Core Web Vitals India: Website Speed Fixes for Sites on 4G

This guide is for owners and marketing managers of Indian business websites, and for the developers who maintain them. By the end you will be able to read a PageSpeed Insights report, know which metric your site fails and why, and decide whether the fix is an afternoon of image work, a hosting move or a rebuild.

Core Web Vitals India: what the three metrics measure and why 4G changes the picture

Core Web Vitals are three numbers Google collects from real Chrome users and reports in Search Console. LCP (Largest Contentful Paint) is how long the biggest visible element, usually the hero image or heading, takes to appear. CLS (Cumulative Layout Shift) measures how much the page jumps around while loading. INP (Interaction to Next Paint) measures how long the page takes to respond after a tap.

The grade is not an average. Google takes the 75th percentile of visits over the last 28 days, so the slowest quarter of your visitors sets your score. A site that feels instant on office broadband can still fail because of the slowest 25% of visitors.

Mobile speed on 4G: the visitor your report is actually scoring

Indian 4G plans advertise 20 to 40 Mbps, but effective throughput in a busy market or an industrial estate at 11 AM is often 2 to 5 Mbps with 80 to 150 milliseconds of latency. A handset sold for 10,000 to 15,000 rupees also takes three to four times longer than a laptop to run the same JavaScript. Bandwidth problems show up as slow LCP; processor problems show up as poor INP. You need to fix both.

Does speed affect ranking?

Yes, but less than most owners fear and more than most developers admit. Google confirms that page experience, including Core Web Vitals, is a ranking signal, and also says that relevant content outranks a faster page with weaker content. In practice speed is a tie-breaker between pages of similar quality.

The bigger effect is commercial. Google's own research found the probability of a bounce rises by about 32% when load time goes from 1 second to 3 seconds. Fix speed for revenue first and treat the ranking lift as a bonus.

PageSpeed Insights score meaning: lab score versus field data

A PageSpeed Insights report has two halves and people confuse them constantly. The top block, labelled as what real users are experiencing, is field data from the Chrome User Experience Report and is the only part that affects your status in Search Console. The lower block is a Lighthouse lab test: one simulated load on a simulated mid-range phone over a simulated slow 4G link, producing the 0 to 100 score everyone screenshots.

The lab score weights Total Blocking Time and LCP most heavily. A 95 lab score with failing field data means your real visitors are still waiting.

Speed test results explained: the numbers worth acting on

If your site gets fewer than a few thousand visits a month the report may show no field data at all. Then use the lab numbers plus a real test: open the page on a 12,000-rupee Android phone with Wi-Fi off and time it with a stopwatch.

LCP fix: getting the largest element on screen in under 2.5 seconds

PageSpeed Insights names the LCP element in the diagnostics section; on most business sites it is the hero image or a full-width banner. LCP time is made of four parts: the server response, the delay before the browser discovers the image, the download, and the render. Fix them in that order.

Image compression and WebP

The most common LCP problem on Indian sites is a 2,000-pixel-wide JPEG at 95% quality, weighing 400 KB to 1.2 MB, served to a 390-pixel phone screen. Resize the image to the largest size it is displayed at, export to WebP at quality 75 to 80, and provide a smaller version for mobile through srcset. A typical hero drops from 450 KB to 60 to 90 KB with no visible difference. Add a preload hint and a high fetch priority to the hero so the browser requests it before the CSS finishes parsing.

Lazy loading images: what to lazy load and what to leave alone

Lazy loading makes the browser download an image only when it is about to scroll into view, and it is right for every image below the fold. The mistake that costs sites their LCP score is lazy loading the hero itself. Many themes and page builders lazy load every image by default, including the first one, which delays the largest element by half a second or more. Exclude any image visible on first load.

Font loading speed

Web fonts are a hidden LCP cost because text does not paint until the font arrives, and a heading is often the LCP element on text-heavy pages. Self-host fonts as WOFF2, preload the one weight used in the heading, set font-display to swap, and cap the site at two families and three or four weights. A remote font service adds a DNS lookup and TLS handshake, which on 4G costs 150 to 300 milliseconds before a glyph downloads.

CLS fix: stopping the page from jumping while it loads

Layout shift happens when something appears late and pushes content down. The usual offenders are images without width and height attributes, a cookie or offer banner injected at the top after load, a font that swaps and changes line lengths, and carousels that resize on each slide.

• Give every image and video explicit width and height, or a CSS aspect-ratio, so the browser reserves the space before the file arrives.
• Put banners, notices and ad slots in a container with a fixed minimum height, or show them as an overlay that does not push content.
• Match the fallback font's size to the web font with size-adjust so the swap does not reflow paragraphs.
• Animate with transform and opacity instead of changing top, height or margin.

INP fix: making taps respond within 200 milliseconds

INP replaced First Input Delay in March 2024 and it is a stricter test. It records every tap, click and key press during the visit and reports close to the slowest one.

The cause is almost always JavaScript holding the main thread. On a budget phone a 1 MB script bundle takes two to three seconds to parse and run, and any tap during that window waits. Break long tasks into smaller pieces, defer scripts not needed for the first screen, and remove sliders and animation libraries that run on every page.

Third-party scripts that slow a site

Count the external scripts on a typical Indian business site and you will find a tag manager container with 15 tags, two analytics tools, a chat widget, a heatmap recorder, a social pixel and an embedded map. Together they are often 400 to 900 KB of the page. Delete anything nobody has looked at in six months, load chat and heatmap tools only after the first interaction, and replace the live map embed with a static image that links to Google Maps. A trimmed tag container alone often moves INP from orange into green.

Web hosting India speed: where the server sits matters

Time to First Byte is the floor under every other metric. A request from Rajkot to a shared server in the United States spends 250 to 350 milliseconds in transit before the server does any work. The same site on a virtual server in a Mumbai, Pune or Hyderabad region answers in 20 to 60 milliseconds of transit plus whatever the application needs. Add full-page caching and Brotli compression and TTFB usually lands under 300 milliseconds.

CDN for India: when you need one

A content delivery network keeps copies of your images, scripts and sometimes whole pages on servers near the visitor. The major networks have edge locations in Mumbai, Chennai, Delhi, Hyderabad and Bengaluru, so a visitor in Surat is served from nearby rather than from your origin. If your server is already in Mumbai and all your customers are in Gujarat, a CDN gives a modest gain on static files plus caching and attack protection. If you sell to buyers in the US, Europe or the Gulf, or your server is abroad and cannot move, a CDN is the cheapest single improvement available.

WordPress speed versus Next.js speed

Platform choice is less decisive than people think, but the failure patterns differ. A WordPress site passes Core Web Vitals when it runs a lean theme, a page cache, automatic image conversion and fewer than ten plugins. It fails when a visual page builder, 30 plugins, a home page slider and five font weights pile up.

A Next.js site generates pages ahead of time, serves static HTML and splits code so each page loads only what it needs. Its problems come from the other direction: too much client-side JavaScript hydrating on load, which hurts INP on budget phones, and hosting outside India, which hurts TTFB. A well-built Next.js site typically lands at an LCP of 1.2 to 1.8 seconds on 4G where the equivalent WordPress site lands at 2.5 to 4 seconds. If you are choosing between them for a rebuild, read our Next.js versus WordPress for SEO comparison.

What we have seen on Rajkot and Gujarat websites

Over the last two years we have run speed audits on more than sixty sites for businesses in Rajkot, Morbi, Jamnagar and Ahmedabad, and the first finding is always the office broadband illusion: the owner sees no problem on a leased line and a flagship phone while Search Console shows 70% of mobile URLs failing.

A bearing manufacturer near Rajkot had a 4.1 MB home page with a 1.1 MB hero, eleven font files and a slider plugin, and a field LCP of 6.8 seconds. After resizing and converting images, cutting fonts to two weights, replacing the slider with a static banner and moving from shared hosting abroad to a Mumbai virtual server, field LCP settled at 2.1 seconds in the next 28-day window. A ceramic tile exporter in Morbi had a 1.4-second TTFB from Rajkot because the server sat in a European data centre; moving the origin to Mumbai and adding a CDN brought it to about 180 milliseconds for Indian visitors and improved European load time as well. A hospital site in Rajkot failed INP at 620 milliseconds with 34 plugins and two chat widgets; removing 19 plugins and deferring the chat tool brought it to about 170 milliseconds.

How to decide what to fix first

1. Open Search Console's Core Web Vitals report and note which metric fails and on how many URLs. That is the only scoreboard that matters.
2. Check TTFB. Above 800 milliseconds, fix hosting and caching before anything else.
3. Find the LCP element and shrink it: resize, WebP, preload, never lazy load it.
4. Cut fonts to two families, self-host them and preload the heading weight.
5. Audit third-party scripts and delete or delay everything not needed for the first screen.
6. Add dimensions to images and fixed heights to banners to settle CLS.
7. Profile the slowest interaction on a budget phone and cut the JavaScript behind it for INP.
8. Wait a full 28-day window and check the report again before declaring victory.

From our projects, an image, font and script clean-up is one to three days of developer time; a rebuild makes sense only when the theme or builder is the root problem.

Related reading and next step

• SEO audit checklist for a website in 2026
• Next.js vs WordPress for SEO
• How long SEO takes to show results
• Website development services
• SEO services in Rajkot

If you want a Core Web Vitals audit of your site with a fixed quote for the fixes, send the URL through the contact page or on WhatsApp and we will reply with the field data, the three biggest causes and the cost to clear them.