In the global cloud era it looks like server location no longer matters. The numbers say otherwise. Latency, the time a packet takes to reach the server and come back, is still physical: bounded by the speed of light in fibre and by every device the packet crosses along the way.
The calculation almost nobody does
A modern web page is not one request. It is the HTML, then stylesheets, scripts, fonts, images, and often several API calls to load dynamic content. Easily 30 to 80 requests for a normal page.
If your server is 20 milliseconds from your visitor and another is 90, the difference per request is 70 milliseconds. Multiplied by the requests that cannot run in parallel because they depend on each other, the result is seconds of difference in perceived load time. And that is with the same machine, the same code and the same bandwidth.
Bandwidth does not fix latency
This is the most common confusion. Bandwidth is how much data fits per second; latency is how long the first byte takes to arrive. Doubling bandwidth does not reduce latency by a single millisecond. It is like widening a road: more cars get through, but no individual car arrives sooner.
That is why a site with light images can feel slow from the other side of the world, and a heavy site can feel fast when the server is nearby.
What this means for a local business
If your customers are in Andorra, the Pyrenees or northern Catalonia and France, a server in the country gives them the lowest possible latency. And there is a nuance that gets overlooked: it is not just about kilometres, it is about which path the traffic takes. Two servers at the same physical distance can have very different latency if one has a direct route and the other goes through a distant hub.
Proximity also brings things that are not measured in milliseconds: support in your language and time zone, the option to physically visit the data centre if it ever matters, and your data under your country’s law.
How to measure it yourself
You do not have to take anyone’s word for it, you can check in two minutes. From a terminal, a ping to your current domain gives you average latency. A traceroute shows the path traffic takes and where time is lost. And your browser’s developer tools, in the network tab, give you time to first byte, the metric that best reflects your server’s real latency.
Compare those numbers by running the same test against a local provider and a distant one. The gap is usually bigger than people expect.
When latency is not your problem
Let us be honest: if your site takes four seconds to load, latency is not your bottleneck. It is probably unoptimised images, too many plugins, slow database queries or missing caching. Fix that first, because the room for improvement is far greater.
Latency matters once you have done your homework: then it is the difference between a fast site and an instant one. And for applications with many API calls, or admin panels where every click is a request, it shows from the very first moment.
Test it with data, not faith
Our virtual servers in Andorra sit in our own La Massana data centre, with an Andorran IP and a direct route to local carriers. And if what you want is managed web hosting, our hosting plans run on the same infrastructure.
Since anything that lives in one place needs a copy somewhere else, our backup services keep your history in two data centres of our own.