Why Speed Tests Give Different Results (and Which One to Trust)
You run three speed tests and get three numbers. That does not mean two of them are broken. They measure slightly different things, and your connection changes from one minute to the next. Here is what causes the gap and how to get a result you can act on.
Some tests pick a server inside or right next to your provider's network. Traffic never leaves the provider, so numbers look high. Others measure to a large content network such as Cloudflare, which is closer to how websites and streaming services reach you. FastSpeedTest uses Cloudflare's nearest location, which is shown with your result.
Wi-Fi is the biggest variable. Moving a laptop one room further from the router can halve the result. Other devices streaming, updating or backing up take from the same line during your test.
On this page
- 1. Different servers
- 2. Different methods
- 3. Your home network
- 4. Your device and browser
- 5. Time of day
- 6. VPNs
- How to get a number you can trust
- Why do different speed test sites show different results?
- Which speed test is the most accurate?
- Why is my speed test lower than my plan?
- Why does the same test give different results minutes apart?
Older phones, battery saving modes, busy tabs and some extensions limit how fast a browser can move data. Above a few hundred Mbps, the device is often the limit rather than the line.
In the evening, more people in your area are online. If the provider's local network is busy, speeds drop and ping rises. Testing in the morning and the evening shows whether this affects you. See peak hours slowdown .
A VPN sends traffic to its own server first. That adds distance and processing, so speed drops and ping rises. Turn it off for one test to see its cost.
If your cabled result is far below your plan at several different times of day, that is worth raising with your provider, along with the times and numbers you recorded.
Key points
- Connections: one connection may not fill a fast line; many in parallel usually can.
- File sizes: small files finish before a connection reaches full speed.
- Averaging: some tests drop the slowest part of the run, some report a peak, some report a plain average.
- Ping: browser tests time web requests, which adds a few milliseconds over a raw network ping.
- Pick one test and stick with it so results are comparable.
- Close other tabs and pause downloads on other devices.
- Run it two or three times and use the typical result, not the best one.
- Run it once on a cable next to the router. That is the closest you will get to your plan speed.
- Repeat at a busy evening hour. Save your results and compare them .
Frequently asked questions
- Why do different speed test sites show different results?: Each test uses its own servers, file sizes, number of connections and way of averaging. A test that picks a server inside your provider's network can show higher numbers than one that measures to a server on the wider internet.
- Which speed test is the most accurate?: None is perfectly right for every use. The most useful approach is to use one test consistently, on the same device, and compare results over time and between Wi-Fi and cable.
- Why is my speed test lower than my plan?: Plans usually advertise a maximum to the router. Wi-Fi, the device, other users and the time of day all lower what one browser can measure. Test on a cable next to the router to see the closest figure to your plan.
- Why does the same test give different results minutes apart?: Other traffic on your network and your provider's network changes constantly. Differences under about 10% between two runs are normal.
More from FastSpeedTest