If you've been told to check your speed and you're seeing a great number, it's natural to think the problem must be at my end. Usually it isn't — and there are two reasons why, both of which are easy to explain once you know what's going on.

The short version

• A general speed test measures the wrong thing. It almost always tests against a server inside your own internet provider's network — a short, fast, empty road. Your streaming data doesn't take that road. It travels overseas to Mediadrive servers, across the public internet.
• Netflix and similar services are not a fair comparison. They keep copies of their content inside your provider's network, and they automatically drop your quality to keep playing without interruption. "It works fine" can genuinely mean "it quietly dropped to a much lower quality".
• Mediadrive servers send the original file at its original quality. Files are often 2–10× the bitrate of the same title on a streaming service.

So a fast speedtest result is good news — it means your line is healthy — but it does not tell you how fast data moves between you and Mediadrive servers.

The longer version

Where a speed test actually measures

When you run a speed test, it automatically picks the closest server with the lowest latency. In most cases that's a server sitting inside your provider's own network. Your test data travels a few short hops and never leaves — it doesn't cross the international links at all.

That's why a speed test is excellent at answering "is my line delivering what I pay for?" — and why it tells you almost nothing about a server on the other side of the world.

It also opens several connections at once and adds them up. That finds the maximum your line can do, but a video stream is one connection. On a long-distance path, a single connection is limited by how long it takes data to get there and back — the further away the server, the lower the ceiling, even on a very fast line. It's entirely normal for a 500 Mbps line to achieve a fraction of that to a distant server.

Why Netflix, YouTube and similar services are different

Two things make them look effortless — and neither is available to a small independent service.

1. They keep copies inside your provider's network.

Netflix runs its own delivery network called Open Connect. These are physical servers installed inside internet providers' own facilities — your provider supplies the rack space and power, Netflix operates the hardware. Netflix's own figures put the hit rate around 98%: your stream is served from a box a few kilometres away, and the data never touches international transit at all.

Mediadrive servers are in Germany. There's no way to put a copy in your local exchange — the traffic has to cross international links, and that's the part of the journey a speed test never tests.

2. They drop your quality automatically to avoid interruption.

This is the big one, and most people don't realise it's happening.

Streaming services pre-encode each title at many different quality levels. As you watch, the player constantly checks how the data is flowing. If it starts to struggle, it switches to a lower-quality version mid-stream rather than stopping. It recovers when it can. You'll often see a brief softness or blocking and never know why.

So "Netflix works fine but your service buffers" frequently means: Netflix quietly reduced your quality and kept going. Mine tried to keep the original quality and stalled instead.

3. Their quality is lower to begin with.

Netflix's own ISP Speed Index — their published measurement of what they actually deliver in prime time — puts the average Netflix stream at around 3 Mbps. That's the honest yardstick, because it's measured delivery rather than a headline ceiling:

An ordinary movie here carries roughly three times the data of the average Netflix stream, and an episode about 1.6×. Their "it looks great" is achieved partly by sending less data in the first place.

Add those three together and the comparison collapses: they're serving a smaller file, from much closer, and silently dropping it further if it struggles.

What actually determines your streaming speed

Your data travels through a chain of separate networks — your provider, their upstream carrier, an exchange point, international links, and finally Mediadrive servers. Each of those hops is a shared link used by many other people, with its own capacity and its own busy periods.

Think of it as driving. A speed test measures the street outside your house. What actually matters is the motorway 400 km away — and how busy it is at that moment. You can have a fast car on an empty driveway and still crawl once you join a congested highway. Nothing at your end is wrong; the road simply has more traffic on it.

And think of it as a flight with connections. There is rarely a direct flight, so your data goes through whatever connections the networks have agreed on that day — you might fly via Singapore, or via Los Angeles. Two people leaving the same city on the same day can be routed completely differently, and arrive with very different delays.

Most of the links are fine most of the time. But if any single one of them is congested, your speed drops — and it drops to whatever that one link can manage. Nothing in your house needs to be wrong for that to happen, and a speed test that never crosses those links will never see it.

This is also why evening congestion is common, and why some providers are noticeably better than others on this specific route.

"It was working fine before and nothing changed"

This is the one that catches people out — and it's one of the most common reasons a service that worked for months suddenly doesn't.

The route is not fixed. The path your data takes is chosen dynamically by the networks in between, and your provider can change it at any time — because of congestion, a capacity upgrade, a commercial agreement, or a cable fault on the other side of the world. Nobody tells you, and nothing in your home changes at all. The part that changed is the part neither of us controls.

This affects Australia more than most countries, because there is no land route out. Every bit of international traffic leaves by submarine cable, and there are only a handful of routes to choose from — so when one becomes busy or gets rerouted, there is far less spare capacity to absorb it.

For a destination like Germany the difference is measurable. The best routes from Australia run through Asia at roughly 280 ms round-trip. If your provider's traffic is sent the long way — out through the USA, then back across the Atlantic — that figure rises sharply, and throughput falls with it. There are documented cases of Australian traffic bound for Singapore being routed via the USA and Europe instead of directly, multiplying the latency several times over.

So if your connection was fine last month and isn't now, and nothing in your home has changed — that is consistent with a routing change, not with anything you did. It's worth mentioning when you contact me, because it points at the route rather than your equipment.

So what does help?

• Compare devices — if one device is much slower than another in the same room, that's Wi-Fi, not the server — see Buffering & Speed
• Test on mobile data — if it plays fine on 4G/5G but not on your home connection, the problem is the home connection or how it reaches overseas services.
• Tell me what you're seeing — if it affects one title, it may be that file. If it affects everything and only at certain times, it's the route. I can sometimes adjust how traffic reaches the servers.

Want to read more?

If you'd like to understand this better — or you're the sort of person who wants to check it for yourself rather than take my word for it — these are worth your time. None of them are mine, so they aren't written to make my case:

Why a speed test isn't the whole story Why Internet Speed Tests Don't Really Matter (and What Does) — written in plain English, and uses much the same analogies as this page: a speed test is like testing a car on a closed race track, or shipping a parcel that reaches the local depot instantly but takes days to cross the country.

What a speed test actually measures (Australian) Internet Speed Tests Explained — the clearest short explanation of the point this whole page makes: a speed test measures the connection between your device and that one test server, at that moment — while the streaming service you're actually using might be on the other side of the world.

If you want to go deeper — how networks connect What is Internet Peering? — from the Internet Society. Explains how providers agree to exchange traffic and why the path your data takes is a matter of commercial agreements between networks, not something either of us controls. A little more technical than the two above, but still readable.