What Internet Speed Do You Actually Need in Boulder?

Providers sell speed in round numbers — 300 Mbps, 500 Mbps, a full gigabit — and the tier a household lands on often comes down to whichever number looked safely oversized on a signup page, not what the household actually asks a connection to do. The Federal Communications Commission publishes a more useful answer: a breakdown of how much bandwidth specific activities actually require, and a baseline definition of "broadband" the agency has updated as households have added more devices, more video calls, and more simultaneous streams to a single connection. Matching a plan to real use — rather than guessing upward "to be safe" — starts with those numbers.

The FCC's Current Baseline, and Why 25/3 No Longer Applies

The FCC's Consumer Guide to broadband speed sets 100 Mbps download and 20 Mbps upload as the current baseline definition of broadband — the threshold the agency uses to describe a connection capable of handling typical modern household use. That's a meaningful jump from the 25 Mbps download / 3 Mbps upload benchmark the agency held for nearly a decade before it, a number set back when a single household streaming one show at a time was closer to the norm than a household running two video calls, a 4K stream, and a cloud backup at once. The upload side of that change matters as much as the download side: 3 Mbps upload was never enough for a household with even one person on regular video calls, and 20 Mbps reflects how much more upload-dependent home internet has become.

That baseline is a floor for describing what counts as "broadband" nationally, not a target for every household. Some homes need less than 100/20 comfortably; others — running several devices, several people, several activities simultaneously — need meaningfully more. The FCC's guide also breaks bandwidth needs down by specific activity, which is the more useful starting point for sizing a plan to an actual household.

Mbps by Use Case: What Each Activity Actually Costs

Streaming video. Standard-definition streaming runs on roughly 3–4 Mbps per stream; HD streaming needs 5–8 Mbps; a single 4K stream needs around 25 Mbps. Those numbers are per stream — a household with two people watching different shows on different devices needs roughly double, not the same single-stream number.

Video calls. Video conferencing looks lighter than streaming on paper — the FCC's guide puts a single call in roughly the 1–4 Mbps range — but the number that actually matters is the upload side, and it's easy to overlook. Streaming is almost entirely download; a video call is symmetric, since your device is uploading audio and video in real time at the same rate it's receiving them. Most residential plans allocate far more capacity to download than upload, which is why a connection that streams 4K without issue can still produce a choppy video call — the download number was never the constraint.

Gaming. Individual online gaming sessions typically need only 3–4 Mbps of raw bandwidth, modest next to streaming. What actually determines whether a game feels responsive is latency — the round-trip time between a device and the game server, measured in milliseconds — and jitter, the variation in that round-trip time from one moment to the next. This is one case where adding more Mbps stops helping past a fairly low threshold; a connection type's latency profile matters more to competitive play than its headline speed number.

Telecommuting and remote work. Remote work rarely runs one activity at a time — a video call in the foreground, a cloud sync running in the background, maybe a second household member streaming or on a call of their own. The FCC's guide treats general telecommuting as its own use-case category precisely because it stacks other categories together rather than isolating one.

Why Your Advertised Speed and Your Actual Speed Aren't the Same Number

The number on a plan page is a ceiling under ideal conditions, not a guarantee. The FCC runs a program specifically to measure the gap: Measuring Broadband America collects throughput data from a volunteer panel of households nationwide using dedicated testing hardware, rather than relying on user-initiated speed tests, and compares what major providers advertise against what those households actually receive. The gap the program has consistently tracked is most pronounced during peak evening hours, when the largest share of a shared network is in use at the same time — a pattern that matters more on cable's shared-node architecture than on a symmetrical fiber connection, but shows up to some degree across technology types.

The practical takeaway isn't that advertised speeds are meaningless — it's that a plan's number describes a best case, and the number that matters for day-to-day use is what a connection actually delivers at the hours a household actually uses it. Running your own tests at different times of day and comparing the results against your plan's advertised speed is the most direct way to see where your own connection sits against that gap; a step-by-step guide to testing correctly covers how to get a reading that isn't skewed by Wi-Fi placement or background app usage.

Stacking Devices: How Household Size Multiplies the Need

None of the per-activity numbers above happen in isolation in a real household. A single 4K stream at roughly 25 Mbps, one HD video call using a few Mbps each direction, and a phone backing up photos in the background can add up quickly even before a second person joins in. Sizing a plan means adding up what the household actually runs at once, not what any single device needs on its own.

As a reasoned planning range built from those per-activity numbers, rather than a single fixed rule:

• Single user, browsing plus one stream at a time: the FCC's 100/20 baseline is comfortable headroom.
• Couple or small household with regular HD streaming and occasional video calls: a plan meaningfully above the 100/20 baseline gives room for two activities running at once without one degrading the other.
• Larger household, or one person working from home on video calls most of the day: upload capacity becomes the constraint to watch as much as download, since calls, cloud sync, and file uploads all draw from the same limited upload pool on most plans.
• Household running multiple simultaneous 4K streams, gaming, and video calls at once: the combined draw from several high-bandwidth activities at the same time is where a plan well above the baseline earns its keep.

The point of a range like this isn't to name one "correct" number — it's to replace the instinct to buy the biggest number on the page with a rough estimate built from what the household actually does simultaneously.

Matching a Plan to What You Actually Do

The most reliable way to find out whether a current plan already covers a household's needs is to test it under real conditions — during the hours it actually gets used, not at 10 a.m. on a weekday when almost nothing else in the house is competing for bandwidth. If a plan's advertised number is well above what the FCC's per-activity math says a household needs, and video calls still stutter or a 4K stream still buffers, the upload side or peak-hour performance — not the advertised download ceiling — is usually where the actual bottleneck sits. Testing first, then comparing that result against both the activity breakdown above and whatever a current plan promises, is a more useful exercise than upgrading to the next tier on faith.

Bottom Line

Most households overpay for download speed they don't use and underpay attention to upload speed they need every day. The FCC's baseline of 100 Mbps download and 20 Mbps upload is a reasonable floor for a typical modern household, and its per-activity breakdown — a few Mbps for a video call, several more for HD streaming, 25 Mbps for a single 4K stream, latency rather than raw speed for gaming — gives a concrete way to add up what a specific household actually needs rather than guessing at a round number. Test the connection that's already running before assuming the fix is a bigger plan; the gap between advertised and delivered speed, tracked by the FCC's own Measuring Broadband America program, is often the real story.

References

• FCC Consumer Guide: Broadband Speed Guide — per-activity Mbps benchmarks and the current broadband baseline definition.
• FCC Measuring Broadband America — nationwide advertised-vs-delivered speed data collected from volunteer households.
• Speedtest by Ookla — a tool for checking what a connection actually delivers against the benchmarks above.