Throttling means your ISP deliberately slowing a specific kind of traffic — usually video streaming or P2P — while your overall connection tests fine. It is hard to prove by feel and easy to prove with a control.
The signature
Throttling looks like this: a general speed test returns close to your advertised rate, but one particular service is persistently slow. The same service is fast at 4am and slow at 9pm. And crucially, it gets faster the moment you route it through an encrypted tunnel.
That last one is the whole test. Your ISP shapes traffic by identifying what it is. Inside a VPN tunnel it cannot identify anything, so it cannot shape selectively.
The ten-minute method
- Run a general speed test with no VPN. Note the result. This is your line capacity.
- Use the slow service and measure its actual throughput — a streaming quality indicator, or a download meter.
- Connect a VPN to a server in your own country, as close as possible, so distance is not a variable.
- Repeat step two through the tunnel.
- Compare. If the service was slow without the VPN and is noticeably faster with it, while your general speed test barely changed, that is throttling.
Ruling out the alternatives
Before concluding throttling, rule out congestion on the service's side — test at a quiet hour and see whether the pattern holds. Rule out peering disputes, which look similar but affect everyone on your ISP reaching that service. And rule out your own Wi-Fi, by testing on a cable.
A VPN is one of the few consumer tools that produces evidence rather than a complaint. Run the test at the same hour twice before you conclude anything.
What to do about it
Where throttling is regulated, evidence like this supports a complaint to your regulator or ISP. Where it is not, a VPN is a practical workaround: if the traffic cannot be identified, it cannot be shaped.
For that use you want a provider fast enough that the tunnel is not itself the bottleneck. NordVPN measured 902 Mbps nearby in our July cycle, 96% of our 942 Mbps control line, which leaves plenty of headroom.
What the results actually look like
A clear positive looks like this: general speed test 300 Mbps without VPN and 290 Mbps with it, but the streaming service delivering 3 Mbps without and 45 Mbps with. The general figure barely moves while the specific service transforms. That pattern is very hard to explain any other way.
A clear negative looks like everything being slow both with and without the tunnel, which points at congestion, your Wi-Fi or the service itself rather than deliberate shaping.
Run it twice, at the same hour
Network conditions vary enough that a single pair of measurements proves very little. Run the whole sequence on two different evenings at roughly the same time. If the pattern repeats, you have something worth acting on.
One measurement is an anecdote. Two at the same hour on different days is evidence.
Documenting it properly
- Record the date, time and exact figures for each of the four measurements.
- Screenshot the speed tests rather than transcribing them.
- Note which VPN server you used and where it was.
- Keep the raw numbers — regulators and ISP complaint desks ask for specifics, not summaries.
