Page speed has a direct relationship with revenue. A one-second delay in load time can reduce conversions by several percentage points, and the effect compounds at scale. A store processing thousands of orders per month can lose significant revenue annually from performance issues that are, in most cases, entirely fixable. For Magento Open Source and Adobe Commerce stores, achieving consistent sub-2 second load times requires a layered approach covering server infrastructure, application configuration, front-end delivery, and ongoing monitoring.
This post highlights the key optimisation techniques Accentika applies when conducting performance audits and remediation projects for UK e-commerce businesses.
Full-Page Cache and Varnish
Magento Open Source includes a built-in full-page cache that stores rendered HTML for cacheable pages, most importantly product pages, category pages, and the homepage. By default this cache is stored in the file system, which limits throughput. Configuring Varnish as the full-page cache backend is one of the highest-impact changes available for a busy Magento store.
Varnish is a reverse-proxy cache that sits in front of the web server and serves cached pages directly from memory, entirely bypassing PHP and the database for cacheable requests. Properly configured, Varnish can serve cached pages in single-digit milliseconds, regardless of the underlying server specification. Magento generates a Varnish Configuration Language (VCL) file that handles cache tagging and invalidation correctly across the Magento page types.
Redis for Session and Cache Storage
Beyond full-page cache, Magento relies heavily on its configuration and block caches during page rendering. Redis is an in-memory data store that provides a substantially faster cache backend than the default file-system storage. Configuring Redis for both the default cache and session storage reduces database load and accelerates the rendering of non-cacheable page components, particularly the cart, checkout, and account pages that cannot be served from full-page cache.
For stores with high concurrent user counts, Redis also prevents the session-locking issues that can occur with file-based sessions, improving perceived performance during peak trading periods.
Database Query Optimisation
Magento’s EAV (Entity-Attribute-Value) data model, while flexible, can generate complex database queries, particularly for stores with large catalogues and many custom product attributes. Slow queries are a common source of performance degradation in mature Magento installations.
Optimisation work at the database level typically covers slow query identification and indexing, review of Magento flat table configuration (which pre-compiles EAV data into flattened tables for faster reads), regular re-indexing schedules, and where relevant, read replica configuration to separate reporting and admin queries from storefront traffic.
JavaScript and CSS Bundling
Magento’s front end can generate a large number of JavaScript and CSS files by default. In production mode, Magento merges and minifies these files, but the default configuration does not always produce the optimal result for Core Web Vitals, particularly the Largest Contentful Paint (LCP) and Total Blocking Time (TBT) metrics.
Custom theme development that audits and removes unused JavaScript, defers non-critical scripts, and implements critical CSS inlining can materially improve these scores. The Accentika development team applies these techniques as standard during theme builds and performance remediation projects.
Image Optimisation
Product images are typically the largest assets on any e-commerce page and a common source of avoidable load time. Optimisation at the image level covers several areas. Serving images in modern formats such as WebP rather than JPEG or PNG where supported, ensures Magento’s built-in image resizing is generating appropriately sized variants for each display context, and implementing lazy loading for images below the fold.
A content delivery network (CDN) for static assets including images, CSS, and JavaScript reduces latency for geographically distributed visitors and offloads bandwidth from the origin server. For UK retailers serving primarily domestic traffic, a CDN with UK edge nodes delivers the most consistent improvement.
Hosting Infrastructure
Optimisation at the application level has limits. If the underlying infrastructure is underpowered or misconfigured, application-level optimisations will not deliver their full potential. Key considerations include the PHP version — Magento Open Source 2.4.8 requires PHP 8.3 or higher, and newer PHP versions offer measurable performance improvements. Web server configuration (Nginx or Apache) and the availability of sufficient RAM for PHP-FPM pools and MySQL buffer pools can all be adjusted for performance gains.
For stores with significant traffic, a horizontally scalable infrastructure that separates web nodes from database and cache layers provides both better performance and higher resilience during peak periods. Accentika’s support and maintenance service includes infrastructure review and ongoing performance monitoring as part of the managed service offering.
Measuring What Matters
Performance optimisation should be data-driven. Key metrics to track include Core Web Vitals (Largest Contentful Paint, Cumulative Layout Shift, and Interaction to Next Paint), Time to First Byte, and page weight. Google Search Console highlights Core Web Vitals issues at URL level, making it a practical starting point for identifying the pages with the most significant problems.
It is important to measure real-user performance (field data) rather than relying solely on lab tools such as Lighthouse which simulates a single user on a throttled connection. Real-user monitoring through Google Analytics 4 or a dedicated tool can provide the most accurate picture of how actual visitors experience the store.
Where to Start
For most Magento Open Source stores, the highest-return optimisation sequence is: Varnish full-page cache configuration, Redis for sessions and cache, image optimisation and lazy loading, JavaScript audit and deferral, and database query review. Done in this order, these changes typically deliver the majority of the achievable improvement before more complex front-end or infrastructure work is required.
Accentika offers free performance audits for Magento Open Source and Adobe Commerce stores. The audit covers Core Web Vitals, server configuration, cache setup, and front-end delivery, and highlights the specific changes most likely to deliver improvement for the store in question. Get in touch to request a performance review.
