Core Web Vitals for marketers: what to fix first
A plain-English guide to LCP, INP and CLS for marketing teams: what each metric means, why it matters for SEO and sales, and which fixes to make first.

Short answer: Core Web Vitals are three measures of how a page feels to real visitors: how fast the main content appears (LCP), how quickly the page responds to taps and clicks (INP) and how stable the layout is while it loads (CLS). Fix the slowest templates first, starting with your highest-traffic landing pages, and focus on images, third-party scripts and reserved space before anything exotic.
Most marketing teams meet Core Web Vitals as a red warning in Search Console or a low score in a report. The good news is that the fixes that matter most are rarely mysterious, and many of them sit within marketing’s control: the images you upload, the scripts you add and the widgets you install.
The three metrics, in plain English
| Metric | What it measures | Good | Usual culprit |
|---|---|---|---|
| Largest Contentful Paint (LCP) | How long until the main content appears | 2.5 s or less | Large hero images, slow servers |
| Interaction to Next Paint (INP) | How quickly the page reacts to input | 200 ms or less | Heavy JavaScript, too many tags |
| Cumulative Layout Shift (CLS) | How much the layout jumps while loading | 0.1 or less | Images, ads and fonts without reserved space |
Google judges these from real Chrome users at the 75th percentile, which means three out of four visits need to meet the threshold. You can see the results in the Core Web Vitals report in Google Search Console, grouped by similar URLs. The official definitions and thresholds are documented on web.dev.
INP replaced First Input Delay in March 2024. If an old report mentions FID, it is out of date.
Why marketers should care
Page experience is a ranking signal, but the bigger effect is on conversion. A faster, steadier page keeps more of the visitors you paid or worked to attract. Slow LCP means people bounce before your headline appears. Poor INP means a ‘Book a demo’ button that seems broken. High CLS means someone taps the wrong link because the page moved under their thumb.
Speed also supports search and AI visibility in quieter ways. Fast, lightweight pages are cheaper for search engines and AI crawlers to fetch, and they tend to be easier to render fully, so your content is actually seen.
Where to look first
Start with data rather than guesses:
- Search Console Core Web Vitals report. Shows which URL groups fail on mobile and desktop, based on real users.
- PageSpeed Insights. For a specific URL, shows field data at the top and a lab diagnosis below with specific opportunities.
- Your analytics. Sort landing pages by sessions and conversions. A slow page with lots of paid or organic traffic is the priority.
Most sites have a handful of templates, such as home, service, blog post, product and category. Fixing a template fixes every page that uses it, so think in templates, not individual URLs.
What to fix first
- Compress and size images. Serve modern formats like AVIF or WebP, at the size they display. A 4,000-pixel photo shown at 800 pixels wastes most of its bytes. See image optimisation.
- Do not lazy-load the hero. Lazy loading is great below the fold, but the main image visible on first load should load immediately, ideally with a preload hint and high fetch priority.
- Cut third-party scripts. Audit chat widgets, heatmaps, A/B testing tools, social embeds and tag manager containers. Remove what nobody uses and load the rest after the page is interactive or on first interaction.
- Reserve space. Give images, video embeds, ad slots and cookie banners fixed dimensions so the layout does not jump.
- Self-host fonts with
font-display: swap, preload the one or two files you need above the fold and use a size-matched fallback to reduce shifting. - Improve server response. A slow TTFB delays everything. Caching, a CDN and better hosting often help more than any front-end tweak.
- Render content on the server. Static or server-rendered HTML beats client-side rendering for both speed and crawlability. See static site generation.
Common traps
- Chasing a perfect lab score. A Lighthouse score of 100 is nice, but real-user field data is what counts. Focus on passing thresholds for real visitors on mobile.
- Adding tools to fix speed. Optimisation plugins that inject more JavaScript can make INP worse. Removing things usually beats adding things.
- Ignoring mobile. Most failures are on mobile devices with slower processors and networks. Test on a mid-range phone, not only a fast laptop.
- Forgetting new content. One huge uncompressed image in a new blog post can undo months of work. Build image limits into your publishing checklist.
A quick example
An online furniture retailer had an LCP of 4.6 seconds on its category pages. The causes were a 1.8 MB lazy-loaded hero banner, three tag manager containers and fonts loaded from two different services.
The team compressed the banner to 140 KB in AVIF, loaded it eagerly with a preload, merged the tag containers into one and delayed non-essential tags until first interaction, and self-hosted a single font family. LCP fell to 1.9 seconds at the 75th percentile within a month, and mobile conversion rate on those pages rose by 11 percent.
How we build for it
Every site our web development team builds starts static-first, ships minimal JavaScript, self-hosts fonts and loads analytics only after the page is interactive. This website is built the same way and scores 100 for performance in Lighthouse. If your site is struggling, a free audit will show which templates to fix first and what each fix is likely to be worth.