Almost every business owner tests their website on office wifi, on a laptop, with the site already cached in the browser. It opens instantly. They conclude the site is fast.
Their customers are on a phone, on mobile data, often with patchy signal, visiting for the first time. That is a completely different experience, and it is the one that decides whether anyone sees your business at all.
Test it the way your customers see it
Do this now, properly:
1. Take your phone and turn wifi off
2. Open an incognito or private tab — this skips the cache
3. Type your website address
4. Count the seconds until you can read something and tap something
If you waited, so do they. Research consistently shows that a large share of mobile visitors abandon a page that takes more than a few seconds — and they do not come back later.
For a second opinion, run your URL through PageSpeed Insights (pagespeed.web.dev) and read the Mobile tab, not Desktop. Desktop scores flatter you and are irrelevant to most Indian traffic.
What Google actually measures
Google’s Core Web Vitals are three numbers. These are the thresholds for “good”:
Metric | What it measures | Good |
LCP (Largest Contentful Paint) | How long until the main content appears | Under 2.5 seconds |
INP (Interaction to Next Paint) | How quickly the page responds when tapped | Under 200 milliseconds |
CLS (Cumulative Layout Shift) | How much the page jumps around while loading | Under 0.1 |
Speed is a ranking factor, but a modest one. The bigger effect is behavioural: slow pages lose visitors, and losing visitors loses enquiries. Fix it for the customers, and the ranking benefit follows.
The causes, in the order they usually matter
1. Images
This is the single biggest cause, and the easiest to fix.
A photo straight from a phone camera can be four or five megabytes. On a mobile connection that is several seconds for one image. A homepage carrying six of them is unusable.
What to do
• Resize before uploading. No image on a website needs to be wider than about 1920 pixels, and most need far less.
• Compress. Tools like TinyPNG or Squoosh typically cut file size by 60–80% with no visible difference.
• Use modern formats. WebP is dramatically smaller than JPEG at the same quality and is supported by every current browser.
• Lazy-load anything below the fold, so images further down do not block the first screen.
• Always set width and height attributes — this stops the page jumping as images arrive, which fixes CLS.
Target: under 200 KB per image, well under 100 KB for most.
2. Too many plugins
Every plugin adds code. On WordPress sites it is common to find thirty or forty installed, half of them unused, each adding stylesheets and scripts to every page.
Go through the list. For each one ask: is this doing something a visitor would notice? If not, deactivate and delete it. Deactivated is not enough — delete.
Common offenders: page builders loading their entire library on pages that do not use it, slider plugins, multiple analytics and chat widgets, security plugins scanning on every request.
3. Hosting
Cheap shared hosting puts hundreds of sites on one server. When another site on that server gets busy, yours slows down, and there is nothing you can do about it.
Two things matter: server location and server quality. If your customers are in Tamil Nadu, a server in India or Singapore responds measurably faster than one in the United States, simply because of distance.
Check your TTFB (Time To First Byte) in PageSpeed Insights. If it is above about 600 milliseconds, the host is the bottleneck and no amount of image compression will fix it.
4. Fonts
Custom fonts are usually loaded from Google Fonts, which means an extra connection and download before any text appears. Three font families at four weights each is twelve files.
• Use one or two families, two or three weights
• Self-host the font files rather than loading them externally
• Add font-display: swap so text appears in a fallback font immediately rather than staying invisible
• For Tamil text, load only the Tamil subset, not the full font
5. Third-party scripts
Chat widgets, analytics, pixels, heatmaps, review widgets, embedded maps, social feeds. Each is code from someone else’s server, loading on your page, outside your control.
Audit them. Remove anything not actively used. Load the rest asynchronously so they do not block the page. An embedded Google Map is particularly heavy — a static image linking to Maps is far lighter and does the same job.
6. No caching or CDN
Caching stores a built copy of the page so the server does not rebuild it for every visitor. On WordPress, a caching plugin alone often halves load time.
A CDN stores copies of your files on servers around the world, so an Indian visitor gets them from a nearby location. Cloudflare has a free tier that covers most small business sites.
A realistic order of work
Step | Effort | Typical gain |
1. Compress and resize all images | Low | Large |
2. Delete unused plugins | Low | Moderate |
3. Enable caching | Low | Large |
4. Add a CDN | Low | Moderate |
5. Fix fonts | Medium | Small–moderate |
6. Audit third-party scripts | Medium | Moderate |
7. Move to better hosting | Medium | Large if TTFB was bad |
8. Rebuild on a lighter theme | High | Large |
Steps 1 to 4 can usually be done in a day and often get a site from unacceptable to acceptable. Only consider step 8 if the rest has been done and the site is still slow.
What slowness costs you
It is worth being concrete about why this matters:
• Visitors leave before the page appears, and are counted as a bounce
• Enquiry forms get abandoned halfway
• Ad spend is wasted — you pay for a click that never became a visit
• Google sees users returning to search results, which signals dissatisfaction
• On a slow connection, a heavy site also burns the visitor’s data
Frequently asked questions
What PageSpeed score should I aim for? Do not chase 100. Mobile scores above about 70, with Core Web Vitals in the “good” range, are a sensible target for a business site. The real test remains your own phone on mobile data.
My site is fast on wifi. Is that good enough? No. Most of your visitors are on mobile data. Wifi testing hides exactly the problems that cost you enquiries.
Does speed really affect Google ranking? Yes, but as one factor among many, and a modest one. Relevance and content quality matter more. Speed’s bigger effect is on whether visitors stay.
Will a faster host alone fix it? Only if the host was the problem. Check TTFB first. If TTFB is fine and the page is still slow, the weight is in images, scripts and plugins.
Is AMP still worth using? It is no longer required for Google News or Top Stories and most sites no longer need it. A well-optimised ordinary page is the better approach now.
How often should I re-test? After any significant change, and once a quarter otherwise. Sites get slower over time as content, plugins and tracking accumulate.


