Benchmark
Public measurements, not promises.
We measure our servers and WordPress sites with an open method and publish the numbers here, even when they aren't the best. The first data arrives with the beta.
What we measure
Eight measures you can actually feel.
| What we measure | Why it matters | How |
|---|---|---|
| Time to first byte (cached page) | How long a visitor waits before seeing anything. | Requests from several cities, median and 95th percentile |
| Time to first byte (uncached) | Shows how fast WordPress is on its own, when cache doesn't help. | Page with cache bypassed, same plugins everywhere |
| Requests per second under load | How many visitors it handles at once before slowing down. | Ramping load test with open tools |
| WooCommerce shop checkout | The step that costs orders when it's slow. | Test order repeated in sequence |
| Core Web Vitals score | What Google looks at to judge speed. | PageSpeed and Lighthouse, repeated |
| Database speed | WordPress writes a lot: a slow disk shows. | Synced writes per second |
| Restoring a backup | How long it takes to get your site back. | Time from request to site back online |
| Migrating a typical site | How long the move from another host takes. | Total time on a sample site |
At server level
Before WordPress, the hardware.
A site is only as fast as the server hosting it. We measure every server with the same script before putting customers on it.
| Server measure | Why it matters for WordPress | Tool |
|---|---|---|
| CPU, single thread | A WordPress page uses a single core: single-core speed is what counts. | sysbench |
| CPU, all threads | How many pages at once the server handles. | sysbench |
| Memory: bandwidth and random access | Cache and database live in memory. | sysbench |
| Disk: random reads and writes | WordPress reads and writes many small files. | fio |
| Disk: writes with fsync | The database does them on every change: it's the measure that separates PC disks from datacentre disks. | fio |
| Database under load | How many operations per second MySQL handles. | sysbench OLTP |
| PHP: compute test | How long PHP takes to do the same work. | php-cli |
| Network: bandwidth and latency | How much flows from the server to the CDN and visitors. | iperf3, ping |
Each measure is repeated three times and we report the median and raw values. The script is public: you can run it on any server.
How we measure
The same test for everyone.
The results
The numbers will appear here.
The table is empty on purpose: we don't put in figures we haven't measured.
| Measure | 1Prime Cloud Grow | Competitor A | Competitor B |
|---|---|---|---|
| First byte, cached | — | — | — |
| First byte, uncached | — | — | — |
| Requests per second | — | — | — |
| WooCommerce checkout | — | — | — |
| Core Web Vitals | — | — | — |
| Database writes per second | — | — | — |
This is how we'll show the bars: shorter is better for timings, longer for requests per second.
Why we expect it to do well
The technical choices, to be checked against the numbers.
These are the reasons we expect good results. They are hypotheses until we measure them.
Multi-layer cache ready to go
The KoaPanel control panel sets up page caching (nginx), object caching (per-site Redis) and OPcache by itself. No plugin to install.
NVMe
NVMe disks for files and database: less waiting when WordPress reads and writes.
The site closest to the visitor
Pages and images go through eight European nodes and are delivered from the closest.
Updates without breakage
The panel checks the site still responds after every update, and rolls back if not.
What we don't promise
The rules of the game.
No made-up percentages
We won't write «78% faster» without saying compared to what and how we measured it.
We publish even if we lose
If a competitor is faster on a measure we say so. If the numbers say we need faster hardware, we change hardware.
FAQ
Benchmark questions
Which PHP version do you use? Can I change it?
We'll publish the list of supported versions before launch, and you'll be able to choose the version for each site from the panel.
Do I need a caching plugin?
No. The server already has page caching (nginx), object caching (per-site Redis) and OPcache. A second caching plugin is usually unnecessary and can conflict: if you already have one, write to us.
How do I make my WordPress site faster?
Start with light images, few plugins and a well-built theme: the server does the rest with caching and the CDN. Then measure with our free tools and compare before and after.
How fast are your servers?
We don't write percentages we haven't measured. We're preparing tests with open method, scripts and data: we publish them on the Benchmark page, even if they aren't the best.
How often are backups made and how long are they kept?
Backups are daily, kept on a different server from your site, and you can take one on demand (on Grow and Geek). We write the retention period into the terms and on the Technology page before launch.
What's the difference between hosting and Cloud?
On shared hosting several sites use the same server, with a limit per site. On Cloud, CPU and RAM are reserved for you, with no overbooking. The panel's features are the same.
Put your own site to the test.
Join the beta: we bring your site over for free and you compare the timings.