Does checking my phone during the test affect the result?

If you are testing on the phone and you turned off the screen or switched to another app, yes. The browser goes to the background and the measurement slows. Glancing at your phone while testing on a computer is fine; what matters is the test tab staying in front on the computer.

Why do I get a different result every time?

Internet speed is not a fixed number; it varies with momentary congestion, Wi-Fi signal and the other devices active at the time. Measuring a few times and reading a general range is more accurate than fixating on a single number.

Is testing in incognito mode more accurate?

Incognito disables extensions, so it removes factors like ad blockers and VPNs that touch your traffic, which can give a cleaner measurement. But the main factors, like background tabs and Wi-Fi, are the same in incognito.

Which browser is best for a speed test?

Every up-to-date browser is capable of measuring correctly. The difference is not the brand but the conditions: a single tab in front, a wired connection and no other traffic improve the result in any browser.

Tek monitörlü, düzenli bir masaüstü bilgisayar çalışma alanı

Measuring correctly

How to run a speed test correctly: background tabs and browser traps

Last updated: July 24, 20265 min read

Browsers deliberately slow down tabs you cannot see. If you switch to another tab during the test, the measuring code is suspended and the result comes out below your real speed. One rule is enough for a reliable measurement: keep the test tab in front and open, and do not switch away. A browser left open for a long time, or with dozens of tabs, can also lower the result.

Why does a background tab lower the result?

Modern browsers restrict tabs you cannot see in order to save battery and CPU. While a tab is in the background its code either runs very slowly or is frozen entirely. A speed test is code too: send the tab to the background mid-measurement and the measurement slows down, so the result comes out low.

This is not a guess; it is documented browser behaviour. Chrome describes how it throttles the timers of a backgrounded page:

“The browser will check timers in this group once per second.”[1]

So code that can only run once per second in a background tab cannot properly drive a speed test that measures on a millisecond scale. Chrome also documents that it can freeze CPU-intensive background tabs outright when Energy Saver is on.

That is why, if you see "you switched to another tab during the test" on your result card, the number may not reflect your real speed. The best move is to measure again with the page in front.

Simple rules for a reliable measurement

Keep the test tab in front. After you press start, do not switch to another tab, window or app; wait until it finishes.
Stop other downloads and uploads. Cloud backups, downloads and updates share your bandwidth and lower the result.
Account for other devices on the same line. If someone at home or the office is streaming, they share the line with you.
Test over a cable if you can. Wi-Fi both lowers and destabilises the result.
Measure a few times at different moments, not once. A single test may have caught a passing spike of congestion.

A browser open for too long

A browser left open for days with dozens of tabs grows heavy over time. Memory fills, tabs compete for the CPU, and the measuring code cannot get the resources it needs. In that state the result can come out below your real speed.

Close tabs you are not using before measuring; let the speed test stand alone.
Fully quitting and reopening the browser clears the accumulated load and is often the fastest fix.
Ad blockers and VPN extensions also touch your traffic; if in doubt, disable them temporarily and re-test.

Know what you are actually measuring

A speed test measures what that device gets, from that spot, at that moment. That is the most honest number for your day-to-day experience. But if you want your subscription's capacity, you have to fix the conditions: a wired connection, a single tab in front, no other traffic. Only then does the number approach your line's ceiling.

If you want to know how we measure and what each value means, our methodology paper is open. Measurement methodology

Frequently asked questions

Does checking my phone during the test affect the result?: If you are testing on the phone and you turned off the screen or switched to another app, yes. The browser goes to the background and the measurement slows. Glancing at your phone while testing on a computer is fine; what matters is the test tab staying in front on the computer.
Why do I get a different result every time?: Internet speed is not a fixed number; it varies with momentary congestion, Wi-Fi signal and the other devices active at the time. Measuring a few times and reading a general range is more accurate than fixating on a single number.
Is testing in incognito mode more accurate?: Incognito disables extensions, so it removes factors like ad blockers and VPNs that touch your traffic, which can give a cleaner measurement. But the main factors, like background tabs and Wi-Fi, are the same in incognito.
Which browser is best for a speed test?: Every up-to-date browser is capable of measuring correctly. The difference is not the brand but the conditions: a single tab in front, a wired connection and no other traffic improve the result in any browser.

Sources

References

[1]Chrome for Developers — Heavy throttling of chained JS timers (Chrome 88)

Images

Fotoğraf: Sharad Kachhi / Pexels

Did this guide help?

Run the test again