How to Run a Fair Internet Speed Test

Why most speed tests are not quite fair

People often run a speed test while a video is playing in another room, a phone is updating its apps and the laptop is two walls away from the router. The result looks poor, and the connection gets blamed. In many cases the test simply measured a busy home network rather than the line itself. A fair test removes as many outside influences as possible, so that the number reflects what your connection can actually deliver.

This guide gives you a repeatable method. If you follow the same steps each time, you can also compare results over days and weeks and notice real changes instead of random noise.

Step one: decide what you want to measure

There are two different questions you might be asking. The first is: what does my internet line deliver at its best? The second is: what do I really get on my phone or laptop in the room where I use it? Both are useful, but they need different setups.

• Line test: a computer connected to the router by an Ethernet cable, with nothing else running. This shows what the connection into your home can provide.
• Experience test: the device you normally use, in the place you normally use it, over WiFi. This shows what you actually get, including the effect of walls, distance and interference.

If the line test is good and the experience test is poor, your internet service is probably fine and the problem lies inside your home network. If both are poor, the issue may be with the plan, the line or the provider.

Step two: prepare the network

Before you start, reduce competing traffic. Pause or close anything that uses data in the background.

• Pause streaming, large downloads, cloud backups and software updates on every device.
• Close extra browser tabs and applications on the device you are testing.
• Ask others at home to stay off heavy use for a couple of minutes.
• Turn off any VPN or proxy, unless you specifically want to test it.
• Check that no smart TVs or game consoles are downloading in the background.

It also helps to restart the router and the testing device if they have been running for a long time without a break, but do this before the test, not in the middle of a comparison.

Step three: choose the connection method

For a line test, plug a laptop into one of the router's LAN ports with a good quality Ethernet cable. Disable WiFi on the laptop so that you are sure the cable is being used. For an experience test, stay on WiFi and stand or sit where you normally use the device. Note which band you are connected to if your router offers more than one, because the faster band usually works best close to the router and the slower one reaches further.

Step four: pick the test server sensibly

Most speed test tools choose a nearby server automatically. That is usually a sensible default. If the tool lets you choose, a server that is geographically close and has a low ping tends to give the most representative result for everyday use. Testing against a distant server will often show lower speeds, because distance and extra network hops add delay. If you want to see how a connection behaves to a far-off location, do that as a second, separate test and label it clearly.

Step five: run the test several times

A single run can be unlucky. Run the test three to five times in a row with a short pause between runs, and write down every result. Look at the range. If the numbers are close to each other, you can trust the pattern. If they swing widely, that is useful information too, and it may point to interference, congestion or an unstable link.

• Record download, upload, ping and, if shown, jitter and packet loss.
• Note the date, time, device, connection type and server location.
• Do not average away a bad result; mention it and look for a cause.

Step six: repeat at different times

Network conditions change during the day. Run the same test in the morning, in the afternoon and in the evening, and on a weekday and a weekend if you can. If results are consistently lower at busy hours, shared capacity in your area may be a factor. If they are low at all times, look at your equipment and plan instead. Keep your notes in a simple table or a notes app so that you can show them to your provider if you need to raise a problem.

Common mistakes to avoid

• Testing on a different device each time: older phones and laptops have slower WiFi hardware, so results are not comparable.
• Comparing different test tools directly: each tool uses its own method and servers, so figures can differ even on the same line.
• Ignoring the unit: remember that speeds are in megabits per second, not megabytes.
• Testing with a VPN on and forgetting it: the tunnel adds delay and can reduce speed.
• Judging by one number: look at upload, ping and loss as well as download.

Using your own results

Once you have a handful of tests, compare them with the speed your plan promises. Plans usually describe a maximum or typical speed rather than a guarantee, so some difference is expected, especially over WiFi. A consistent and large gap on a wired line test is worth raising with your provider, and your notes make that conversation easier. A gap that appears only over WiFi is more likely a home network matter, which you can often improve by moving the router, changing the band or reducing interference.

A short checklist

• Close other apps and pause other devices.
• Turn off VPN and proxy.
• Use Ethernet for a line test, WiFi in your usual spot for an experience test.
• Use a nearby server.
• Run at least three tests and record every result.
• Repeat at different times of day.

With this routine, a speed test stops being a one-off guess and becomes a reliable way to understand how your connection really behaves.

Frequently asked questions