Internet works but is unusually slow

• Estimated time: 5 min
• Easy
• Low risk

A precise observation avoids testing the wrong hypothesis.

1. Compare two unrelated services and note whether startup is slow, transfer is slow, or access drops. For no access, use ISS-0004/ISS-0019; for disconnects, ISS-0021. If Steam shows 0 B/s with disk activity, use ISS-0018 before judging the entire network.

2. If only one service fails while others work normally at the same time, retain that scope for its support. If several are slow, compare another device on the same network without cellular data distorting the comparison.

• Estimated time: 5 min
• Easy
• Low risk

One test does not establish permanent service speed.

1. With users’ permission, pause identifiable nonessential transfers; do not interrupt work or critical backups. If using a trusted speed test, record device, Wi-Fi/Ethernet, time, and concurrent traffic. Keep tool and conditions comparable, and avoid continuous testing.

2. Compare cautiously with the subscribed plan, not a promise of exact Wi-Fi speed. Improvement after pausing traffic implicates shared load. If unchanged, retain measurements and separate device from path; do not run resets to improve one isolated number.

• Estimated time: 5 min
• Easy
• Low risk

Another normal device reduces the likelihood of a general failure without excluding every shared cause.

1. Compare another authorized device under similar conditions. If only the first is slow, inspect its downloads, synchronization, and managed VPN/proxy use; do not disconnect workplace controls without permission. Compare applications before reinstalling adapters.

2. If stopping your own transfer improves performance, manage that load and stop. If only that device remains slow without explanatory traffic, give device support the comparison. If several fail, proceed to Wi-Fi/Ethernet.

• Estimated time: 5 min
• Easy
• Low risk

Change one variable while keeping the same service and device.

1. On a trusted network, compare the device near the router with its usual location without blocking ventilation or moving unsafe wiring. Improvement makes signal, obstacles, or interference worth reviewing; it does not alone prove a faulty router.

2. If Ethernet is available, compare through a supported connection and confirm the test actually uses it. Normal Ethernet with slow Wi-Fi points toward wireless support/configuration; both slow across devices points toward the shared network or upstream. Without Ethernet, report that limitation rather than inventing the comparison.

• Estimated time: 5 min
• Easy
• Low risk

The pattern decides where to take the evidence.

1. If transfers sustain good throughput but calls or games respond late, record latency/variation when the tool exposes it and whether it occurs under load. Do not automatically call that insufficient Mbps. If only one page starts slowly, service or resolution may be involved; DNS is not a universal accelerator.

2. If several devices worsen at similar times, record a few equivalent comparisons with local traffic identified. A time pattern justifies asking about congestion or upstream paths, not claiming throttling. Give provider/router support those details if it persists over Ethernet too, without random MTU, IPv6, or Registry changes.