Investigating a slow home connection
Compare a controlled wired baseline with the problem in the room, rather than collecting isolated peak speeds.
Slow can describe several different problems. A page may hesitate before loading, a download may transfer little data, or a call may break up despite a high speed-test result. Describe the specific symptom first: when it started, when it happens and which devices are affected. A short record often saves more time than running ten speed tests in succession.
Create a useful baseline
Where possible, connect a suitable computer to the router with Ethernet. Check that the operating system and adapter are actually using that connection. Pause large downloads, cloud backups and other avoidable transfers during the comparison. Record the test service and use similar conditions for subsequent measurements.
A test built into the router may measure a different path from one running in a laptop browser. The former can assess the router's connection outwards, while the latter also includes the route to the client. Both are useful, but their results should not be treated as interchangeable. The computer or a USB adapter can also impose a limit.
Keep download, upload and latency results separate. A large download figure does not establish that a video call will be stable. A concurrent upload can affect other activity, and a remote service can itself be slow. Also check the unit: megabits per second and megabytes per second are different measurements.
Investigate the wireless path separately
Repeat on the same computer over Wi-Fi, first near the router and then in the troublesome location. A good wired result with much poorer wireless performance suggests investigating the local radio path. If both are similarly affected, weak Wi-Fi is not necessarily the cause. The comparison narrows the possibilities; it does not prove a particular component is faulty.
If the wired measurement is unexpectedly low, inspect the negotiated link speed as well. A port, damaged cable or adapter can constrain the whole route. Try a cable known to work and another suitable LAN socket, changing one thing at a time. A more expensive internet plan cannot remove a bottleneck inside the house.
Look for repeatable behaviour
Make a few comparisons at the times when the problem normally occurs. Record date, time, connection type, client device and other activity. One excellent result late at night does not explain evening failures. Equally, one poor result could reflect a temporary problem at the test server.
Check whether only one application is affected. If several services work normally but one download host is slow, replacing the router is not the obvious first step. High computer load or an audio problem can also resemble a network fault during a call. Testing a second device and a different application helps keep those explanations separate.
If controlled wired tests remain repeatedly abnormal, contact the provider with the observations. Include the connection type and router version, but do not publish login credentials. Formal evidence about contractual performance may require a particular measurement procedure. These ordinary comparisons are for technical diagnosis, not a legal assessment of the service.
