Our TYPO3 backend runs as fast as the frontend — here's how we got there

TYPO3 is slow, old, complicated? We've optimised our backend so it responds like a modern frontend. A look at the concrete measures.
You meet it regularly, this prejudice: TYPO3 is slow, sluggish, basically a relic. Anyone saying that has either not looked in a long time — or has never seen a properly set up TYPO3 backend. We work with it daily and can tell you: our backend is no slower than a modern SaaS product. In many editorial day-to-day situations it's even noticeably faster, because it's tailored exactly to your content and your structure.
That didn't happen by itself. We invested deliberately in performance. In this post we show which screws we turned — and why the result is more than just a nice feeling on click for you as the customer.
Why backend performance isn't a luxury
The frontend of a website is optimised almost everywhere today. The backend only crosses your mind when editors complain. That's short-sighted. Every second your editorial team spends waiting for a loading backend multiplies across the day, the week, the year. A sluggish backend means less content, sloppy maintenance and, in the worst case, frustration that ends in resignations.
That's why backend performance is a fixed part of CMS as a Service for us. If we run a system, it shouldn't just run, it should work. Specifically: for the people who earn money with it every day.
The screws we turned
A caching strategy across several layers
Caching in TYPO3 isn't a single switch but a system of several layers. We rebuilt our cache strategy: TYPO3's own caching framework no longer sits on disk or in the database, but in a central Redis cluster. The backend user cache, used too quietly in many projects, is actively configured here and gets cleared in a controlled way during deployments. You notice it immediately when switching pages in the backend — calls that used to take half a second now sit in the few-millisecond range.
Asset pipeline for the backend
The TYPO3 backend is a web application. It benefits from the same techniques as a modern frontend: bundled JavaScript and CSS assets, Brotli compression, efficient browser caching. We adapted the asset pipeline so backend assets aren't reloaded on every login, but stay long-cacheable.
PHP OPcache and JIT properly configured
Many systems run with an OPcache standard configuration intended for an average project. We tuned OPcache deliberately to our typical TYPO3 workloads — including a generously sized opcache.memory_consumption, an opcache.max_accelerated_files matched to the extension footprint, and JIT enabled where it actually helps. The result is measurable: noticeably lower TTFB times in the backend under load.
HTTP/2 and HTTP/3 everywhere
Backend communication today isn't a single request. It's many parallel calls for listings, AJAX content, previews. HTTP/2 and, where possible, HTTP/3 cut these parallel connections massively. Both are standard in our hosting base, not optional.
Critical CSS and lazy loading in the backend
The backend also has templates with blocks and components rarely needed straight away. We separate critical from non-critical styles so the first render is quick, and load less important areas asynchronously. Not a spectacular trick, but the sum makes a noticeable difference.
Database tuning
Many performance problems in the TYPO3 backend are, ultimately, database problems. We review index strategies for project-specific tables, optimise queries on critical paths, and use a MariaDB or MySQL setup that harmonises with your project's extension set. Where needed, we work with read replicas to take load off large listings.
Why this gives you more than just faster clicks
A faster backend doesn't just mean a better feeling. It specifically means:
More output per working day in the editorial team. A noticeably faster backend leads to more maintained content, because nobody postpones edits out of frustration.
A smaller frustration curve for new editors. Anyone joining compares TYPO3 with the tools they use privately every day. The bar is high — a fluid backend keeps up.
Better conditions for integrations. If your CMS backend reacts promptly to inputs, external editorial workflows, imports and AI-assisted tools can be deployed more confidently.
Not a one-off effect, but a working standard
Important to us: these optimisations aren't the result of a one-off tuning project. They're part of our standard. Every new TYPO3 project starts with these conditions. Existing systems we take over are brought to the same level step by step — visible in Lighthouse measurements, in monitoring, and above all in your team's day-to-day feel.
![[Translate to English:] Nahaufnahme einer frisch geölten Fahrradkette und Kettenblatt mit scharfen Details im Studiolicht.](/fileadmin/_processed_/5/a/csm_98fe9205cf4752b782d22dc0696e6c37227385b4ddd908e69cc2142f9455ad56_45810731e4.jpg)
An offer
If your editorial team experiences your TYPO3 backend today as a brake rather than a tool, a short conversation is worth it. We'll take 30 minutes and tell you where the biggest levers in your setup are. No pitch, no abstract recommendations — a concrete assessment.
Frequently asked questions
What customers most often ask us on this topic — answered openly.
What does this performance work cost us?+
For new projects this is part of our standard and not invoiced separately. For existing systems we classify the levers in a short analysis by effort and impact and build a manageable budget from that. No black-box tuning — every measure is traceable.
Doesn't something break with all that caching?+
Caching is only stable if invalidation is right. We use cache tags consistently and check invalidation on every deployment. The backend therefore never sees stale content — it just sees the current content faster.
How much effort is it to retrofit an existing system?+
It depends on the starting point. The caching and hosting layer can usually be brought up to our level within a few weeks. Database and extension optimisation happens iteratively, prioritised by your editorial team's biggest pain points.
Does this even bring anything for a small team?+
Especially for small teams: if an editor waits 20 fewer minutes a day for load times, that adds up to several working weeks over a year. Performance has impact regardless of team size.
How do you actually measure backend performance?+
We don't measure only synthetically; we use real-user monitoring out of the backend. Relevant metrics are TTFB, time to interactive and the actual click-to-click time of typical editorial workflows — not abstract Lighthouse scores.