Average world broadband subscriber speeds rise 23% in a year

Average world broadband subscriber speeds rise 23% in a year


The average broadband speed of millions of users around the world has gone up in one year by 23% from just over 7Mbps to just over 9Mbps.

The findings come from more than 163 million tests carried out on networks from Singapore – the fastest of 200 countries surveyed last year and this – to Yemen, the slowest in both years. (Results here.)

“The average global broadband speed measured during the period from 11 May 2016 to 10 May 2017 was 7.40Mbps,” says M-Lab, which carried out the research. “The average global broadband speed measured during the period from 30 May 2017 to 29 May 2018 was 9.10Mbps – a rise of 23%. That’s considerable.”

M-Lab is a partnership between the New America Open Technology Institute, Google Open Source Research, Princeton University’s PlanetLab and other supporting partners, and the results were compiled by Cable, a UK-based telecoms comparison site.

Dan Howdle, consumer telecoms analyst at Cable, said: “Europe, the US and thriving economic centres in the Asia-Pacific region – Singapore, Japan, Taiwan and Hong Kong – are leading the world when it comes to the provision of fast, reliable broadband, which suggests a relationship between available bandwidth and economic health. Those countries leading the world should be congratulated, but we should also be conscious of those that are being left further and further behind.”

Within Cable’s spreadsheet of the findings are rises and falls by individual countries that will get politicians to congratulate themselves or provide excuses.

Romania went up 13 places between the two surveys, and is now fifth in the world in terms of mean download speed – 38.6Mbps in more than 175,000 tests. The previous test showed 21Mbps, so the absolute rise was good – but other countries rose faster.

Hong Kong’s mean download speed went down from 27.16Mbps to 26.45Mbps, and the special autonomous region of China went down 10 places in the league.

After Singapore in the lead, the next six top countries for speed are all European – Sweden, Denmark, Norway, Romania, Belgium and the Netherlands.

The US is at twentieth place, up one from the last survey, with a mean speed now of 25.86Mbps. The UK is at 35, down four places, with a mean speed in the latest survey of 18.57Mbps.

“A number of other countries have leapfrogged [the UK] since last year, including France and Madagascar,” said Howdle. “Compared to many other countries both in and out of Europe, the UK has simply come too late to a full fibre solution, relying instead on copper to cover the last mile. Despite plans to roll out full fibre to UK homes across the next decade or so, the UK is likely to fall even further behind while we wait.”

This point was seized upon by Greg Mesch, CEO of CityFibre, which is building rival fibre networks in a number of cities in the UK. He called the fall “depressing but not surprising, given the UK’s lack of investment in fibre to the premises and other nations’ new networks increasingly coming online”.

He added: “This situation must change – and quickly – as successfully rolling out this superior digital infrastructure is critical for the success of the UK economy and our ability to compete internationally.”

He called for new rules to stop rivals using the term “fibre” when the last connection to premises is still copper, “so that consumers know that when they see fibre, it is a fibre to the premises connection they are buying. Copper is dead.”

Cable, which analysed the results, pointed out that “Ofcom, the UK telecoms regulator, as well as many regulatory and governmental bodies around the world, uses physical equipment to constantly measure the maximum speed available on particular lines across a long period of time, and it is from this they will derive their average.” These results, however, used regular speed tests. This “explains why Ofcom’s average speeds measured in the UK, for example, are higher than those measured in our tests”, said Cable. “The M-Lab data, and therefore the national averages presented in this report, should be regarded as a realistic reflection of real-world user experience rather than an absolute measurement of available bandwidth.”

Gift this article