Why it shows up in your week
The paid landing page hero is a 3MB PNG. On office Wi-Fi it feels fine. On a phone in a train the offer appears after the user already tapped back. That late paint is LCP.
In plain language
Identify the LCP element in field tools (CrUX, RUM) and lab traces. Optimise that element first: size, format, priority, and server response.
A passing lab score on a developer laptop can still fail mobile field data. Field data is what Google and users experience.
Priority hints, modern image formats, and correct dimensions help. So does cutting render-blocking CSS and JS in the critical path.
TTFB feeds LCP. If the HTML arrives late, paint starts late.
Money templates (home, pricing, campaign landers) deserve LCP budgets in CI, not only in an annual audit.
Facts worth keeping
- Good threshold
- ≤2.5s at the 75th percentile
- Needs improvement
- 2.5s–4s
- Poor
- >4s
- Common LCP nodes
- Hero image, large heading, video poster
- Family
- Core Web Vital with INP and CLS
Not the same as
- TTFBTime until the first byte of HTML. Influences LCP but is not LCP.
- INPMeasures interaction responsiveness after load/interaction, not the main content paint.
- Fully loaded timeOlder marketing metric. LCP focuses on when the main content appears.
Where it bites
LCP hurts on paid landing pages and relaunch templates where a heavy hero ships without priority hints.
Conversion drops while the design still looks premium in screenshots taken on a fast laptop.
What to check
- What is the LCP element on the money templates in mobile field data?
- Is the hero image correctly sized, compressed, and prioritised?
- Is TTFB low enough that the browser can start painting early?
- Which render-blocking CSS/JS delay first paint on those templates?
- Do campaign landers share the same budget as the homepage?
Common questions
What is LCP?
Largest Contentful Paint measures how long until the largest visible content element in the viewport has rendered.
What is a good LCP score?
Google treats 2.5 seconds or faster at the 75th percentile as good. Between 2.5s and 4s needs improvement. Over 4s is poor.
What should you fix first for LCP?
Fix the actual LCP element: image weight and priority, critical CSS, and server response time on commercially important templates.
Does LCP affect SEO?
Yes. It is a Core Web Vital used in page experience signals, and slow main content also hurts conversion.
Lab or field data?
Use both. Field data (CrUX/RUM) decides whether users are fine. Lab traces explain why.