What an advertised internet speed leaves out

Three numbers, three questions

Write the three numbers in separate columns. They answer different questions, and mixing them is how a plan that looked fine in an ad disappoints in a back bedroom.

The FCC consumer page, updated July 23, 2026, says labels must disclose prices, introductory rates, data allowances, and broadband speeds. Most providers had to show a label at the point of sale by April 10, 2024. Providers with 100,000 or fewer subscriber lines followed on October 10, 2024. Ask for the label that belongs to the plan at your address, not a sample from another city.

The money side of that label, including how long a headline rate lasts, is covered in the full cost of home internet. This page stays with the speed lines.

What typical download and upload mean

The glossary defines typical download speed as the speed at which data arrives at a device, usually in megabits per second. Higher means files and video arrive sooner. It defines typical upload speed as the speed at which data leaves the device. Upload is what a video call, a photo backup, and a cloud drive use when you send.

The April 2024 machine-readable spec, revision 1.0 dated April 17, 2024, stores both figures in megabits per second, with the sample download allowed two decimal places. The sample row is an illustration in the spec, not a plan you can buy. Use it only to see the unit. Your label has its own numbers.

A plan name that includes a speed still has to match the label. The spec says that if a provider names a plan by its speed tier, that speed tier must be accurate and consistent with the speed metrics elsewhere on the label. When the name and the typical-download line disagree, ask which one the provider will stand behind at your address.

Download and upload are often far apart on cable and closer together on fiber. Fiber, cable, and 5G walks through that difference. How internet reaches a house explains why the line to your unit can differ from the line next door.

Latency is a separate line

The glossary calls typical latency the time for data to travel from one endpoint to another, in milliseconds. People call the delay lag. Lower latency indicates a higher quality connection, but a small delay from distance is unavoidable. The same paragraph says latency can be noticeable for streaming video, live audio, and online games, and that high latency can make real-time conversation hard, including a telehealth visit or a conversation in American Sign Language.

Buying a higher download tier does not, by itself, shrink that delay. The label already has a latency line. Read it. If the work that matters is calls, start with internet for working from home. If the work is cloud gaming, start with cloud gaming and latency. Both pages use the same idea: capacity and delay are different limits.

Two checks before you blame the plan:

• Run the call or the game on a wire from the router, then again on Wi-Fi in the room you actually use.
• If the wired attempt is fine and the room is not, the path inside the house is the limit. The plan tier will not move the router.

Why a speed test can miss the label

The glossary says typical speeds may not be the speed you see on a test, and that once service is connected to a modem and any routers, the number of devices in the home may change the speed any one device receives. A test is one device, one path, one minute. The label is the provider figure for the plan.

Before you decide the plan is short, write down four facts about the test:

• Wi-Fi or a cable into the router.
• The room, and whether a wall or a floor sits between you and the router.
• How many other devices were on, including TVs and cameras.
• The time of day. A quiet morning and a busy evening are different loads on a shared line.

A weak result in one room, with a strong result beside the router, points at Wi-Fi placement. Why your Wi-Fi is slow is the room-by-room version of that test. A weak result on a wire, next to the router, repeated at two times of day, is the result to show the provider, next to the label.

Do not average the ad and the test and call that the truth. Keep the label, the test screenshot, and the time. If they disagree, you have something specific to ask about.

A data allowance is not a speed

The glossary describes the data included with the monthly price as the amount you get before extra fees or slower performance. It says the label must show those extra charges or reductions, often as a cost for more data, and may link to the provider policy. That line caps how much you can use. It does not raise or lower the typical speed while you are under the cap.

The unit is easy to misread. The glossary text updated March 11, 2024 says the allowance is typically represented in gigabits (GB) or terabytes (TB). The April 17, 2024 machine-readable spec describes the same monthly allowance field as gigabytes. Those are different units that share the letters GB. Read the unit printed on your label. Do not convert a remembered abbreviation in your head.

If you move large files or several people stream at once, ask whether the plan has a monthly cap and what happens after it: a fee, a slower speed, or neither. Our data cap guide goes further on that choice. The label is still the first place the number should appear.

What to ask before you order

Take this list to the checkout. You want answers for your address, not for the provider footprint in general.

• Show me the broadband label for this plan at this address, including the unit number.
• What are the typical download, typical upload, and typical latency on that label?
• If the ad says up to a higher figure, which number is the typical speed?
• Is there a monthly data allowance, and what happens after it?
• Does the plan name a speed that matches the typical-download line?

If you are comparing a cable plan with a fixed-wireless or 5G home plan, the label still uses the same fields. The path to the house is what changes. A Verizon hub page such as where Verizon lists Fios and 5G Home is a provider article, not a speed promise. Use it to see where that provider says it offers service, then come back to the label for the plan you were actually offered.

Save the label with the date. If the first month feels slower than that label, the notes from your speed test are the comparison. The research method explains how we date a source and how to tell us when a guide is wrong.

What to do next

If the open question is the bill rather than the speed line, use the cost guide and the household needs tool. If the open question is whether a wire even reaches the unit, start with how internet reaches a house.