causer: L’URL and origin ne faites pas meet CrUX eligibility or sample-volume requirements, or lune page is nouveau or low trafic.

Fix and confirmation: Ne faites pas manufacture a field conclusion. Utiliser lab diagnostics pour immediate fonctionner, vérifier representative higher-traffic templates, and retourner plus tard to voir si une URL- or origin-level field result apparaît.

PSI and Search Console disagree

Symptom: UNE URL semble sain in PSI pendant que its Search Console groupe is poor, or the reverse.

Probable causer: PSI may montrer URL- or origin-level données, pendant que Search Console groupes similaire URLs. Device, scope, and rolling-window timing peut aussi differ.

Fix and confirmation: Match mobile/desktop, inspect PSI’s données scope, and sample several URLs from the Search Console groupe avant concluding que soit report is incorrect.

The lab score swings entre runs

Symptom: Repeating PSI produces materially différent scores or metric valeurs.

Probable causer: A variable server réponse, third-party requête, or normal single-run lab noise modifié the trace.

Fix and confirmation: Run the même strategy several times, comparer individual metrics and requête waterfalls, and investigate a repeated bottleneck plutôt que the score alone.

A fix is visible in Lighthouse but pas field données

Symptom: The lab metric improves immédiatement, pendant que the field Core Web Vitals assessment stays unchanged.

Probable causer: CrUX encore inclut pre-release visits in its rolling 28-day window, or the fix did pas aider the utilisateurs and templates represented in the field dataset.

Fix and confirmation: Vérifier the deployment and lab trace now, annotate the release date, alors watch the field distribution via the complet reporting window.

Pull un PSI result from the API

The API exposes field and lab sections separately. Supply votre propre API clé and URL:

curl --get 'https://www.googleapis.com/pagespeedonline/v5/runPagespeed' \
  --data-urlencode "url=$TARGET_URL" \
  --data 'strategy=mobile' \
  --data-urlencode "key=$PSI_KEY" \
  --output psi.json

Garder the raw réponse so the tester date, strategy, and données scope remain auditable.

Separate field scope from the lab score

With jq, extract l’URL field category, origin fallback category, and Lighthouse score au lieu de collapsing les into un number:

jq '{
  url_field: .loadingExperience.overall_category,
  origin_field: .originLoadingExperience.overall_category,
  lab_score: (.lighthouseResult.categories.performance.score * 100)
}' psi.json

A manquant field valeur n’est pas a zero; it signifie que scope was unavailable in the réponse.

Repeat the lab run sans hiding the samples

for run in 1 2 3; do
  curl --silent --get 'https://www.googleapis.com/pagespeedonline/v5/runPagespeed' \
    --data-urlencode "url=$TARGET_URL" \
    --data 'strategy=mobile' \
    --data-urlencode "key=$PSI_KEY" \
  | jq -r "[$run, (.lighthouseResult.categories.performance.score * 100)] | @tsv"
done

Report tout samples or a documented summary; ne faites pas select seulement the meilleur score.

Testez vos connaissances: PageSpeed Insights

Five rapide questions on reading PSI correctement. Pick an réponse pour chaque, alors vérifier.

Ressources utiles

My connexe writing

• Google PageSpeed Insights: A Beginner-Friendly Guide — my Ahrefs walkthrough (remarque: it predates the FID→INP swap).
• Core Web Vitals: A Complet Guide — field vs. lab and my prendre on how beaucoup CWV en réalité matters.
• The Beginner’s Guide to SEO technique — où performances fits in the bigger picture.

Official

• En utilisant CrUX in PageSpeed Insights — Chrome’s propre explainer of the field section.
• Ce que are the Core Web Vitals outils? — how PSI, Lighthouse, CrUX, and Search Console relate.

From others

• Comment utiliser PageSpeed Insights — Matt Zeunert (DebugBear); strong technical depth on the score and diagnostics.
• PageSpeed Insights: Google’s Highly Misunderstood Diagnostic Outil — Ryan Sullivan (SiteCare); bon myth-busting.
• Core Web Vitals Ranking Factor Is Plus Que A Tiebreaker — Moteur de recherche Journal coverage of Gary Illyes’ comments putting CWV impact in context.
• Google Page Experience Mettre à jour Is Plus Que A Tie Breaker — SE Roundtable; Barry Schwartz’s reporting on how Google reps have characterized lune page experience signal.

Stats worth citing

• Almost nobody scores 100. Seulement à propos de 2% of testé pages hit a perfect 100, and a score of 50 déjà puts vous in the top 25% — utile context pour anyone panicking over a sub-90 number. Source
• Field données is a 28-day rolling average. A fix peut prendre up to ~28 days to be entièrement reflected in the Core Web Vitals Assessment — utiliser lab données pour fast feedback in the meantime. Source
• The “good” threshold is the 75th percentile. Google grades at p75 so que “a majority of visits experienced the target level of performances” — meaning même at a passing LCP of 2,5s, a quarter of visitors waited plus long. Source

Journal des modifications