Website speed optimisation matters most on mobile data, which is how most people in Pakistan reach your site. A 1.7 MB page with 90 requests may feel acceptable on office broadband; on a phone in a market or a court corridor, it is a blank screen for several seconds.
The Three Numbers Google Measures
Google’s Core Web Vitals measure what a visitor feels. A page passes when at least three-quarters of real visits meet each threshold:
| Metric | What it measures | Good |
|---|---|---|
| Largest Contentful Paint (LCP) | How soon the main content appears | 2.5 seconds or less |
| Interaction to Next Paint (INP) | How quickly the page responds to a tap or click | 200 milliseconds or less |
| Cumulative Layout Shift (CLS) | How much the layout jumps while loading | 0.1 or less |
INP replaced the older First Input Delay metric on 12 March 2024, so advice written before then may quote the wrong measure.
Where the Weight Usually Is
- Images. One hero photograph saved as PNG can weigh more than the rest of the page. On RightLaw.pk the homepage image was 1.4 MB; as WebP at the size it is displayed, it is 67 KB.
- Page-builder CSS and icon fonts. Theme and builder files often load hundreds of kilobytes of styles and icon sets the page never uses.
- Scripts. Sliders, animations, chat widgets and tracking tags each add requests and block the main thread.
- Redirects. Every internal link that redirects costs a round trip before anything loads.
- Web fonts. Two or three families in several weights can add 200 KB on their own.
Our Website Speed Optimisation Method
- Measure the real pages — homepage, top service pages and top landing pages — and record weight, requests, LCP, INP and CLS.
- List every resource above 50 KB and every request that does not need to load before the content.
- Convert images to WebP at the displayed size, give each one width and height, and lazy-load everything below the first screen.
- Remove unused CSS, fonts and scripts; replace heavy widgets with lighter equivalents or plain links.
- Fix redirects and caching headers.
- Measure again and give you the before-and-after table.
When Website Speed Optimisation Is Not Enough
Some sites are slow because of what they are built on: a theme demo with a dozen plugins, each loading its own files on every page. You can trim that, but the floor stays high. In those cases a rebuild in Next.js usually costs less than a year of patching, and the URLs, content and rankings can be carried across intact.
Frequently Asked Questions
Does speed really affect rankings?
Google uses page experience, including Core Web Vitals, as one of many signals. It is rarely the deciding factor for relevance, but between two similar pages the faster one has the edge, and a slow page loses visitors whatever its ranking.
Will speed work change how my site looks?
Usually not. Images look the same in WebP; removed scripts are ones the page did not need. Where something visible would change, such as an animation, we ask first.
Which speed test do you use?
PageSpeed Insights for Google’s field data, and lab tools such as Lighthouse and Pingdom for page weight and request lists. Field data is what Google uses; lab tools show why.
Related Pages
Sources
Reviewed by: Mohsin Ali Shah, founder of SEO Content Master
Published by: SEO Content Master, seocontentmaster.com
Last updated: 29 October 2026