Why is my internet slow?

“Why is my internet slow?” is almost never a question about a single magic number. It is a complaint about an experience: a film that pauses to buffer, a video call that freezes on someone’s forehead, a page that sits on a white screen, a game that rubber-bands three rooms away from the router. Those symptoms can share a cause, or they can come from completely different parts of the path. A free internet speed test is the fastest way to stop guessing. Measure download, upload and ping, write the three numbers down, then change one thing at a time.

People often treat the first result as a verdict on the internet service provider. That is too fast. The tester sees whatever path your device is using in that moment. If you are on Wi-Fi two walls from the access point, the measurement includes the radio hop. If a laptop is backing up a camera roll, the measurement includes that traffic. If a VPN is on, the measurement includes the VPN server. Slow internet is real. The location of the slowdown is what you need before you buy a new plan or argue with a support script.

Slow is not always low megabits

A 200 Mbps plan can still feel miserable. Ping can be high. Wi-Fi can drop packets. Several 4K streams can share a crowded 2.4 GHz band. Upload can be too small for two people on camera at once. Download, upload, ping and jitter describe different jobs. Streaming mostly cares about a steady download. Video calls care about upload as well as download. Games and remote desktops care about ping even when the megabits look generous. If the capacity numbers look healthy and the app still stutters, read packet loss, jitter and lag before you assume the plan is too small.

The opposite mistake is also common. Someone sees 40 Mbps on a phone in the garden, compares it with a “300 Mbps” bill, and decides the provider is lying. Plans are typically a maximum under good conditions, often sold as “up to.” Wireless is rarely as clean as a short Ethernet run. Units get mixed up as well: testers speak Mbps, while many download windows speak MB/s. Those are not the same quantity. Mbps versus MB/s walks through the conversion so you are not comparing eight times the wrong number.

The usual causes, in the order they actually happen

Distance and walls come first in most homes. Radio does not punch through brick and metal the way a cable does. A router in a cupboard, a basement utility room, or the far end of a long apartment will starve the rooms you actually sit in. Neighbour access points, a microwave, a baby monitor, or a crowded 2.4 GHz band make that worse. The pattern is distinctive: a test next to the router looks decent, and the same tester in the living room collapses. That gap is Wi-Fi, not the street cable.

Other devices are the second cluster. Phones updating overnight, a TV catching up on 4K episodes, a console downloading a 90 GB patch, and a laptop syncing photos can flatten everyone else’s calls. The line is not “broken.” It is busy. Pause the obvious transfers, test again, and you often get the plan back without touching a setting.

Software on the same device matters too. A VPN often lowers throughput and raises ping because traffic takes an extra encrypted hop. Browser extensions, corporate filters, and “optimizer” apps can intercept or throttle paths in ways that look like a bad ISP. Pause the VPN, test once, then turn it back on if that is how you actually work. You want two measurements, not a lecture about privacy.

Evening congestion sits on the ISP side. Cable and some fibre products share a node. When the neighbourhood streams at once, your quiet-hour speed can sag. Aging combo modem-routers also get tired: a reboot restores speed for an hour, then the radio or the WAN session falls over again. That pattern is a hardware or last-mile clue, not a reason to keep rebooting forever.

A two-minute check that actually isolates the hop

Pause extra downloads. Stand near the router. Run the tester. Then run it again in the room that feels slow. A large gap is almost always the wireless hop. If both spots are poor, you have not proven the ISP is at fault yet, but you have ruled out “just this armchair.” Follow how to debug slow internet next: other devices, VPN off, a reboot, then Ethernet if the computer has a port or a USB adapter.

Three timestamps beat one screenshot. Morning, early evening, and late night tell you whether the problem is constant or only at busy hours. Constant and wired-and-still-low points at the plan or the last mile. Only-in-the-far-bedroom points at placement, mesh, or a cable. Only-when-the-VPN-is-on points at the VPN. That is the whole method. It is boring, and it works.

When the plan itself is the bottleneck

Sometimes the line is doing exactly what you paid for, and the household outgrew it. One person browsing is a different job from two 4K streams, a cloud backup, and a classroom call. What a good internet speed looks like is a better yardstick than the marketing headline. If Ethernet already saturates a 50 Mbps product every evening, buying a faster plan can be the right lever — after you have confirmed the wire can use the megabits you already have. Paying for gigabit on a distant 2.4 GHz radio is how people waste money.

If the wired or near-router tests stay far below the advertised “up to” number at several hours of the day, after a reboot, you have a fair case for the provider. Bring the numbers and the timestamps. Support teams respond better to “82 Mbps on Ethernet at 19:10 and 21:40 on a 300 Mbps plan” than to “it feels slow.” More on that mismatch lives in why a speed test can sit below the plan. Practical fixes that do not require a new contract are in how to improve internet speed at home.

Common questions