⚡ Tech
Web Performance
Ecommerce
PageSpeed
Next.js

Vape Website Performance: Inside Cigge's 99/100 Result

I cut redundant browser work across Cigge’s storefront. The October 12 Vercel mobile snapshot shows 99/100, LCP 1.63 s and INP 128 ms—with checkout CLS still to fix.

Uygar DuzgunUUygar Duzgun
Oct 2, 2026
Updated Oct 13, 2026
13 min read
Vape Website Performance: Inside Cigge's 99/100 Result

Cigge scored 99/100 in Vercel's mobile Real Experience Score. I checked its vape website performance using recorded production traffic on October 12, 2026. The selected week's 75th-percentile Largest Contentful Paint was 1.63 seconds. Interaction to Next Paint was 128 milliseconds.

My work covered the loading sequence: first-screen CSS, optional JavaScript, HTML sanitization, repeated catalogue data and cache invalidation. A visitor should be able to read a category page while the browser finishes the less urgent work. Prices, filters and availability still have to behave like a real shop.

The recorded live PageSpeed comparison also showed a specific gain: simulated LCP fell from 4.77 to 4.20 seconds, about 12%, in the tested late-paint group. That percentage covers one comparison. The broader work spans several parts of the storefront, and the October visitor snapshot shows good loading and interaction response in its selected sample.

There is more to fix. Vercel still flagged layout stability, with checkout showing a CLS of 0.32. I keep that result alongside the 99/100 score because a shop needs stable controls as well as fast content.

Cigge is a Swedish retailer with a vape and e-cigarette catalogue, e-liquids and accessories. I work on its technical storefront. Cigge's company page provides the business context; this article covers the engineering work from September 25 to October 2 and the October 12 measurement follow-up.

99/100 on mobile: the October 12 visitor snapshot

On October 12, I checked Cigge's production project in Vercel Speed Insights with Mobile, Last 7 Days and P75 selected. The dashboard displayed October 5–12, 2026; the final day was still in progress. No individual route or country filter was selected.

Largest Contentful Paint: 1.63 seconds, rated Great; the LCP view showed 796 metric events.
Interaction to Next Paint: 128 milliseconds, rated Great; the INP view showed 618 metric events.
Cumulative Layout Shift: 0.1 as displayed, rated Needs Improvement; the CLS view showed 432 metric events.
Real Experience Score: 99/100, Vercel's summary score for the selected mobile production view.
Cigge mobile results: lab LCP 4.77 to 4.20 s and PageSpeed 72 to 79; separate Vercel visitor score 99/100
Cigge mobile results: lab LCP 4.77 to 4.20 s and PageSpeed 72 to 79; separate Vercel visitor score 99/100

Source: the historical September 30 live PageSpeed comparison described in this article, and the Vercel mobile production snapshot checked October 12. The lab improvement applies to the late-paint subset; the newer visitor figures have a separate reporting scope.

These are recorded visitor measurements. P75 means the value at the 75th percentile, rather than an average. The event counts belong to each metric; they are not unique visitor counts. Vercel explains collection and scoring in its Speed Insights metrics guide.

The 99/100 score covers the selected sample as a whole. Vercel displayed CLS as 0.1, but labelled it Needs Improvement and stated Above 0.1. That rounded value is not precise enough to call a pass. Checkout showed 0.32 across 86 CLS events, so I have a specific flow to investigate.

The earlier 237 ms INP came from Chrome UX Report data for September 2–29. Vercel's 128 ms comes from October 5–12 and different collected traffic. I cannot turn that difference into a controlled improvement percentage or assign it to one deployment. The selected Vercel week does show good mobile loading and response; checkout layout shifts remain visible.

This follow-up adds production visitor data to the original lab comparison. It does not replace the category-specific CrUX figures, repeat the PageSpeed experiment or establish a ranking or conversion gain.

The overhaul: make the browser do less

Cigge uses a Next.js storefront with PrestaShop behind the catalogue. That split lets me change what the browser receives while keeping the shop's business data in its existing backend.

A category page has to show product information, keep prices and availability correct, and support filters, sorting and back navigation. A quick first paint is useful only if those controls keep working. I treated the work as a series of scoped changes with regression checks.

The starting problem was how much work we handed the browser before a visitor could use the page. I traced the HTML, first-screen styles and client bundles, then removed or deferred work the route did not need immediately.

Nearly 1 MB of HTML: where the investigation started

The September 28 public homepage payload audit recorded 987,836 bytes of HTML. Inside that document, the embedded React Flight stream accounted for 492,120 bytes. Flight is the serialized data React uses to carry server-rendered component information to the browser.

The same audit identified 270,853 bytes of page-builder initial state, 96,628 bytes of translation messages and 47,640 bytes of bootstrap data.

Cigge homepage payload snapshot from September 28, 2026, showing overlapping document and serialized-data sizes
Cigge homepage payload snapshot from September 28, 2026, showing overlapping document and serialized-data sizes

Source: our recorded public-page payload audit. These are serialized byte counts, not compressed network transfer sizes. The categories overlap: the Flight stream sits inside the document, and expanded component data can overlap within Flight. Do not add the bars together. This is a diagnostic snapshot, not a before-and-after result.

Those numbers helped me ask more useful questions. Does this route need the full brand list immediately? Does a product card need product-detail fields? Are we copying identical pricing rows into multiple variants?

Each answer gives me a smaller, specific change to make and verify. The payload snapshot identifies where to look; it does not establish how many bytes the finished work removed.

Critical CSS: get the first screen visible

I separated the CSS needed for the first visible screen from the larger storefront style profiles. The server includes the critical subset in the HTML, so the browser can style that screen without waiting for the full profile stylesheet.

The later category-page changes defer the larger profile until the first contentful paint. That means the browser can display the content before requesting styling that the initial screen does not need.

Client-side navigation still needs the right styles. Background tabs need a loading path that does not depend on a paint event. Visitors with JavaScript disabled need a fallback. I kept those cases in the implementation and its checks.

There is a trade-off: on slower connections, styling farther down the page can arrive later. Critical CSS has to cover the actual first viewport, and an early scroll deserves testing. Deferring a stylesheet without checking that boundary would risk exchanging a loading delay for a visibly unfinished page.

Stop optional JavaScript competing with the first paint

On the category page we investigated, the Largest Contentful Paint element was description text, not a product image. Compressing another image would not address that particular bottleneck.

I moved the recent-category recorder into a separate, deferred client chunk. It still records the visit for the quick-access menu, but it does not need to compete with the first visible content.

The header and footer logos also moved to same-origin image paths through the existing image proxy. The image content stayed the same; the request no longer needed a separate storefront-to-CDN origin connection for those assets.

I also kept home-only client code on the homepage instead of loading it through category and product routes. Search and catalogue interactions remain available where visitors need them. The aim was to make each route carry its own responsibilities.

15.4 KB less initial JavaScript on the category route

Another change moved storefront HTML sanitization to server producers. Previously, public client bundles included DOMPurify to sanitize content the server also rendered.

The recorded build comparison found about 15.4 KB less initial JavaScript, compressed with Brotli quality 3, on the Cigge category route. That is a bundle measurement, not another percentage improvement in loading time.

The implementation keeps sanitization on the server and sends sanitized fields to the rendering components. Administrator editors can still load the sanitizer when editing requires it. A build guard checks that the library does not return to ordinary storefront client chunks.

The server still has to sanitize untrusted HTML. I traced the content producers, dynamic responses and browser-storage paths before moving that responsibility. The review record includes those checks, rendered-markup comparisons and interactive flows.

Cut duplicated catalogue data and scope cache purges

The week also included a smaller client catalogue projection. Some variant records repeated the same product-level pricing rows. The implementation removes identical copies while retaining variant-specific differences and the full records where product detail needs them.

Brand data follows a similar principle. The storefront no longer needs to include the full manufacturer list in the initial bootstrap for every page. Search and brand controls request it when needed, with shared loading and retry handling.

Variant-specific prices and fields stay intact. Product detail still receives the full records it needs. The smaller projection removes identical copies while keeping the data customers need to choose a variant.

Cache work followed the same logic: invalidate the content that changed, avoid rebuilding unrelated pages, and warm selected catalogue routes after deployment. Stock-sensitive information still needs its own correctness rules. A warm cache is useful; stale availability is a defect.

PageSpeed results: the improvement I can quantify

How the comparison ran

The recorded comparison used PageSpeed Insights mobile tests against two preview environments on the same code base and staging data: a control and the changed version. It ran eight alternating rounds per store, with at least 75 seconds between requests to the same URL.

The test record separated runs by when the observed first paint happened relative to the document finishing. The graph below shows only the late-paint group, with four to five runs per side. It does not summarize all eight rounds.

Cigge category-page mobile simulated LCP comparison: control median 5.620 seconds and changed preview median 5.176 seconds in the late-paint subset
Cigge category-page mobile simulated LCP comparison: control median 5.620 seconds and changed preview median 5.176 seconds in the late-paint subset

Source: our recorded preview comparison. Lower is better. The difference between these medians is 444 milliseconds, or approximately 7.9%. These are simulated lab timings for the late-paint subset, not production visitor measurements.

The preview gain: 444 milliseconds

The control range was 5.476-5.721 seconds; the changed-preview range was 5.176-5.326 seconds. The ranges did not overlap in that recorded subset. Earlier-paint runs behaved differently, so I would not turn this result into a site-wide percentage.

This comparison isolates a specific measured delay under the tested conditions. The October Vercel snapshot covers real mobile traffic across the project. Each answers a different question.

The live PageSpeed gain: about 12% lower LCP

Preview runs on heavier staging data than production, so the numbers above sit higher than what visitors meet. After the September 30 deployment I ran PageSpeed Insights against the live category page: simulated LCP fell from 4.77 to 4.20 seconds (about 12%) and the score rose from 72 to 79, in the same late-paint group (11 and 5 runs). It is still a simulated lab time from PageSpeed's slow test environment.

The historical visitor baseline is a different measurement. The Chrome UX Report for phones (September 2-29, queried October 2) showed a 75th-percentile LCP of 1.01 seconds on the category page, with 94.5% of visits rated good, and 1.73 seconds across the whole site, with 87.8% good. Google's threshold for a good LCP is 2.5 seconds. That window ends on September 29, before the category first-paint change (September 30) and the server-side sanitization change (October 1) went live, so it shows the starting point for those two, not their effect.

That September field window showed two values on the wrong side of Google's "good" line: interaction responsiveness (INP) at 237 milliseconds against a 200 millisecond threshold, with 68% of visits good, and layout shift (CLS) at 0.12 on the category page against 0.10. Those were the next targets at the time. The dated Vercel snapshot near the start of this article reports the newer sample.

Vape website performance and SEO

For this vape website performance case, I keep laboratory measurements separate from visitor experience. Google's guide to measuring Web Vitals explains the difference between controlled lab data and real-user data. PageSpeed Insights field reporting covers the preceding 28 days, so a week of changes does not immediately replace the whole reporting window.

LCP describes loading of the largest visible content, INP describes interaction responsiveness, and CLS describes unexpected layout movement. Google's Web Vitals overview explains the metrics and their thresholds. A loading test alone cannot establish the quality of every interaction.

A Swedish vape catalogue also needs useful category information and reliable mobile navigation. I want visitors to find readable content and working controls, rather than a page built around repeating a search term.

Google says Core Web Vitals contribute to its ranking systems, but good scores do not guarantee top positions. Its page-experience guidance also warns against chasing a perfect score purely for SEO. I am not claiming a ranking or conversion increase from this week's work.

The result—and the next bottleneck

I changed how the storefront delivers its first screen, which JavaScript ordinary visitors receive, and how much repeated catalogue data travels with the page. Cache work follows the same rule: invalidate the content that changed and avoid rebuilding unrelated pages.

The measured results cover different parts of that work: about 15.4 KB less initial JavaScript in the recorded category build, roughly 12% lower simulated LCP in the live late-paint comparison, and a 99/100 Vercel mobile visitor snapshot on October 12. They show the scope of the investigation without turning separate measurements into one invented site-wide percentage.

Checkout layout stability is the next clear target. My remaining checks are:

Investigate the checkout flow behind the 0.32 CLS route result.
Repeat a clean-profile mobile lab comparison on the deployed category page.
Check filters, sorting, product pages and back navigation.
Test early scrolling on a slower connection.
Follow INP and CLS in the same data source over comparable periods.

The historical engineering checks include builds, browser navigation and markup comparisons. This editorial update adds no new browser-flow test or performance experiment. The October 12 visitor snapshot remains dated evidence, rather than a claim about every future visit.

I have covered preserving an existing backend in Headless WordPress AI Migration in One Day. My Headless WordPress Deployment: DNS, SSL, and AI follow-up covers release problems. In Vercel ISR Writes: How I Hit 200,000 and Fixed It, I explain why cache scope also deserves a separate audit.

If your ecommerce storefront sends more work than its first screen needs, contact me about a performance review. I can help trace payload, rendering and cache costs, then define checks that protect the customer flows.

Visual note: the hero is an AI-generated editorial illustration. The three data visuals use the recorded measurements described in this article.

FAQ

What is this Cigge case study about?+
This case study follows Cigge’s storefront performance work across critical CSS, JavaScript, catalogue payloads and cache invalidation. It includes recorded September lab comparisons and an October 12 mobile Vercel snapshot showing RES 99/100, LCP 1.63 seconds and INP 128 ms.
Is the 99/100 result a Google PageSpeed score?+
No. It is Vercel’s Real Experience Score for the selected October 5–12 mobile production sample, checked October 12 at P75. The historical live PageSpeed lab score moved from 72 to 79. Those scores come from different measurement methods and should not be presented as a 72-to-99 improvement.
How much did mobile LCP improve in the recorded test?+
In the late-paint subset of a controlled PageSpeed Insights preview comparison, median simulated LCP fell from 5.620 to 5.176 seconds: 444 milliseconds, approximately 7.9%. The subset had four to five runs per side. A later run on the live page, in the same late-paint group, went from 4.77 to 4.20 seconds (about 12%). These are simulated lab timings, not a site-wide result or a measurement of production visitors.
Is Cigge's category page slow for real visitors?+
The October 12 Vercel Speed Insights snapshot for mobile production traffic, October 5–12 at P75, showed LCP of 1.63 seconds and INP of 128 ms, both rated Great. This is project-wide recorded traffic, not a new category-only measurement. CLS displayed 0.1 but remained Needs Improvement; checkout showed 0.32 across 86 CLS events. The older category-specific CrUX LCP of 1.01 seconds belongs to September 2–29.
Why was the largest category-page content not an image?+
The investigated category page's LCP element was description text. The work therefore focused on first-viewport CSS and optional pre-paint requests rather than assuming another product-image optimization would fix that measured delay.
Does a higher PageSpeed score guarantee better rankings for vape?+
No. Google uses Core Web Vitals in its ranking systems, but good performance scores do not guarantee top rankings. Relevant content, crawlability and the wider page experience still matter. No ranking or conversion increase is claimed here.
Were the results newly measured on the production site for this article?+
The October 12 follow-up checks Vercel Speed Insights for recorded mobile production traffic during October 5–12. The older preview comparison and payload snapshot remain historical evidence. The live PageSpeed comparison dates from September 30, and the CrUX baseline queried October 2 covers September 2–29. These sources and periods do not establish a controlled before-and-after visitor comparison.
✻

Recommended for you

Headless WordPress AI Migration in One Day

Headless WordPress AI Migration in One Day

I rebuilt a WordPress site into a headless frontend in one day using AI, Next.js, and WPGraphQL. Here’s the exact workflow.

10 min read
Headless WordPress Deployment: DNS, SSL, and AI

Headless WordPress Deployment: DNS, SSL, and AI

A real headless WordPress deployment can break DNS, SSL, images, and forms. Here’s how I fixed it in production.

8 min read
Vercel ISR Writes: How I Hit 200,000 and Fixed It

Vercel ISR Writes: How I Hit 200,000 and Fixed It

AI made Vercel the easy default. A broad Next.js cache strategy exhausted my ISR quota. This is the audit and fix I wish I had first.

8 min read