Speed test lower than your plan

People search this after a speed tester shows 80 Mbps on a “300 Mbps” plan. That screenshot is a measurement of one path. The bill is a maximum under conditions the living room Wi-Fi rarely matches. Both can be true at once. The useful question is not “who is lying.” It is “which hop ate the megabits, and is that hop something I can fix.”

Providers advertise a ceiling because the last mile is shared, the radio is messy, and the contract language is built around “up to.” That language is frustrating. It is also why a single low result on a phone is a weak argument. You need a method. The same method you would use to debug slow internet — quiet house, near the router, Ethernet if possible, more than one hour of the day.

Why the numbers disagree

Plans are often sold as “up to,” especially on cable or shared fibre nodes. The neighbourhood streams at once, and your quiet-hour speed sags. That is congestion, not a broken tester. Wi-Fi rarely delivers the full line rate two rooms away. Other devices already using the connection take a cut you will not see on the bill. A VPN adds a hop. And many people compare Mbps on the invoice with MB/s in a download window. Those are not the same unit. A factor-of-eight “deficit” is sometimes just arithmetic.

Protocol overhead exists even on a clean cable. Ethernet, TCP and Wi-Fi headers are not free. Hitting a round marketing number to the last megabit is uncommon. A 10 percent gap on a wired evening test is a different conversation from a 70 percent gap that never recovers at 11 a.m. on Ethernet. Keep those two stories apart.

What “close enough” looks like

On Ethernet, a quiet-hour test near the plan is a healthy line. On Wi-Fi, a noticeable discount is normal, especially through walls. If you need a household yardstick for streaming and calls, use what a good internet speed is rather than the marketing headline alone. A family that is happy at 120 Mbps on the sofa does not have a crisis because the bill says 300. A family that cannot hold a call on a wired 20 Mbps product does.

Device limits hide under the plan too. An old laptop with a 100 Mbps Ethernet port will never show 400. A phone on a weak 2.4 GHz radio will never show a gigabit. Test a second device before you accuse the street. How to test internet speed is the prep so the comparison is fair.

When to push the ISP

Wired or router-side tests that stay far below the plan morning and evening, after a reboot, are worth a ticket. Take three timestamps from the same tester. Note whether a VPN was off. Note the room. That is stronger than a single low result on a phone in the garden. If the provider asks you to test on their own page, do it — and keep your SpeedMetric numbers as a second instrument, not as a fight about which brand is “correct.”

Ask about node congestion if the sag is only at peak hours on a cable. Ask about the modem if a reboot restores speed for an hour. Ask about the drop and the connectors if rain seems to correlate. You still might not get a perfect answer. You will get further than “it feels slow.” For the feeling itself, why internet is slow covers Wi-Fi, devices and delay as well as the plan.

If Ethernet already matches the product you pay for, more advertised megabits are a purchase decision, not a repair. Look at how the house actually uses the line. Two 4K streams plus uploads may justify an upgrade. A single browser tab will not. Improving speed at home is cheaper than a new tier when the radio is the missing piece.

What to put in the support note

Write the plan speed, the tester you used, three timestamps, the room, and whether the run was Wi-Fi or Ethernet. Mention that extra streams were paused and that the VPN was off if that is true. Attach the numbers, not a collage of the highest and lowest screenshots you ever captured. A short table is easier for a human to read than a paragraph of anger, and it survives a transfer to the next agent.

If they ask you to test on their page, do it on the same cable and write that result next to yours. Two instruments that agree are a stronger story than a fight about brands. If they disagree wildly, you still have a clue: one path is different. Repeat both at a second hour before you assume either number is the whole truth. Then decide whether you are asking for a repair or shopping for a different product.

Common questions