You pay for a 500 Mbps internet plan. Your speed test shows 487 Mbps. You start downloading a 1GB file and notice it's downloading at 45 MB/s. Is something wrong? Why isn't it downloading at 500 MB/s if you're paying for 500 Mbps?
The answer is a fundamental units difference that ISPs exploit — and once you understand it, every speed test result makes perfect sense.
The Bits vs. Bytes Confusion
Internet speeds are measured in Megabits per second (Mbps). File sizes are measured in Megabytes (MB). The critical difference: there are 8 bits in one byte.
So to convert your speed test result to the download speed you'll see in a file manager:
Actual download speed (MB/s) = Mbps ÷ 8
Examples:
- 100 Mbps ÷ 8 = 12.5 MB/s
- 500 Mbps ÷ 8 = 62.5 MB/s
- 1,000 Mbps (1 Gbps) ÷ 8 = 125 MB/s
If you have a 500 Mbps plan and see 60 MB/s in your download manager, that's actually excellent performance — you're getting 480 Mbps, 96% of your plan's advertised speed.
Why ISPs Advertise in Mbps, Not MB/s
It's purely a marketing choice. Mbps numbers are 8× larger than the equivalent MB/s figure."500 Mbps" sounds significantly faster than"62 MB/s," even though they describe the same connection. Most regulators don't require ISPs to advertise in consumer-friendly units.
What"Up to X Mbps" Actually Means
ISPs advertise"up to" speeds, not guaranteed speeds. Under optimal conditions — direct Ethernet connection to the modem, off-peak hours (2–5 AM), no devices sharing bandwidth — you might reach the advertised peak. In typical usage, expect 60–90% of the advertised rate under good conditions, and 30–60% during peak evening hours.
In most markets, ISPs are required to publish actual average speeds versus advertised speeds. In the US, the FCC's"Measuring Broadband America" program documents the gap. In the EU, BEREC regulations require ISPs to publish minimum, average, and maximum speeds.
The Real Reasons Your Speed Is Lower Than Advertised
1. WiFi Signal Loss
WiFi introduces protocol overhead (20–40% of raw radio throughput is consumed by protocol headers and acknowledgment packets) plus physical signal attenuation. Through one concrete wall at 10 meters, a 5GHz WiFi 6 router may deliver 200–300 Mbps to a device on a 1 Gbps plan. This is not an ISP problem — it's a WiFi physics problem.
2. Network Congestion
Residential ISPs oversell their infrastructure capacity — the assumption is that not all customers use their full allocation simultaneously. On a busy Friday evening, shared fiber or cable segments may be 80% saturated, slowing all users on that segment.
3. Server-Side Limits
The other end matters too. If you're downloading from a server that caps transfer speed at 100 Mbps, you can't download faster than 100 Mbps regardless of your 1 Gbps connection. Netflix, YouTube, Steam, and other major platforms have high-throughput CDN infrastructure. Random websites may not.
4. Your Hardware
Old routers often have 100 Mbps or Fast Ethernet ports — not Gigabit. Your network adapter in a laptop may only support 100 Mbps. Older USB WiFi adapters may cap at 150 Mbps. These create hard ceilings regardless of your ISP plan.
5. ISP Infrastructure Issues
Genuinely degraded service — line noise, congestion at the ISP's peering points, modem provisioning errors — can reduce speeds below what you should receive. Running multiple speed tests at different times of day (morning vs. evening, weekday vs. weekend) helps identify this pattern.
How to Get the Most Accurate Speed Test Result
- Connect via Ethernet directly to your modem (bypass the router) if possible.
- Test during off-peak hours (early morning).
- Close all other applications and browser tabs.
- Run the test 3–5 times and take the average.
- If using WiFi, position yourself as close to the router as possible in the same room.
Use the ToolsPixel Internet Speed Test, powered by Cloudflare's global edge network (300+ locations worldwide), to get an accurate benchmark with download, upload, ping, and jitter measurements in one free test.
What Is a Good Speed for Your Needs?
- Single user, streaming only: 25–50 Mbps
- Single user, working from home with video calls: 50–100 Mbps
- Family of 3–4 with mixed use: 200–300 Mbps
- Power users, 4K streaming on multiple devices, gaming: 500 Mbps+
- Prosumer / small business content creators uploading large files: 500 Mbps+ symmetric (fiber required for high upload)
The most common mistake is buying a faster plan when the real bottleneck is old WiFi hardware. Before upgrading your ISP plan, test whether a newer WiFi 6 router or a direct Ethernet connection closes the gap.
Understanding Ping, Jitter, and Packet Loss
Download and upload speeds aren't the only metrics that matter. The full speed test also measures:
- Ping (latency): The round-trip time for a data packet to reach a server and return, measured in milliseconds. Under 20ms is excellent; 20–50ms is good for gaming; over 100ms causes noticeable lag in real-time applications.
- Jitter: The variation in ping times between successive packets. High jitter (over 30ms) causes choppy voice and video calls even when average ping is acceptable. Jitter is caused by network congestion causing inconsistent packet routing.
- Packet loss: The percentage of data packets that fail to reach their destination. Even 1% packet loss significantly degrades TCP/IP performance because the protocol retransmits every lost packet, effectively stalling the connection during retransmit waits. 0% is expected on a healthy connection.
If your download speed is fine but video calls stutter, high jitter is often the culprit — not low bandwidth. Run a complete test to see all five metrics before diagnosing a connection problem.