Home ›Insights ›Performance
PerformanceEngineeringSeptember 24, 202612 min read

Application Performance Optimisation: What Actually Moves Revenue in 2026

Portent's analysis put conversion at 3.05% for pages loading in one second and 0.41% at five seconds. Seven times the revenue from the same traffic and the same page. Yet most performance work targets a lab score that has almost no relationship to what Google actually measures — which is why teams optimise for weeks and watch Search Console refuse to move.

VE
Vikgol Engineering Team
Performance & Cloud Engineering · Vikgol
Share
Load Time vs Conversion RateSame traffic. Same page. Seven times the revenue.CONVERSION RATE BY LOAD TIME · PORTENT1 second3.05%2 seconds1.68%3 seconds1.12%5 seconds0.41%CORE WEB VITALS THRESHOLDS · 75TH PERCENTILE, FIELD DATALCPunder 2.5s · loadingusually a server problemINPunder 200ms · responsivenessusually a JavaScript problemCLSunder 0.1 · visual stabilityusually a missing dimensionYour PageSpeed score is a simulation. Google ranks on what real users experienced.VikgolBelieve In Doers

There is a conversation that happens in almost every performance project. Someone runs PageSpeed Insights, gets a score in the nineties, and cannot understand why Search Console still reports the site as failing Core Web Vitals.

Both numbers are correct. They are measuring different things, and only one of them affects your ranking.

PageSpeed Insights gives you a lab score — a simulation on a throttled connection and a simulated device. It is useful for diagnosis and useless as a target.

Google ranks on field data. The Chrome User Experience Report collects what real users on real devices and real networks actually experienced, over a rolling 28-day window, evaluated at the 75th percentile. That means three-quarters of your visitors must have a good experience — including the ones on a mid-range Android phone on patchy mobile data.

A site can score 95 in the lab and fail in the field. This happens constantly, and it is the single most common reason performance work appears not to work.

📌 Quick answer: what should we actually measure?

Three metrics, all at the 75th percentile of field data: LCP under 2.5 seconds, INP under 200 milliseconds, CLS under 0.1. Get them from Search Console's Core Web Vitals report or the CrUX API — not from a PageSpeed run. Use lab tools to find out why a metric is bad, and field data to know whether it is bad.

What Performance Is Actually Worth

7.4×
Conversion difference between a 1-second and 5-second page: 3.05% against 0.41%
Portent
53%
Of mobile visits abandoned when a page takes longer than 3 seconds
Google
~1%
Conversion lost for every additional 100ms of latency
Amazon

Those are correlational figures, which is worth saying plainly — fast sites tend to be well-built sites, and well-built sites convert better for reasons beyond speed alone.

The more persuasive evidence comes from controlled tests published through Google's own case study programme, where performance was the only variable changed:

CompanyWhat they changedMeasured result
Rakuten 24LCP optimisation, run as an A/B test+53.4% revenue per visitor, +33.1% conversion rate
Vodafone ItalyLCP improved 31% via server-side rendering and reduced render-blocking JavaScript+8% sales
RedBusINP optimisation — responsiveness, not load time+7% sales
Ray-BanSpeculation Rules API to prerender key pages+101.5% mobile conversion on product pages

The RedBus result is the one most teams overlook. They improved responsiveness after load, not load time — and it moved sales by 7%. INP is the metric most sites are quietly failing while they focus entirely on LCP.

The Lab-Versus-Field Trap

⚠️ Lab data
PageSpeed Insights, Lighthouse
  • · Simulated device and throttled network
  • · Single run, controlled conditions
  • · Results appear immediately
  • · Excellent for diagnosis
  • · Has no effect on ranking
  • · Can pass while real users suffer
✅ Field data
CrUX, Search Console
  • · Real users, real devices, real networks
  • · Rolling 28-day window, 75th percentile
  • · Takes 4–8 weeks to reflect a fix
  • · Tells you whether a problem exists
  • · This is what Google ranks on
  • · Cannot be gamed by a good test run
⚠️ The timing problem this creates

You deploy a fix. Lab scores improve within minutes. Search Console does not move for four to eight weeks, because the 28-day rolling window has to fill with post-fix data. Teams that do not know this conclude the work failed and revert it — or worse, keep optimising a metric that was already fixed. Set the expectation before the work starts, not when someone asks why the dashboard is unchanged.

Most LCP Problems Are Server Problems

The standard LCP advice is about images: compress them, use WebP, lazy-load below the fold. All correct, and for many sites it is not where the time is going.

LCP decomposes into four parts, and it is worth measuring each before optimising any:

PART 01 — OFTEN THE LARGEST
Time to first byte
How long your server takes to respond at all. Slow database queries, uncached API calls, cold serverless starts, a server in the wrong region. No amount of image optimisation fixes a 900ms TTFB.
PART 02
Resource load delay
The gap between the HTML arriving and the browser starting to fetch the LCP element. Usually caused by render-blocking CSS and JavaScript in the head, or by the hero image being discovered late.
PART 03
Resource load time
Actually downloading the image. This is the part everyone optimises — format, compression, dimensions, CDN. Real, but often the smallest slice of the four.
PART 04
Element render delay
The browser has the resource but has not painted it — blocked by a long JavaScript task, a font that has not loaded, or a hydration pass that has to finish first.
✅ Where full-stack visibility pays off

This is the practical argument for treating performance as an application problem rather than a frontend one. If TTFB is the largest part of your LCP, the fix is a database index, a cache layer, a regional deployment — infrastructure work, not asset optimisation. Teams where frontend and infrastructure sit in separate silos frequently spend a quarter compressing images while the real cost sits in an unindexed query.

INP: The One Most Sites Are Failing

INP replaced First Input Delay in 2024 and is considerably harder to pass. FID measured only the delay before the browser could begin processing the first interaction. INP measures the full latency from input to visual response, across every interaction in the visit.

A site can load fast and still fail INP badly. The usual causes:

  • Long tasks on the main thread — anything over 50ms blocks the browser from responding to input at all
  • Heavy hydration — the page looks interactive several seconds before it actually is
  • Third-party scripts — analytics, chat widgets, session recording and tag managers competing for the same single thread
  • Unnecessary re-renders — a state change that re-renders half the component tree on every keystroke
  • Synchronous work in event handlers — doing the expensive thing before yielding back to the browser

The fix is usually the same idea applied in different places: break long tasks up, yield to the main thread, and defer anything that does not need to run before the first interaction.

The Thresholds, and What Each One Really Means

MetricGoodNeeds workPoorUsual root cause
LCPunder 2.5s2.5–4sover 4sServer response time, render-blocking resources
INPunder 200ms200–500msover 500msLong main-thread tasks, third-party scripts
CLSunder 0.10.1–0.25over 0.25Images without dimensions, injected banners, late fonts

All three are evaluated at the 75th percentile of field data. Passing on your own laptop means nothing — your laptop is not the 75th percentile of your users.

A 90-Day Sequence

1

Get field data before touching anything

Search Console's Core Web Vitals report, segmented by mobile and desktop. Mobile is almost always worse and almost always where the revenue is. If you have no field data, install a real-user monitoring script and wait — optimising blind wastes the quarter.

Week 1
2

Rank pages by traffic times revenue, not by score

A page scoring 40 that nobody visits is not the problem. Your highest-traffic conversion pages scoring 65 are. Sort by business impact and fix the top three templates — most sites have fewer distinct templates than they assume.

Week 1
3

Decompose LCP before optimising it

Measure all four sub-parts. If TTFB is 800ms, start with the backend — caching, query optimisation, regional deployment. If the image is the bottleneck, then compress. Doing this in the wrong order is how performance projects deliver nothing.

Week 2–4
4

Audit third-party scripts honestly

List every one, what it does, and who asked for it. Most sites carry several nobody can account for. Each is a main-thread competitor and an INP contributor. Removing two dead tags often beats a week of code splitting.

Week 3
5

Fix INP with task breaking, not rewrites

Find long tasks in the Performance panel. Break them up, yield to the main thread, defer non-critical work past first interaction. This is usually a set of surgical changes rather than an architectural overhaul.

Week 4–8
6

Set a performance budget and enforce it in CI

Performance regresses by default — every feature, script and font adds weight. A budget checked on every pull request is the only thing that holds the gain. Without it you will repeat this whole exercise in eighteen months.

Week 8+

Frequently Asked Questions

Why does my PageSpeed score say 95 while Search Console says we fail?
They measure different things. PageSpeed Insights runs a lab test — a simulated device on a throttled connection, one run, controlled conditions. Search Console reports field data from the Chrome User Experience Report: real users, real devices, real networks, over a rolling 28-day window at the 75th percentile. Field data is what affects ranking. Lab data is a diagnostic tool. A site can pass in the lab and fail in the field, which is extremely common on sites with heavy JavaScript or a mobile-skewed audience.
How long before performance fixes show up in Search Console?
Four to eight weeks. The CrUX dataset uses a rolling 28-day window, so it has to fill with post-fix data before the report moves, and updates land weekly. Lab scores improve immediately, which creates an expectation gap. Set this out before the work starts — teams that do not know it frequently conclude a successful fix did not work.
What is the single highest-impact fix for most sites?
There is no universal answer, which is why the decomposition step matters. That said, two things recur: server response time on sites with dynamic content, and third-party scripts on marketing sites. If you have to guess, measure TTFB first — if it is above 600ms, everything downstream is paying for it, and no amount of image work will compensate.
Do we need to rebuild on a modern framework?
Usually not as a first step. Framework-based platforms do pass Core Web Vitals at higher rates than traditional CMS platforms — Next.js around 58% against WordPress at 38% — largely due to code splitting, image optimisation and server-side rendering being built in. But a rebuild is a six-month project, and most sites can move from failing to passing with image optimisation, script removal, deferred JavaScript and caching. Exhaust those first. Rebuild when the architecture is genuinely the constraint, not because the score is red.
Is INP really worth separate attention from LCP?
Yes, and it is the more commonly neglected of the two. RedBus improved INP alone and saw a 7% increase in sales. A page can load quickly and still feel broken if interactions lag — and INP measures every interaction across the visit, not just the first. Sites heavy on client-side rendering frequently pass LCP and fail INP, because the page looks ready several seconds before it can actually respond.
How do we stop performance regressing after we fix it?
A performance budget enforced in CI, and real-user monitoring in production. Performance degrades by default — every new feature, tag and font adds weight, and nobody notices until it is bad again. A Lighthouse CI check that fails a pull request when LCP or bundle size exceeds a threshold is a small amount of setup that prevents repeating the entire exercise in eighteen months.

Not Sure Where Your Time Is Going?

We do full-stack performance work — frontend, backend and AWS infrastructure. Book a free 30-minute call and we'll look at your field data and tell you honestly where the bottleneck actually is.

#WebPerformance#CoreWebVitals#LCP#INP#PageSpeed#AWS#Engineering#Vikgol
VE
Vikgol Engineering Team
Performance & Cloud Engineering · Vikgol
The Vikgol engineering team has shipped 90+ AI, web and cloud projects for startups and enterprises across US, UK, UAE and India. We treat performance as a full-stack problem — frontend, backend and infrastructure — because that is usually where the time actually goes.
Available Now · 72-Hour POC

Ready to Ship Your AI Product?
Let’s Build It Together.

Senior engineers on demand. Working prototype in 72 hours. NDA before we discuss anything. 100% code ownership to you — no lock-in, ever.

70+
Senior engineers on staff
90+
Projects delivered globally
72h
Working POC guaranteed
5★
Client satisfaction rating