What a speed test actually measures
The principle is straightforward: your browser exchanges packets of data with a remote server for a few seconds, and the tool measures how much gets through in each direction. Download speed is how fast data comes in. That's the number that determines whether a 4K video streams smoothly or whether a page takes forever to load. Upload speed is the reverse, how fast you send data out. It matters most for video calls, live streaming or sending large files.
A speed test also reports latency, often called ping. That's how long a packet takes to make a round trip between your device and the server. The lower the number, the more responsive your connection feels. For regular browsing, anything under 40 ms is unnoticeable. For online gaming, you really want to get below 20 ms if you can.
Why the internet speed test doesn't match your plan
This is the question everyone asks after their first test. You're paying for gigabit fiber, the internet speed test shows 300 Mbps, and you start wondering if your provider is shortchanging you. Short answer: probably not. The speed advertised in your contract is a theoretical maximum measured under ideal conditions that almost nobody actually has at home.
Several things eat into your real-world speed. Wi-Fi is the biggest one. Even with a good router, Wi-Fi can easily cut your internet speed test result in half or worse compared to a wired Ethernet connection. Distance, walls, interference from neighboring networks, all of it adds up. Your hardware matters too. An older laptop with a 100 Mbps Ethernet port will cap at 100 Mbps no matter how fast your fiber line actually is. And if someone in the house is streaming Netflix during your test, of course the bandwidth gets split.
Wi-Fi vs Ethernet: what your speedtest really measures
If you want to measure the actual performance of your line, plug an Ethernet cable directly from your router to your computer. End of story. Running a speedtest over Wi-Fi measures your Wi-Fi network, not your internet plan. Those are two completely different things, and people regularly confuse them when troubleshooting.
Wi-Fi introduces too many variables. The frequency band (2.4 GHz vs 5 GHz), the standard your device supports (Wi-Fi 5, Wi-Fi 6, Wi-Fi 7), the distance from your router, congestion from neighboring networks. A speedtest over Wi-Fi will always be lower and more variable than one over a wired connection. If you get good results on Ethernet but poor ones over Wi-Fi, the problem is with your Wi-Fi setup, not your ISP.
When you should run a wifi speed test
A wifi speed test gives you the most realistic measure of what you actually experience day to day, since most people use Wi-Fi for almost everything. But the result is a snapshot, not a permanent score. Run a wifi speed test from your kitchen, and then again from your bedroom, and you'll likely see two very different numbers. The further you are from the router, the slower it gets, sometimes dramatically so in older homes with thick walls.
For a fair picture of your overall network, run a few wifi speed test sessions at different times and in different rooms. Pay attention to the location where you actually use the connection the most. If your kitchen drops to 30 Mbps but your living room hits 200 Mbps, the problem isn't your plan, it's how the signal reaches the room where you're working. Mesh systems, range extenders or even just relocating the router can make a much bigger difference than upgrading your subscription.
The right way to test internet speed
The most common mistake when people test internet speed is doing it once and treating the result as gospel. A speed test reflects your connection at a specific moment, not its average performance. Running it on a Sunday evening at 9 PM when half the neighborhood is streaming will give you a different number than Tuesday morning at 10 AM. Both are valid measurements, they just tell different stories.
To get a reliable picture, test internet speed several times throughout the day, ideally over a couple of days. Before each run, close any browser tabs loading content in the background, pause downloads and check that nobody else on the network is streaming or gaming. Otherwise you're not really measuring your line, you're measuring whatever bandwidth is left over from everything else happening at the same time.
Fiber, DSL, 4G, 5G: what to expect from each
DSL connections typically return download speeds between 5 and 15 Mbps in a wifi test or wired test. That sounds slow but it's enough for browsing, HD video and voice calls. Upload tops out under 1 Mbps in most cases, which makes video calls painful. If you're on DSL and your wifi test shows 8 Mbps down, your line is probably fine, that's just where the technology caps.
Fiber-to-the-home (FTTH) ranges from 300 Mbps to over 2 Gbps depending on your plan. A speed test will commonly show 400 to 800 Mbps on a 1 Gbps plan over Ethernet. On 4G, results vary wildly with antenna proximity and network load, anywhere from 10 to 150 Mbps. 5G can go significantly higher but depends heavily on coverage and whether you're on the high-band or low-band variant.
Latency, the metric nobody talks about
When people run an internet test, they almost always focus on the download number. Latency gets ignored. But in a lot of situations, latency is what actually shapes the experience. You can have 500 Mbps of bandwidth with 80 ms of latency, and your video calls will feel worse than someone on a 50 Mbps line with 10 ms of latency.
Latency depends largely on your connection type. Fiber typically runs 3 to 15 ms. DSL falls between 20 and 50 ms. 4G sits between 30 and 80 ms with more variability. Satellite is in another league entirely, often 500 ms or more, which makes VoIP calls genuinely uncomfortable. If your internet test consistently shows high latency, the issue usually lies with your provider's network rather than your home setup.