A speed test measures how fast and how reliably data moves between your device and the internet. Beyond a single headline number, the metrics that matter — download, upload, latency, and jitter — together explain why streaming buffers, calls stutter, or pages feel slow.
A speed test sends and receives data to a nearby server and times the result. It reports four core figures: download speed, upload speed, latency, and jitter. Download and upload describe bandwidth — how much data fits through your connection per second — while latency and jitter describe responsiveness. A connection can have huge bandwidth yet still feel sluggish if its latency is high, which is why all four numbers matter.
Download speed measures how quickly data travels from the internet to your device, expressed in megabits per second (Mbps). It governs how fast pages load, how quickly files arrive, and what video quality you can stream without buffering. For most households this is the headline number providers advertise. Streaming 4K video comfortably needs roughly 25 Mbps, while a busy household with several users benefits from 100 Mbps or more.
Upload speed measures how quickly data goes from your device to the internet. It matters for video calls, sending email attachments, cloud backups, online gaming, and publishing content. Many home plans, especially cable, provide far lower upload than download speeds, which is fine for browsing but can cause frozen video or laggy screen sharing in meetings. Fiber connections typically offer symmetric speeds where upload matches download.
Latency, or ping, is the round-trip time for a small packet to reach a server and return, measured in milliseconds. Low latency means a responsive connection where clicks and actions register instantly; high latency causes noticeable lag even on fast connections. It is the single most important metric for online gaming and video calls, where a delay of even 100 milliseconds is felt by the user.
Jitter is the variation in latency from one packet to the next. Even with a low average ping, high jitter makes real-time applications unstable because data arrives unevenly, causing audio dropouts and stuttering video. Packet loss, where data never arrives and must be resent, has a similar effect. Stable, consistent timing often matters more for calls and gaming than raw bandwidth.
Many factors shape real-world speed: your plan's limits, Wi-Fi signal strength, the age of your router, network congestion at peak hours, and the distance to the test server. A VPN adds encryption overhead and an extra hop, so it usually reduces measured speed. For the most accurate baseline, test over a wired connection, close bandwidth-heavy apps, and run the test a few times to average out fluctuations.
A speed test measures the live path between your device and a specific nearby test server, so the figure is a snapshot of that route at that moment, not a fixed property of your connection. The distance to the chosen server, its current load, and the short duration of the test all shape the result, which is why two back-to-back runs — or two different test services — can disagree. Treat any single number as one sample: a few runs against a close server give you a realistic range, and a reading far below your plan only means something once a distant or busy server has been ruled out.
Speed tests report megabits per second (Mbps), while file sizes and download managers use megabytes per second (MBps), and the two differ by a factor of eight, because one byte is eight bits. So a 100 Mbps connection downloads at roughly 12.5 MBps at best. This is the most common reason people think their connection is eight times slower than advertised — it is simply the unit difference. Knowing the conversion lets you sanity-check whether a slow download is your connection or something else, such as a throttled server.
It depends on use. Basic browsing and HD video work on 10–25 Mbps download, 4K streaming wants around 25 Mbps per stream, and a busy multi-user home is comfortable at 100 Mbps or more. For calls and gaming, low latency and jitter matter more than a high download number.
Advertised speeds are maximums under ideal conditions. Wi-Fi interference, an older router, peak-hour congestion, background downloads, and distance to the server all reduce real throughput. Testing over Ethernet with other devices idle gives a truer picture; if wired speeds are still far below your plan, contact your provider.
Ethernet almost always tests faster and more consistently because it avoids wireless interference and signal loss. Wi-Fi is convenient but its speed drops with distance, walls, and competing devices. For an accurate measurement of what your plan actually delivers, run the test on a wired connection whenever possible.
Usually a little. A VPN encrypts your traffic and routes it through an extra server, which adds overhead and distance. The slowdown is often modest with a nearby, high-quality server but grows when the server is far away or heavily loaded. Testing with the VPN on and off shows the real impact.
Run the test on a wired connection if you can, close other apps and pause background downloads, and repeat it a few times rather than trusting one reading. Wi-Fi, other devices, and time-of-day congestion all affect the result. Comparing several runs gives a realistic range, and testing against a nearby server reflects your connection rather than distant network conditions.
Mbps is megabits per second — the unit internet plans and speed tests use — while MBps is megabytes per second, the unit downloads are shown in. Because one byte is eight bits, you divide Mbps by eight to get MBps: a 100 Mbps connection tops out around 12.5 MBps. That mismatch is why downloads look far slower than your plan's headline number.
