The claim that over 99 percent of data traffic runs through gets quoted everywhere, usually with no source attached. TeleGeography went looking for where it came from and found it traced back to their own FAQ, which rested on filings. The figure holds up. The data behind it stopped being collected a decade ago.
The trail is short and a little awkward. The number came from a TeleGeography FAQ citing FCC data, which showed accounting for a tiny share of United States international capacity. The most recent FCC data of that kind is from 2013 and reached the public in 2015, and the reporting requirement later changed, so nothing newer exists. TeleGeography can confirm the claim is true but cannot calculate it precisely.
Almost every article about internet infrastructure contains a version of the same sentence: over 99 percent of intercontinental data traffic travels through submarine cables. It is quoted confidently and almost never with a citation. TeleGeography, a research firm that tracks exactly this kind of infrastructure, went looking for the original.
The trail led back to them. The claim traces to a TeleGeography FAQ, which in turn rested on FCC filings showing that satellites accounted for about 0.37 percent of United States international capacity. On those numbers, cables carrying well over 99 percent of the traffic is the obvious conclusion.
The awkward part is the vintage. The most recent FCC data of that kind covers 2013 and reached the public in 2015. The reporting requirement then changed, so no newer series exists. TeleGeography’s own conclusion is careful: they can confirm the claim is true, but they cannot do precise calculations without data on intercontinental satellite traffic.
That leaves the figure in an unusual position. It is almost certainly right, because satellite capacity is nowhere near that of fiber on the seabed. But it is no longer a measurement. It is an inference from a decade-old filing, repeated until it sounds like a current reading. The distinction is worth keeping in mind when you see it cited.
wat. The best-known statistic about the internet rests on filings nobody collects any more.
wat wat
A statistic can be both true and unverifiable. This one survives on a reporting requirement that lapsed, which is worth knowing before anyone quotes it as a current measurement.
How do we know?
TeleGeography’s own mythbusting post traces the claim to its source, names the FCC filings it rested on, gives the satellite share those filings reported, and states plainly what can and cannot be calculated now.
What is still unknown
A post by a telecom research firm about its own widely quoted figure, so it is both the best available account of where the claim came from and an interested party. The underlying satellite traffic data does not exist in public form.
Learn something?
Where to next
What was the first webcam pointed at?internet
What was in the first email ever sent?internet
What makes a whirlpool in the open sea?oceans