Tools and providers
Part of Core Web Vitals audit checklist for UK ecommerce sites
Core Web Vitals audit checklist for UK ecommerce sites
Core Web Vitals audit checklist for UK ecommerce sites: test LCP, CLS and INP, compare UK hosting benchmarks and fix mobile-first indexing issues.
What to take away
- Google's pass marks are LCP 2.5 seconds, CLS 0.1 and INP 200 milliseconds, measured at the 75th percentile of real visits.
- A template passes only when all three metrics pass. Two out of three is a fail, and there is no separate UK threshold.
- Origin response time sets the floor under LCP, and a CDN edge cannot repair a slow checkout query.
- Audit field data first, then reproduce each failure in the lab on a 4G mobile profile.
Thresholds and the pass rule
Google's published marks are LCP 2.5 seconds, CLS 0.1 and INP 200 milliseconds. They apply at the 75th percentile of real visits, so a quarter of sessions can be slower while a page still passes. These numbers are the same for UK stores.
Core Web Vitals pass marks
- 2.5sLCP good threshold
- 0.1CLS good threshold
- 200msINP good threshold
| Metric | Pass mark | Usual UK ecommerce failure |
|---|---|---|
| LCP | 2.5s or less | Slow origin at checkout, unoptimised hero image |
| CLS | 0.1 or less | Cookie banner, promo bar, web font swap |
| INP | 200ms or less | Filter and add-to-cart scripts, third-party tags |
Judge each template on its own. Product, category, cart and checkout pages score differently, so a passing homepage says nothing about the rest of the store.
Hosting and CDN effects on LCP
Time to first byte is the floor under LCP. A UK shared server that returns 800ms before any HTML arrives leaves little room to reach 2.5 seconds on 4G.
Hosting delay to LCP
- Shared hosting: many sites on one server
- Slow database query at checkout
- Delay passes into time to first byte
- TTFB delay passes into LCP
- UK CDN edge cannot fix slow origin
A CDN edge helps with static assets, images and fonts. It does not fix a slow database query behind your checkout.
Target TTFB to verify
- Cloudflare, London edge
- under 200ms
- Amazon CloudFront, London region
- under 200ms
- Fastly, London POP
- under 200ms
- Krystal or Kinsta, UK managed hosting
- under 400ms
When it helps LCP
- Cloudflare, London edge
- Caching HTML and assets near UK shoppers
- Amazon CloudFront, London region
- Stores already hosted on AWS
- Fastly, London POP
- Dynamic checkout traffic needing quick connection setup
- Krystal or Kinsta, UK managed hosting
- Smaller stores wanting edge caching without a CDN project
Those figures are test targets, not published vendor benchmarks. Confirm them from your own UK test node. Your server logs record origin response time per template, and log file analysis for technical SEO shows how to read them alongside crawl and checkout paths.
Auditing each page template in turn
Run this sequence on home, category, product, cart and checkout pages.
- Open the Core Web Vitals report in Google Search Console, set it to mobile, and export the page groups so URLs land in template buckets. Google Search Console documents the grouping and export options.
- Run PageSpeed Insights on one live URL per template using the mobile 4G profile.
- Split LCP into time to first byte, resource load delay and resource load time, then fix the largest share first.
- Set width and height on product images, serve AVIF or WebP, and add fetchpriority="high" to the LCP image only.
- Load web fonts with font-displayswap and preload one heading font.
- Record long tasks while you exercise filters and add-to-cart buttons. JavaScript rendering and third-party tags are frequent INP causes.
- Reserve space for the consent banner so it cannot shift layout, and keep it within ICO guidance on cookies and similar technologies.
- Repeat the lab test and recheck that template's field data after each release.
For repeatable grouping, the guide to exporting Google Search Console bulk data covers the BigQuery route.
Example: product template before and after
These illustrative figures come from a mobile product template measured at the 75th percentile.
Product template before and after
| Metric | Before | After | Result |
|---|---|---|---|
| LCP | 4.3s | 2.1s | Pass |
| CLS | 0.26 | 0.04 | Pass |
| INP | 380ms | 170ms | Pass |
Before the work, all three metrics failed. After it, all three passed, so the template passed its assessment.
The fixes were a 1.2MB hero image replaced with a 180KB AVIF with fixed dimensions, fetchpriority="high" on that image, a reserved min-height for the consent banner, and a deferred filter script.
Measuring progress after fixes
Field data is the final check because it reports real UK sessions. Lab data explains why a metric failed and confirms a fix before release. Compare the same template and device before and after each change.
Keep one sheet per template per device with the values, the fix applied and the date. Record the review date for the next audit while the current one is fresh. Quarterly suits a stable store, and monthly suits stores that deploy often or add tags.
For the wider checks behind these numbers, see technical SEO audit checks UK ecommerce sites.
Common questions
Do UK Core Web Vitals thresholds differ from Google's pass marks?
No. Google applies one set of thresholds worldwide. UK work concerns test locations, mobile networks and origin hosting, not different numbers.
Which tool gives template-level field data?
Page groups in the Search Console Core Web Vitals report are the quickest route. The BigQuery export lets you regroup URLs by pattern and keep history.
What causes the worst CLS on UK product pages?
Cookie banners, promotional bars, late web fonts and images without width and height attributes. Reserve space for each element before it renders.
Will a CDN fix a slow LCP?
Only for static assets, images and fonts. If TTFB stays high from the origin, fix the server or database query before adding an edge.
