“Core Web Vitals Assessment: Failed.” It sits in a red-tinted box at the top of PageSpeed Insights, and it is hard not to read it as a verdict on your whole website. Perhaps a developer or an SEO agency has sent you the screenshot with a worried note. Perhaps you found it yourself after a sale went quiet.
A “Core Web Vitals assessment failed” message is telling you something specific and fairly narrow, and the right response depends on reading that box carefully. Read it properly and you can usually fix a failed CWV assessment one metric at a time, without throwing every optimisation plugin at it.
What “Core Web Vitals assessment failed” actually means
The assessment uses field data: real visits recorded by Chrome over the previous 28 days. It looks at the three Core Web Vitals at the 75th percentile, and passes only if all three are in Google’s good range:
- Largest Contentful Paint at 2.5 seconds or less
- Interaction to Next Paint at 200 milliseconds or less
- Cumulative Layout Shift at 0.1 or less
So “failed” means at least one of the three was outside the good range for more than a quarter of recorded visits. It does not mean the site is broken, that Google has penalised you, or that all three metrics are bad. Often it is one metric, on mobile only, slightly over the line. If you want a fuller picture of what each metric measures, the overview of LCP, INP and CLS is the place to start.
Read the box before you do anything
Take a minute with the result itself. Four questions answer most of what you need to know.
1. Which metric is red or amber?
Each metric in the field section is colour-coded and has a bar showing the share of visits that were good, needing improvement and poor. Write down which one failed and by how much. An LCP of 2.7 seconds and an LCP of 6 seconds are different projects.
2. Mobile or desktop?
PageSpeed Insights has separate tabs. It is common for desktop to pass and mobile to fail. Check both before you plan anything, and check that the screenshot someone sent you is the tab you think it is.
3. This URL or the origin?
Near the top of the field section is a switch between “This URL” and “Origin”. URL data is for that single page. Origin data is every page on your domain combined. If the page you tested has too few visits of its own, the tool may only have origin data to show.
This matters a great deal. A failing origin with a passing homepage means some other part of the site, perhaps the blog or the shop, is dragging the average down. A failing URL with a passing origin points at that one page or its template.
4. Is there data for every metric?
Quieter pages sometimes have figures for some metrics and not others, and very quiet sites have none. If you see no field data at all, there is no assessment to fail; the red you are looking at may be the lab score further down the page, which is a different thing entirely.
Low-traffic sites: a different situation
Many small business sites have little or no field data, especially for individual pages. In that case:
- An origin-level result may be based on only your busiest pages. Your homepage and one or two popular posts might be shaping the whole figure.
- Small changes in who visits can tip it. A burst of visitors on slow phones from one campaign can swing the 75th percentile when the total is modest.
- There may simply be no assessment. Then you work from lab tests and checks on a real phone, and judge yourself against the same thresholds.
None of this is a reason to ignore the result. It is a reason to check which pages your field data is really describing before you rebuild something that was not the problem.
A prioritised plan to fix a failed CWV assessment
The order below is the one that tends to waste the least time. It does not cover how to fix each metric in depth; each has its own post, linked as you go.
Step 1: pick one metric, the one furthest from good
Do not try to move all three at once. If LCP is at 4 seconds and CLS is at 0.12, LCP needs far more work, but CLS may be one fixed banner away from passing. Decide based on both the gap and the likely effort, then start.
Step 2: find which pages and template are responsible
In Search Console, go to the Core Web Vitals section under Experience and start with the mobile issues. Its URL groups usually make it obvious whether the problem is the homepage, all product pages, all blog posts, or everything. Fixing a shared template does more than fixing a single page.
Step 3: diagnose on one example page
Take an example URL from the worst group and run the lab test. Use the diagnostics to see the cause:
- For LCP, identify the LCP element and which part of its timing is longest. The method is in our post on breaking LCP into its four parts.
- For INP, the lab cannot measure it; you need to reproduce a slow tap in Chrome DevTools. Tracking down the slow interaction explains how.
- For CLS, make layout shifts visible and see what moves. The common causes are covered in finding what makes the layout jump.
Step 4: make the smallest change that addresses the cause
Exclude the hero image from lazy loading. Add dimensions to the images in the blog template. Remove an abandoned tracking script. Small, targeted changes are easier to test and easier to undo than switching on twenty options in a performance plugin.
Step 5: confirm in the lab, then wait for the field
Re-run the lab test several times and compare against your notes. When you are satisfied, start validation in Search Console for the affected issue. Then wait. Because field data covers 28 days, the assessment improves gradually as older visits age out.
Step 6: move to the next metric
Only once the first is on its way. Fixing LCP sometimes changes CLS or INP slightly, for better or worse, so re-check all three before starting the second.
Common reasons sites fail
Every site is different, but on WordPress and WooCommerce sites the same causes come up again and again.
- Mobile LCP from a large or late hero image, often lazy-loaded by mistake or set as a page-builder background.
- Slow server response on pages that are not cached, which pushes LCP back before anything else starts.
- Mobile INP from too many scripts: analytics, pixels, chat, consent tools and builder add-ons all starting at once.
- CLS from cookie notices, promo bars and images without dimensions, especially on first visits.
- A recent change: a new theme, a redesign, a marketing tag added for a campaign. Check the Search Console chart for a date when things got worse.
What a failed assessment does not mean
It is worth keeping this in proportion.
- It is not a penalty notice. Google’s own position is that page experience is one consideration among many, and relevant, helpful content comes first. The question of how much it weighs in search is covered in whether Core Web Vitals affect SEO.
- It does not mean every visitor has a slow experience. Most visits may well be fine; the result fails once more than a quarter of them are not.
- It does not mean your lab score must reach 90 or 100. The lab score is not what the assessment measures.
That said, the visitors behind a failed result are real, and many are on phones. Fixing the cause usually makes the site better for the people you most want to reach. For how this fits into speed work more broadly, see the website loading speed guide.
Frequently asked questions
My homepage passes but the assessment for my site fails. Why?
You are probably seeing the origin result, which combines every page on your domain. Other templates such as product pages, blog posts or a slow landing page can pull the combined figure below the threshold even when the homepage is fine. Use Search Console’s URL groups to find which kind of page is responsible.
The lab test shows good numbers but the assessment still fails. What now?
The assessment reflects 28 days of real visits, so a recent fix may not have worked through yet. Real visitors also use slower devices and interact with the page, which the lab does not. Give a recent change a few weeks, and test interactions on a real mid-range phone in case INP is the hidden problem.
Can a failed assessment pass again by itself?
Sometimes, if it was close and your visitor mix changes, for example after a campaign bringing slow mobile traffic ends. It can also fail again for the same reason. Relying on that is risky. If a metric is close to the line, a small targeted fix gives you some margin rather than leaving the result to chance.
Why did my assessment change from passed to failed when nothing on the site changed?
Something outside the site often has. A third-party script such as a chat widget or ad tag may have been updated, your visitor mix may have shifted towards slower phones, or a page that was quiet now has enough traffic to be counted. Check the Search Console chart for the date and compare with any campaigns or tag changes.
We would rather help you fix the one metric that is failing than sell you work on all three. Our website speed optimisation service starts with an audit of the failing pages, sets out the order of work, and applies fixes only once you agree.
Next in this series: Do Core Web Vitals Affect SEO? What Google Has Said
