Faster Core Web Vitals, higher Lighthouse scores, and better conversion — on Luma, Hyva, or headless.
A slow Magento 2 store quietly loses money every day. Shoppers abandon pages that take too long to load, mobile conversion suffers most, and because Core Web Vitals are part of how Google ranks pages, speed problems can cost you traffic as well as sales. Magento is a powerful platform, but its default Luma frontend, a growing stack of extensions, and an under-tuned server combine to produce slow Time to First Byte, heavy JavaScript, and poor LCP, CLS, and INP scores.
Magento 2 performance optimization is the work of finding and fixing those bottlenecks across the whole stack — not just chasing a single Lighthouse number. That means the server and Time to First Byte (caching, Redis, MySQL, PHP, hosting), the frontend (render-blocking CSS and JavaScript, images, fonts), and the third-party scripts that often do more damage than the store itself. The goal is faster real-world pages for real users on real devices, which is what actually moves Core Web Vitals and revenue.
This page is for merchants whose stores feel slow, whose Core Web Vitals are failing in Google Search Console, or whose conversion has drifted down as the site has grown. We optimize any Magento 2 store regardless of how it was built — Luma, Hyva, or a headless frontend — and we work from measurement, not guesswork.
We scope each engagement to where your store is actually losing time, but a performance project typically includes:
Lab data (Lighthouse, WebPageTest), field data (Core Web Vitals from CrUX and Search Console), and server timing, so we fix what real users feel — not just what one synthetic test reports.
Targeted fixes for LCP (largest contentful paint), CLS (layout shift), and INP (interaction responsiveness) — the three metrics Google actually measures and ranks on.
Full-page cache and Varnish configuration, Redis, MySQL query and index tuning, PHP-FPM and OPcache settings, and Elasticsearch — because no amount of frontend work fixes a slow server response.
Critical CSS, JavaScript bundling and deferral, responsive AVIF/WebP images, and a font-loading strategy that removes render-blocking and layout shift.
Auditing and taming the tag managers, chat widgets, and marketing pixels that often cause the worst INP and main-thread blocking.
Edge caching and CDN configuration so repeat visits and global shoppers get consistently fast pages.
We target 90%+ Lighthouse where the architecture allows, and we are clear up front about the ceiling on a heavy Luma store versus what Hyva or headless can reach.
Real-user monitoring and Lighthouse CI budgets so a future extension or campaign tag does not quietly undo the gains.
Every change documented with its impact, so you can see what moved the numbers and what is worth doing next.
We baseline both synthetic scores (Lighthouse, WebPageTest) and real-user Core Web Vitals from CrUX and Search Console. Field data tells us what your actual customers experience, which is often very different from a single desktop test.
We separate server problems from frontend problems from third-party problems. Most 'slow Magento' complaints are a mix, and fixing the wrong layer first wastes budget.
A fast frontend on a slow server still feels slow. We bring Time to First Byte down — caching, database, PHP, and hosting — before polishing the frontend.
We attack LCP, CLS, and INP directly: the hero image, render-blocking resources, JavaScript execution, and layout stability.
We confirm each change against lab and field data, keep what works, and roll back what does not. The Lighthouse number has to hold up in the real world.
We add monitoring and CI performance budgets so the store stays fast after we leave, instead of slowly sliding back.
Performance work is the core of what we do. When we rebuilt Royal Creation's storefront headless, the measured results were:
45 → 96
Lighthouse performance
8.2s → 1.4s
Load time
+34%
Conversion rate
Headless Next.js rebuild of a slow Magento 2 Luma storefront (handcrafted rugs, Switzerland). Royal Creation
Speed work delivers excellent ROI on the right store. But optimization has limits, and sometimes the honest answer is that tuning a heavy store is treating the symptom, not the cause.
When optimization has a low ceiling on your setup, we say so and point you to the rebuild path — a headless build or a Hyva re-theme — rather than bill hours against a wall.
It depends almost entirely on your architecture. On a heavy Luma store with many extensions there is a real ceiling — meaningful gains are possible, but a 95+ mobile score often is not realistic without deeper changes. On Hyva or a headless frontend, 90%+ is achievable and we target it. After the audit we give you a specific, honest target for your store rather than a generic promise. For reference, our headless rebuild of Royal Creation moved Lighthouse performance from 45 to 96.
Usually a combination: a high Time to First Byte from an under-tuned server or missing full-page cache, a heavy Luma frontend shipping large CSS and JavaScript, unoptimized images, and third-party scripts (chat, analytics, marketing tags) blocking the main thread and hurting INP. Catalog size and extension bloat make all of it worse over time. The audit's job is to quantify which of these is costing you the most, so we fix the biggest levers first.
Core Web Vitals are a confirmed ranking signal, but one factor among many — fixing them will not outrank great content on its own. The bigger, more reliable win is usually conversion: faster pages reduce bounce and cart abandonment, especially on mobile. We treat ranking as a benefit and revenue as the goal, and we are honest that speed is necessary but not sufficient for SEO.
Sometimes. If your server is genuinely undersized, that has to be addressed — but often the bigger issue is configuration (full-page cache, Varnish, Redis, PHP and database tuning) rather than raw capacity. We measure first and only recommend a hosting change when the data shows it is the actual constraint, not as a default upsell.
The audit is typically a matter of days. The fixes depend on what we find: backend and caching wins can land quickly, while deeper frontend and Core Web Vitals work runs over a few weeks. We sequence the highest-impact changes first, so you usually see measurable improvement early rather than waiting for the whole project to finish.
We don't publish fixed prices, because the work is scoped to your store and quoted per client. Cost depends on your current store size, how many extensions are in play, your hosting setup, and the depth of the engagement — a focused audit with quick wins is a far smaller commitment than a full re-architecture of a heavy store. A performance audit is the usual starting point: it shows both of us exactly what the work involves before you commit. Get in touch and we'll scope it and send a tailored quote.
Get a free performance audit with your current Core Web Vitals and the realistic gains for your architecture.
Get a free performance audit