You run a speed test, it says 200 Mbps download, and your video call still stutters. Or it says 50 Mbps and everything feels fine. Speed test results are useful, but they’re frequently misunderstood, partly because “speed” is the wrong metaphor for what’s actually being measured.
Here’s what each number on a speed test result actually tells you, what it doesn’t tell you, and which number matters for what you’re doing.
Download Speed: How Much Data, Not How Fast
Download speed, measured in megabits per second (Mbps), is bandwidth, not speed. The distinction matters.
Think of it like a highway. Bandwidth is how many lanes the highway has. More lanes mean more cars can travel simultaneously, but each car doesn’t go faster. A 200 Mbps connection doesn’t load a single web page faster than a 50 Mbps connection in most cases, because a web page isn’t big enough to saturate either connection. Where bandwidth matters is when you’re moving large amounts of data: downloading a game, streaming 4K video, or having multiple people and devices sharing the connection at once.
Here’s what different activities actually need:
- Web browsing, email: 5-10 Mbps is plenty
- HD video streaming (1080p): 5-10 Mbps per stream
- 4K video streaming: 25-35 Mbps per stream
- Video calls (Zoom, Teams): 3-5 Mbps up and down
- Online gaming: 5-15 Mbps (gaming uses surprisingly little bandwidth; latency matters far more)
- Large file downloads: as much as you can get
If you have 100 Mbps and live alone, you have more bandwidth than almost anything you do will use. If you have 100 Mbps and four family members are each streaming in 4K while someone downloads a game update, you might feel squeezed.
Upload Speed: The One Most People Ignore
Upload speed is how quickly you can send data from your device to the internet. On most residential connections, upload is significantly slower than download, often 10-20% of the download speed. This is by design: cable and DSL technologies allocate more capacity to download because historically people download far more than they upload.
Upload speed rarely mattered until remote work and video calling became common. During a video call, you’re uploading a continuous stream of your own video and audio. If your upload is too slow, other participants see you freeze, pixelate, or drop out, even if your download speed is excellent. You can be watching them in crystal clarity while they’re seeing a slideshow of your face.
Activities that need decent upload:
- Video calls: 3-5 Mbps upload minimum, more if you’re sharing your screen
- Live streaming: 6-10+ Mbps depending on quality
- Cloud backups: as much as you can get (affects how long backups take)
- Uploading files/photos: matters for large files, irrelevant for email
If your video calls are choppy and your download speed looks fine, check your upload. It’s the culprit more often than people realize.
Latency (Ping): The Number Gamers Care About
Latency, often displayed as “ping” and measured in milliseconds (ms), is the time it takes a small packet of data to travel from your device to a server and back. Unlike bandwidth, this one actually is about speed: how quickly a single piece of information makes the round trip.
Low latency means things feel responsive. High latency means things feel sluggish, even if your bandwidth is enormous. You could have a 1,000 Mbps connection with 500 ms latency and your experience would feel terrible. Every click, every keystroke in a remote application, every action in an online game would have a half-second delay.
What different latency levels feel like:
- Under 20 ms: Excellent. You won’t notice any delay in anything.
- 20-50 ms: Good. Fine for gaming, video calls, and everything else.
- 50-100 ms: Acceptable. Casual gaming and video calls work fine. Competitive gaming starts to suffer.
- 100-200 ms: Noticeable. Video calls have a slight delay that makes conversation feel slightly off. Competitive online games become frustrating.
- Over 200 ms: Problematic. Conversations have awkward delays. Fast-paced games are unplayable.
Latency is determined mostly by physical distance (data can only travel so fast through fiber optic cables, roughly 2/3 the speed of light) and by the number of network hops between you and the server. A speed test to a server in your city might show 10 ms, while a connection to a server on another continent might show 150 ms. Neither result is “wrong”; they’re measuring different routes.
Jitter: Why Your Call Keeps Cutting Out
Jitter is the variation in latency over time. If your latency is a consistent 30 ms, things feel smooth. If it bounces between 15 ms and 200 ms unpredictably, real-time applications suffer.
Jitter is particularly destructive for video calls and voice calls because these applications need data to arrive at a steady pace. When packets arrive at wildly varying intervals, the application has to buffer more aggressively (adding delay) or drop packets entirely (causing audio glitches and video freezing).
Common causes of high jitter:
- Wi-Fi congestion: the most common cause in homes. Other devices, neighboring networks, and physical obstacles all cause wireless signal quality to fluctuate.
- Network congestion: your ISP’s network may be overloaded during peak hours.
- Background downloads: a large download on your network can cause latency spikes for everything else.
If your video calls are intermittently bad rather than consistently bad, jitter is likely the issue, not raw bandwidth.
Why Speed Test Results Vary
If you run the same speed test three times in a row, you’ll often get three different numbers. This isn’t because the test is broken. Network conditions genuinely fluctuate moment to moment.
Factors that affect your results:
Server selection. Speed tests connect to a specific server. A nearby server will give different results than one across the country. Most tests auto-select the “best” server, which tends to give you an optimistic result that may not reflect your experience connecting to services hosted elsewhere.
Time of day. Cable internet connections share bandwidth among neighbors. Evening hours (roughly 7-11 PM) typically show lower speeds as more people in your area stream video and use the internet.
Wi-Fi vs. wired. If you’re testing over Wi-Fi, your result reflects both your internet connection and your wireless link. A poor Wi-Fi signal can make a fast internet connection look slow. Plug in over Ethernet and you get a cleaner measure of your actual internet service.
Your device. Older devices, devices with many background apps running, or devices with outdated network drivers can bottleneck the test before your internet connection does.
What to Actually Do With Your Results
Run the test a few times at different times of day, ideally on a wired connection if possible. Look at the consistency of results as much as the peak numbers. If you’re getting 80% or more of your plan’s advertised speed on a wired connection, your service is working normally. If you’re consistently getting much less, call your ISP.
For troubleshooting specific problems, match the symptom to the metric. Buffering video is usually bandwidth. Laggy games or calls are usually latency or jitter. Choppy outgoing video is usually upload speed.
Try the Internet Speed Test on ToolzHQ to check your connection. Now you’ll actually know what the numbers mean.