The plan says 500 Mbps. The test says 180. Neither number is a lie, and the gap between them is mostly definitional — advertised speed is a carefully worded claim about a maximum under conditions that do not describe a normal home.
What the advertised number means
In most markets the headline figure is a theoretical line rate: the maximum the technology can carry to the premises, measured at the point the line enters the building, with nothing else running. Advertising rules in several countries additionally allow “up to” claims based on what a minority of customers achieve.
Everything between that point and your device — the router, the Wi-Fi, the device’s own network hardware — sits outside the claim entirely.
The five places speed disappears
| Stage | Typical loss |
|---|---|
| Protocol overhead | 5–15% — unavoidable; headers are not payload |
| Line quality and distance | Large on copper, negligible on fibre |
| Contention at peak | Highly variable — shared local capacity |
| Your router | Older hardware caps well below gigabit |
| Wi-Fi | Usually the biggest single loss in the chain |
The bottom row dominates. Wi-Fi is a shared half-duplex medium competing with every neighbouring network on the same channels; real-world throughput of half the advertised wireless rate is normal even in good conditions.
READ ALSO: Suchbegriffe als Frühwarnsystem für lokale Google-Ads-Kampagnen
Speed is not the number that determines the experience
Above about 50 Mbps, additional bandwidth stops being noticeable for ordinary use. What people experience as “slow internet” past that point is almost always latency, not throughput.
Three measurements matter more than the headline figure:
- Latency. Round-trip time. A ping test under 30 ms feels immediate; over 100 ms feels sluggish regardless of bandwidth.
- Jitter. Variation in latency. This is what breaks video calls — a stable 80 ms is better than an average of 40 ms that swings wildly.
- Loaded latency. Latency measured while the connection is busy. A line that shows 20 ms idle and 400 ms under load has bufferbloat, and it will feel broken whenever anyone is downloading.
Loaded latency is the single most useful and least advertised measurement in home broadband.
Building a case your provider cannot wave away
- Test wired. A provider will dismiss any Wi-Fi measurement, correctly.
- Test repeatedly over several days, including evening peak. One result proves nothing.
- Record times and figures. A log of twenty timestamped results is evidence; a screenshot is an anecdote.
- Compare against the minimum guaranteed speed, not the headline. Most contracts specify one, and it is the number that carries any obligation.
- Check what others on the same provider see. Aggregated ISP speed test results show whether your experience is typical or an individual fault, and those are different conversations.
Choosing between providers
Advertised speeds are close to useless for comparison because every provider quotes them the same way. Aggregated measurements from real customers are more informative, and ISP rankings built from them show the two things advertising never mentions: how much the evening figure falls below the daytime one, and how consistent latency is.
A provider delivering a steady 200 Mbps with low jitter at 21:00 gives a better experience than one advertising 900 Mbps that collapses every evening — and only the second kind of data will tell you which is which.
