Start with measurable causes
Do not infer the source from one speed number. Compare the same device, destination and time under controlled conditions.
CONNECTION DIAGNOSIS
How to Check for ISP Throttling
One slow test does not prove throttling. Build evidence with repeated tests, controlled devices, different times and destinations, then compare the results with the plan terms and provider status information.
Compare, don’t guess
Test over Ethernet with background traffic stopped. Keep other variables stable, then change one condition at a time.
What can create this pattern
Normal congestion resembles intentional throttling
Confirm this possibility with a repeatable comparison before changing equipment or service.
Wi-Fi or device limits distort results
Confirm this possibility with a repeatable comparison before changing equipment or service.
Different services use different network routes
Confirm this possibility with a repeatable comparison before changing equipment or service.
Plan traffic-management terms may apply
Confirm this possibility with a repeatable comparison before changing equipment or service.
Test it in this order
Test over Ethernet with background traffic stopped
Record the result, keep the test comparable and continue only after this step.
Repeat at consistent times
Record the result, keep the test comparable and continue only after this step.
Compare multiple reputable endpoints
Record the result, keep the test comparable and continue only after this step.
Document the plan and the full result history
Record the result, keep the test comparable and continue only after this step.
Use more than download speed
Read throughput together with ping, jitter and packet loss. A browser result describes this device, route and moment; it does not by itself prove an ISP fault.
Continue the diagnosis
speedtest monitor your connection tracerouteFrequently asked questions
How to Check for ISP Throttling
One slow test does not prove throttling. Build evidence with repeated tests, controlled devices, different times and destinations, then compare the results with the plan terms and provider status information.
What should I test first?
Test over Ethernet with background traffic stopped. A consistent, reproducible pattern is more meaningful than a single speed number.
Does one test prove the cause?
No. Repeat comparable tests over Ethernet and Wi-Fi, on more than one device and at different times.
MEASURE BEFORE YOU CHANGE
Start connection test →Find where the connection slows down
Save the result, repeat under comparable conditions and use the difference as evidence.