What PageSpeed score do you need for an ad landing?

A useful guide is about 90+ on mobile and green LCP, INP and CLS. It is not a hard law and not a lead guarantee, but a calm target before scaling campaigns. If vitals are green and the score is 80–85 — you can often work while polishing the rest in parallel.

What matters more — the 0–100 score or Core Web Vitals?

The vitals matter more. The score is a lab/field summary. Watch LCP, INP and CLS: they are closer to what people feel on a phone. A red LCP with a “pretty” score still means the main content appears late.

Where should I start if the site is slow?

With the checklist: PageSpeed on mobile, then first-screen images (WebP/AVIF compression), deferred chats and analytics, and a Slow 4G test. Write down 2–3 causes with the biggest effect, not the whole Opportunities list.

Is a WordPress cache plugin enough?

Often no. If the first screen has heavy photos and many plugins, cache alone will not fix it. You need content changes (image size and format) and script cleanup. Cache helps on repeat visits, but it does not replace hero-block optimization.

Does speed affect Google Ads and SEO?

Yes, gently: a slow landing less often reaches a lead (you already paid for the click), and weak Core Web Vitals get in the way in search when other factors are equal. Speed does not replace the offer or content, but it removes extra friction between interest and action.

When is optimization enough, and when should I change platforms?

If after compressing images, deferring scripts, cleaning plugins and adding a CDN the metrics look fine — keep the platform. If the mobile score/vitals stay weak with ongoing ads — a clean fast build is wiser than more plugins on top of a builder ceiling.

Lab is green but field is red — which should I trust?

Field (real user data), when it exists. Lab is a controlled test; field reflects real networks, devices and geography. In Uzbekistan the gap often comes from 4G and distant hosting.

Web development18 minMay 29, 2026

Site Speed and PageSpeed in 2026: What a Business Owner Should Check

Robert MirovUpdated Sep 5, 2026

If you already have a site or landing page in Tashkent, Samarkand, or the regions, the question is usually: “Why do people leave on mobile?” Often it is not a “weak offer” — the page opens slowly, buttons feel stuck, and the form block jumps at the moment of a tap. PageSpeed (PageSpeed Insights — Google’s free page-speed report) and Core Web Vitals (three key “health” metrics for loading) are not a beauty contest of scores for developers. They show what a client feels on 4G: when the main content appears, how quickly the site responds to a tap, and whether the layout jumps. In this article UZNEO calmly covers: why speed matters for business, what LCP, INP and CLS mean in plain words, how to run diagnostics without panic, what usually slows sites in Uzbekistan, and which plan to choose — for a day, a week, or a serious rebuild. No promises that “leads will jump by N%” and no scary percentages — only a clear order of actions.

★ TL;DR

▸Speed affects leads, ads, SEO and trust — especially on mobile 4G in Uzbekistan.
▸PageSpeed is Google’s report; more important than the 0–100 score are three Core Web Vitals: LCP, INP, CLS.
▸LCP — when main content appears (good ≤ 2.5 s); INP — response to a tap (≤ 200 ms); CLS — no layout jumps (< 0.1).
▸Diagnostics: PageSpeed → mobile tab → vitals → Opportunities → Slow 4G → 2–3 main causes → re-test after fixes.
▸In UZ the usual blockers are heavy photos, WordPress/Tilda clutter, third-party scripts, distant hosting without CDN, and heavy fonts.
▸Plan: day — images and widgets; week — plugins, CDN, CLS, ad landings; refactor — when basics are done and mobile is still weak.
▸Speed the page first, then scale ads; one cache plugin rarely saves a heavy first screen.

01Why site speed matters for a business owner

Speed is not a vanity metric for programmers. It affects whether someone reaches the text, the price, and the lead form. In practice, speed ties to four things: 1. Leads. On mobile in Uzbekistan, clients often open a site between tasks. Four to six seconds of a white screen — some people close the tab before reading the offer. 2. Ads. In Google Ads and other systems you pay per click. A slow landing page (the page an ad leads to) lowers the chance of reaching the form: the click is already billed, the lead is not. 3. Search (SEO). Google uses Core Web Vitals among other signals. Speed alone will not “put you on top”, but weak metrics get in the way when content is already comparable to competitors. 4. Trust. Several seconds of an empty screen on 4G in Tashkent feel like “the site is broken” — even if everything looks fine after eight seconds.

Before	After
Checking the site only on office Wi‑Fi	Testing the same page on a phone / Slow 4G
A “pretty” layout with heavy uncompressed photos	A readable first screen in 2–3 seconds first
Ads running, form “somewhere below”	Fast first screen + a form that does not jump
PageSpeed score “for the checkbox”	LCP, INP, CLS — what the client feels

You can check for free: PageSpeed Insights — paste the full URL and open the mobile tab first. The mobile report is closer to a real client in UZ.

02What PageSpeed and Core Web Vitals are

Short glossary — so the report is readable without confusion. PageSpeed Insights is Google’s tool: you paste a page URL and get a score plus recommendations. The 0–100 score is a summary guide (lab and/or field). Useful — but not the only truth. Core Web Vitals are three metrics Google treats as key to user experience: • LCP (Largest Contentful Paint) — when the main content appears on screen. • INP (Interaction to Next Paint) — how quickly the page responds to a click/tap. • CLS (Cumulative Layout Shift) — how much the layout jumps during load. Reports often also show: • Lab data — a Google lab test (controlled conditions). • Field data — real Chrome user data (when there is enough). If field is red and lab is green — trust field. • TTFB (Time To First Byte) — how long until the first server response. High TTFB often means distant hosting or a weak server. • Opportunities / Diagnostics — hints: huge images, spare JavaScript, slow server response. How to read the 0–100 score roughly: • 90+ — a solid level, especially for an ad landing. • 50–89 — room to improve; often images, scripts, fonts. • Below 50 — worth fixing before scaling ads, or in parallel with them. One red metric in Google Search Console (Core Web Vitals) can drag attention across a whole site template — not only one page.

03LCP, INP and CLS in plain words

Below — Google’s thresholds and a business-language translation.

Metric	What a person feels	Good	Needs improvement	Poor
LCP	When the main thing is visible (headline, photo, offer)	≤ 2.5 s	2.5–4 s	> 4 s
INP	Button / menu / form “woke up” after a tap	≤ 200 ms	200–500 ms	> 500 ms
CLS	Text and buttons do not jump under the finger	< 0.1	0.1–0.25	> 0.25

LCP — when the main content appears Until LCP happens, people stare at emptiness, a spinner, or a skeleton. On a landing, LCP is often a large headline or the hero image (main first-screen photo). A 3–5 MB phone photo without compression is a classic cause of poor LCP in UZ. INP — how quickly the page responds to an action Someone taps “Leave a request”, opens a menu, or focuses the phone field — and waits for a reaction. A long delay makes the site feel stuck, even if images have loaded. A common cause is heavy JavaScript: chats, pixels, animations, builder widgets that all start at once. CLS — whether the layout jumps Text shifts, a button slides down, a banner or chat suddenly appears at the top — that is CLS. On a phone it is easy to miss the button you meant to tap. Fix it by reserving sizes for images/banners and loading widgets carefully.

Before	After
Looking only at “score 73”	Seeing which of the three metrics is red
Hero photo 4 MB JPEG from a phone	WebP/AVIF, screen-sized, compressed
Chat and 5 pixels start immediately	Deferred until interaction or after LCP
Banner with no height in the layout	Reserved width/height → less CLS

UZNEO’s rule: fix the red vital first, then chase a “pretty hundred”.

04PageSpeed diagnostics checklist

Follow the steps so a long report does not overwhelm you. 1. Open PageSpeed Insights and paste the full URL of the page you care about (not only the homepage — also check the ad landing). 2. Start with the mobile tab. Most clients in Uzbekistan browse on phones. 3. Note the summary score, then go straight to LCP, INP and CLS — they matter more than a “nice” number. 4. Compare lab and field (when field exists). Field is real users; lab is the lab. On conflict, trust field. 5. Open Opportunities / Diagnostics. Typical hints: oversized images, unused JavaScript, reduce unused CSS, reduce initial server response time. 6. Test the same page on Slow 4G in Chrome DevTools (Network → throttling). Office Wi‑Fi often lies: “everything flies for us”. 7. Write down 2–3 main causes, not a full list of 15 items. Usually: first-screen images → third-party scripts → server response / CDN. 8. Re-run the same URL after fixes. Compare the same metrics (LCP in seconds, INP in ms, CLS), not only the score. 9. If you have Search Console — cross-check the Core Web Vitals report: sometimes a whole site template is broken, not one page.

Before	After
One run “for the checkbox”	Before / after on the same URL
Fixing everything in Opportunities	2–3 causes with the biggest effect
Testing only from a laptop in the office	Mobile report + Slow 4G
Panic over score 48	Plan: day → week → platform question

Diagnostics take 15–20 minutes. That is enough to see: fix content and scripts — or already think about changing the stack.

05Common causes of a slow site in Uzbekistan

In UZ the blockers are usually simple and familiar — not “Google algorithm magic”. 1. Heavy photos. Phone shots of 3–8 MB without compression on the first screen. You need WebP or AVIF, a sensible screen size, and lazy-load (deferred loading) for images below the fold. 2. WordPress with a plugin “zoo”. Easy to turn features on, but each plugin brings scripts and styles. The page piles up. More in the CMS comparison. 3. Tilda and other builders. Easy to assemble a landing. Harder not to overload it with dozens of blocks, animations, video backgrounds and widgets. On mobile 4G you feel it immediately. 4. Third-party scripts. Live chats, Meta/Google pixels, maps, booking widgets, pop-up promos — if they all start at the beginning of load, LCP and INP suffer. 5. Hosting far from the audience. A server in Europe or Asia without a CDN (Content Delivery Network — nearby edge points for static files) → high TTFB for people in Tashkent and the regions. 6. No CDN for images and files. Everything comes from one distant server. 7. Fonts. Five Google Fonts weights blocking text paint. One or two weights and `font-display: swap` are enough. On Next.js and a proper CDN, a landing often lands in the green zone on mobile — if you do not overload the first screen with sliders and heavy JS.

Before	After
Hero 5 MB, chat + 4 pixels at once	Compressed WebP + deferred widgets
15 plugins “just in case”	Only what this page needs
Server in the EU, no CDN	CDN / hosting closer to the UZ audience
6 fonts and a video background on the first screen	1–2 weights, a strong static offer

06Improvement plan: day, week, refactor

You do not need to do everything at once. Pick a horizon that matches the job and your attention budget.

Horizon	Goal	Typical actions	When it is enough
1 day	Remove crude blockers	Images, lazy-load, defer chat, fonts	Score rises, vitals turn yellow/green
1 week	Systemic cleanup	Plugins, pixels, CLS, CDN, ad landings	Stable mobile result
Refactor	Lift the platform ceiling	Lighter stack, clean first screen, hosting+CDN	Basics done, mobile still weak

In 1 day — quick wins without changing platforms: • Compress first-screen images (Squoosh, TinyPNG) to WebP or AVIF. • Enable lazy-loading for images below the fold. • Remove or defer chat and spare widgets until user interaction. • Keep one or two font weights; enable `font-display: swap`. • Test Slow 4G and re-run PageSpeed. In a week — systemic improvements: • Review plugins / blocks: what this page actually needs. • Defer analytics and pixels after the main load (or until consent/interaction — per your policy). • Reserve image and banner sizes to reduce CLS. • Add a CDN for static files if the server is far from Uzbekistan. • Check key ad landings, not only the homepage. Serious refactor — when tweaks are no longer enough: • Rebuild the landing on a lighter stack (turnkey landing). • Drop a heavy builder or a plugin “zoo”. • Design the first screen without spare JS and huge sliders. • Plan proper hosting and CDN from the start. Do the day first, then the week. Move to a refactor when the basics are done and mobile metrics are still weak — especially if ads run continuously.

07When optimization is enough vs when to change platforms

Not every slow site needs a full rewrite. And not every site is “saved” by another cache plugin. Optimization is usually enough if: • After image compression and deferred scripts the score and vitals clearly rise. • LCP, INP and CLS are already in the green or a stable yellow zone. • The site is a brochure, catalog, or occasional campaigns — not constant ad traffic. • The platform (WordPress, Tilda, etc.) still works for editing content. It makes sense to change platforms or rebuild if: • Basic fixes are done and the mobile score / vitals stay weak for a long time. • Ads run continuously and the landing stays slow on 4G. • Every new “optimization” pulls in another plugin or workaround. • The builder is easy to edit but cannot deliver the speed you need on phones at your block volume.

Before	After
Another cache plugin on top of a heavy hero	Content and first-screen scripts first
“Rewrite everything from scratch” with no day fixes	Day → week → stack decision by numbers
Ads ↑, same slow landing	Speed the landing before/while raising bids
Convenient builder at any cost	Easy editing + mobile speed

Short version: first squeeze clear gains from the current platform. If you are near the ceiling and the site is a sales tool, calmly plan a fast clean build. More on the stack in the WordPress, Tilda and Next.js comparison.

08Speed, ads and SEO: how they connect

A fast landing does not replace the offer, price, or ad. But it removes extra friction between the click and the lead. Ads. In Google Ads the landing matters: a slow page less often gets people to the form. It makes sense to speed up the landing before or while raising bids — not only to raise bids. The same applies to Yandex: one slow page hurts both systems. SEO. Google looks at Core Web Vitals among other signals. Speed alone will not “put you in the top”, but weak metrics get in the way when texts and links are already comparable to competitors. If the page is slow, content work loses some of its value — see also how to grow in Google. What usually does not help: • A cache plugin alone while the hero still has heavy uncompressed photos. • Waiting for a high score on very cheap hosting with no content changes. • Copying a competitor’s pile of fonts, sliders and video backgrounds on a heavy builder. • “We’ll optimize later” while ads already run to the same page.

Before	After
Ads budget first, speed “later”	A readable mobile first screen first
One cache plugin = “optimization”	Images + scripts + server/CDN
SEO texts on a page with LCP 6 s	Content + vitals in green/yellow

Speed the page first, then scale traffic — calmer for both ads and search.

Summary

Open your site in PageSpeed Insights and check the mobile tab. For an ad landing, a useful guide is a high score and green (or stably yellow with a plan) LCP, INP and CLS. The order is simple: diagnostics checklist → quick day fixes → systemic week work. If mobile metrics are still weak after that and the site supports ongoing ads — calmly plan a platform change or a clean build, not endless workarounds. UZNEO builds sites and landings with solid speed out of the box. Send a URL — we will say what usually slows your page and where it is wiser to start.