Broadband News
News, views and analysis
Dutch broadband speeds don't measure up
18 Dec 2009 | 14.50 Europe/London
The Netherlands may be one of Europe’s leading broadband nations, but it’s suffering from a familiar problem – the actual broadband speeds received by consumers are significantly lower than advertised. A study carried out by Telecompaper in partnership with iPing shows that Dutch subscribers currently receive about 60 percent of the average advertised broadband speeds.
The results echo findings from a study carried out earlier this year by Ofcom in conjunction with technical partner Samknows, which showed that actual broadband speeds in the UK are also substantially below advertised speeds – and consumer expectations.
According to the report from Telecompaper, the average download speed achieved in the Netherlands is 8.98 Mbps, compared to an average speed offered by the ISPs of 14.9 Mbps. Previous Telecompaper studies found the average actual download speed was 4.95 Mbps in December 2008, and 2.8 Mbps in November 2006. But while speeds have been improving in leaps and bounds, the gap between the service being advertised and the consumer’s experience has been maintained.
The tests were carried out by software on the consumers’ computers, which automatically checks the speed of broadband. More than 2 million tests on 20,000 users were generated over the three month period from July to September. Owing to the set up, the speed test checks the actual speed perceived by the end-user, which may include other factors like the wireless connection from the router to the home computer, explains Ed Achterberg, senior research analyst with Telecompaper and author of the report.
Many households in the Netherlands have a broadband connection in the utility closet – a point where services like gas, electricity and telephone lines enter the property. From that point the signal is delivered over a wireless network, and can be affected by the position of the computer in the house, and even by the type of concrete from which the building is constructed, he says.
Alex Salter from Samknows commented:
One of the interesting things about the Dutch broadband study is that it collects speed data about fibre-to-the-home connections for the first time. Fibre achieved the best average download speed in the study by quite a high margin, at 40 Mbps, followed by cable with 14.43 Mbps and DSL with 5.81 Mbps. Fibre and cable also achieved the highest average performance ratios, delivering respectively 62 percent and 65 percent of average offered speeds, while DSL averaged just 55 percent of advertised speeds.
But wait a minute, isn’t fibre supposed to deliver the maximum speed available on the line. In this case it doesn’t, and that could be related to the fact that the tests are performed at the computer. “We know that some providers deliver the maximum speed that is included in the package, but that’s not always received on the PC or desktop,” says Achterberg.
In Achterberg’s opinion, service providers need to be aware that the service they deliver depends on elements in the complete chain, which are outside their control but that impact upon the perception of the performance. Therefore, if a fibre-to-the-home network includes poor in-home wiring, or a badly positioned router or modem, this could detract from the consumer’s experience.
If you’re Dutch, then you may be interested in the fact that cable operators seemed to come out on top – at least as far as download speeds are concerned (upload and latency were not measured). Ziggo offers the best performance in the low-end segments of 0-1.5 Mbps and 1.5-3 Mbps, as well as in the 6-12 Mbps segment. Meanwhile, UPC scored best for high-speed, coming out top in the 12-24 Mbps, 24-48 Mbps and 48+Mbps segments, and tied with Tele2 for the best performance in the 3-6 Mbps segment.
The results echo findings from a study carried out earlier this year by Ofcom in conjunction with technical partner Samknows, which showed that actual broadband speeds in the UK are also substantially below advertised speeds – and consumer expectations.
According to the report from Telecompaper, the average download speed achieved in the Netherlands is 8.98 Mbps, compared to an average speed offered by the ISPs of 14.9 Mbps. Previous Telecompaper studies found the average actual download speed was 4.95 Mbps in December 2008, and 2.8 Mbps in November 2006. But while speeds have been improving in leaps and bounds, the gap between the service being advertised and the consumer’s experience has been maintained.
The tests were carried out by software on the consumers’ computers, which automatically checks the speed of broadband. More than 2 million tests on 20,000 users were generated over the three month period from July to September. Owing to the set up, the speed test checks the actual speed perceived by the end-user, which may include other factors like the wireless connection from the router to the home computer, explains Ed Achterberg, senior research analyst with Telecompaper and author of the report.
Many households in the Netherlands have a broadband connection in the utility closet – a point where services like gas, electricity and telephone lines enter the property. From that point the signal is delivered over a wireless network, and can be affected by the position of the computer in the house, and even by the type of concrete from which the building is constructed, he says.
Alex Salter from Samknows commented:
“I would stress that the only way to accurately measure a broadband provider's performance is by running tests 24x7 whilst there is no other traffic from the home network. Only then are you actually testing the ISP rather than an individual's home network, which may be being used by someone else in the house.”
One of the interesting things about the Dutch broadband study is that it collects speed data about fibre-to-the-home connections for the first time. Fibre achieved the best average download speed in the study by quite a high margin, at 40 Mbps, followed by cable with 14.43 Mbps and DSL with 5.81 Mbps. Fibre and cable also achieved the highest average performance ratios, delivering respectively 62 percent and 65 percent of average offered speeds, while DSL averaged just 55 percent of advertised speeds.
But wait a minute, isn’t fibre supposed to deliver the maximum speed available on the line. In this case it doesn’t, and that could be related to the fact that the tests are performed at the computer. “We know that some providers deliver the maximum speed that is included in the package, but that’s not always received on the PC or desktop,” says Achterberg.
In Achterberg’s opinion, service providers need to be aware that the service they deliver depends on elements in the complete chain, which are outside their control but that impact upon the perception of the performance. Therefore, if a fibre-to-the-home network includes poor in-home wiring, or a badly positioned router or modem, this could detract from the consumer’s experience.
If you’re Dutch, then you may be interested in the fact that cable operators seemed to come out on top – at least as far as download speeds are concerned (upload and latency were not measured). Ziggo offers the best performance in the low-end segments of 0-1.5 Mbps and 1.5-3 Mbps, as well as in the 6-12 Mbps segment. Meanwhile, UPC scored best for high-speed, coming out top in the 12-24 Mbps, 24-48 Mbps and 48+Mbps segments, and tied with Tele2 for the best performance in the 3-6 Mbps segment.