My broadband experience in Europe, Asia, Latin America and the US

This is an article about my experience with broadband service during my travels in Europe, Asia, Latin America and the United States and what I consider to be the important factors (often ignored by commentators) that determine the speed and price of broadband in different countries.

In recent weeks, I have read several articles claiming that broadband service is better in the US than in Europe; others claim it’s the other way around. Some insist that the fastest and cheapest broadband is in Asian cities such as Singapore. The situation in the real world is more complicated than what these articles would have you believe. Even within a country, the speed and price of broadband service differ dramatically.

(1) In my experience, both in Europe and the US, the dividing line between lousy and amazing broadband is urban (dense) versus suburban/rural.

In the San Francisco Bay Area, my broadband experience has been vastly different from that of my friends who reside just outside San Francisco. From 2008-2011 I was living in an apartment building a few blocks from downtown San Francisco. This building was served (and continues to be served) by an ISP called Webpass. It provided me with the best wired broadband service I have every had. It was faster than what I currently have in Paris, which is fiber-to-the-home (FTTH) service (promising 100 Mbps downstream and upstream, but delivering barely 35 Mbps downstream). Webpass San Francisco consistently provided me with speeds of 60 to 70 Mbps downstream and 70 to 80 Mbps upstream for $45 per month (they are now charging $50 per month). You cannot get broadband speeds like this just south of San Francisco in the suburbs for $50 per month. Unfortunately Webpass serves buildings in the urban core; they don’t do suburbs. In a community called Redwood Shores (where Oracle’s headquarters are located and which is halfway between San Francisco and Silicon Valley), the best deal for broadband is Comcast, a cable company, where you can get 25 Mbps downstream/5 Mbps upstream (actual speed) for about $35. The most expensive Comcast package — 105 Mbps downstream and 20 Mbps upstream (advertised speeds) — costs over $115. You would think that the areas south of San Francisco where a lot of startups and tech people live, would have broadband service as good and cheap as that in San Francisco, but that’s not the case.

As I’ve observed, population density is an extremely important factor when one is in the business of delivering wired broadband. It’s the difference between being cash-flow positive or going bankrupt. That is why rural areas don’t have many high-speed broadband options and suffer from poor broadband connections, both in Europe and the United States, unless these are subsidised by the local or national government. Of course density alone is not a determining factor but it’s critical. Therefore, to say that broadband service in Europe is better than in the US (or vice-versa) is inaccurate. It depends on where you live.

(2) Another factor that determines the quality of the broadband connection (especially FTTH) seems trivial but is of critical importance: the nature of actual connection from the street into the building and into the apartment.

In Paris, I have a FTTH connection from Orange (France Telecom) that should theoretically give me 100 Mbps symmetrical speed. Shortly after FTTH service was turned on in our building, I was getting only 20-30 Mbps downstream so I called a technician from Orange to come to the apartment to check my connection. Apart from the stress of having to communicate with the technician in French, I was heartbroken to find out that the previous technician who had installed the FTTH box in the apartment (while I was away), had placed it in the kitchen close to the service door that leads to the rear of the building. This is far from the living room and as a result, the FTTH box has to be connected to a wireless repeater and it is from this wireless repeater that I am getting my broadband connection. Had the FTTH box been placed in the living room, my broadband speed would have been much better, according to the technician. Indeed, he moved the wireless repeater next to the door of the kitchen closer to the living room and behold, I started getting 30 to 40 Mbps downstream. But if I close the door, the speed drops down again. The apartment building was built between 1910 and 1930 in the Haussmannien style which means thick walls, kitchens separate from dining and living areas and lots of volume (high ceilings). The thickness of the walls means good sound insulation, but poor wireless transmission. After the technician departed, I inspected the connection from the kitchen to the stairwell and into the street. It looked very 19th century. Details, details. Alas, they all matter when it comes to the quality of your broadband experience.

paris home

The Paris apartment: the FTTH connection to the outside seems as old as the furniture

(3) Now, let’s leave the US and Europe, and venture out to Australia, New Zealand, and Latin America. Here’s my experience in Santiago, Chile where I spent six months in 2011 and in New Zealand and Australia, where I traveled in 2011 and 2012: It’s not just the speed of the broadband service delivered by the ISP in the particular country that matters, but the total capacity of the undersea cable that goes from that country to Europe and North America. Look at this map which shows the undersea cables between the continents.

undersea cables

Undersea cables in Asia-Pacific

Nobody every talks about this but believe me, when you’re in Chile and your favourite website (located in the US or Europe) is loading very slowly, you ask yourself: is it my ISP or is it the fact that when I ping a European website from Chile, I have to go through Argentina and Brazil (or Peru and Colombia), then the US and finally, to Europe? Broadband in Santiago, Chile was very expensive. I paid about US$60 per month for 10-15 Mbps downstream (actual speed). Bear in mind that in Chile, the average household net-adjusted disposable income is 11,039 USD a year, much less than the OECD average of 23,047 USD a year (statistics from the OECD). That’s less than $1000 per month. Rent, food, petrol and education are expensive in Chile so most people cannot afford the kind of broadband connection I had.

The worst broadband experience for me was in New Zealand. They have only one undersea cable to the United States and it is owned by Southern Cross. The other cable goes to Australia. It’s not surprising that Wi-Fi, and broadband in general, in New Zealand is very expensive and metered like precious diamonds. The only place it wasn’t metered and was reasonably fast was in the McCafe (of McDonald’s). In Australia, I encountered the same issue: metered broadband, expensive Wi-Fi. I understand that Australia has a Next-Generation Broadband Network project in the works whose aim is to transport Australians into broadband Nirvana, but unless that includes the deployment of several undersea cables to Asia and the United States, it’s not going to happen.

For Australia, New Zealand and Latin America, the end user’s broadband experience depends upon the number of undersea cables, unless the user simply wants to visit local websites. It’s critical for those regions to invest in building more undersea cables and not just to focus on local broadband deployments.

(4) I did not experience broadband utopia in Singapore either (those undersea cables again, among other things).

Even in Singapore, which is touted as a Broadband Utopia, the ISPs state in their broadband offers that the speeds for accessing websites outside Singapore will be slower (see Growing Pains Plague Singapore Fiber Rollout and Starhub Singapore’s broadband offerings). You can have 100 Mbps to a website in Singapore but only 20 Mbps to a website in the US. My experience in a rented apartment in Singapore for three months is the service was not as good as I would have expected in this so-called Broadband Utopia. I was getting between 5 to 15 Mbps downstream, much slower than what I get in Paris today or what I was getting in San Francisco via Webpass. I do not find Singapore FTTH inexpensive (see Starhub Singapore’s current FTTH plans below) although I had been led to believe otherwise by inaccurate reports in the mainstream press.

starhub fibre plans

Starhub Singapore FTTH: Focus on the assured international speed – even the 1 Gbps gives you only 30 Mbps

Summary: it’s almost a pointless exercise to make a broad comparison between broadband in the US and in Europe because the quality and speed of broadband service varies dramatically even within each continent. When talking about broadband in Asia, Latin America, Australia and New Zealand, the number of undersea cables that connect those regions to Europe and North America affects the broadband experience of people trying to connect to websites outside the regions.

* * * * * *

NEXT UP: The mobile and Wi-Fi experience around the world . . .


  1. Just for fun, take a trip around the Caribbean. That’s where it gets real interesting.

  2. Bhupinder Misra says

    Do you factor in the increase in fees every year? Promotion for 1st year and then double the price after that.
    How does this work abroad?

  3. Many broadband companies have a promotion offer in the first year but they do not double the price in the second year if they are operating in a place that has a lot of competition. In France, for example, Orange fears losing customers to SFR, Bouygues, Free and Numericable (the cable company). So actually Orange ends up lowering the price in some cases, to keep their customers. In the US, where there is a duopoly, you are more likely to find price increases although in a densely populated area, the increase won’t be 2X.

  4. fun article… but it leads us to a very slippery slope on statistics.

    AT&T, in 22 states, provides U-verse, with a current top speed of 24 mbps. It has only 5 million locations (residential and business) with TV capable and out of 76 million locations at best, less than 50% will ever get any upgraded service. At the same time, AT&T has filed to get rid of all of their copper -that is not upgraded — a line in the sand and put customers onto expensive wireless.

    Verizon has the same plan and with 27 million ‘households’ (we don’t know locations), and only 5 million TV households– and with only 14-15 million ‘passed, the rest can be simply ‘shut off’

    Even though Comcast has a service over 100 mbps, Time Warner’s top speed is less than 50mbps, and they don’t overlap and only cover specific areas of the US>

    Also, while some people in some cities have high speed — there are people in FiOS footprint as part of the NYC cable franchise that can’t get the service.

    But this leaves massive gaps in customers who can get fast service, much less competitive TV — which I consider to be a critical part of this picture as the majority of the US buy the triple play when offered.

    Meanwhile The FCC/NTIA goverment maps are all corrupt to hell — My address showed 4 providers — 2 didn’t exist and both of the others claimed 100mbps, which also didn’t exist from the companies (dec 2012 data)

    Meanwhile, 19 states’ legislatures have voted to ban some, if not all muni competitive services.

    So, how do we collect accurate data or figure out what the next steps are when in the US at least, the wired providers are planning on not offering competitive service to at least 50% of their territories?

    Much less compare the rest of the world? I would argue, almost jokingly, that since everywhere has a problem with reporting and actual use, then it’s statistically a wash and we should simply discount every speed given — and that the benchmarks we see are equally as skewed.

    We must also remember that the corrupted data is being used to create public policies. – Everyone quotes the NCTA number that 80% of the US can get 100 Mbps or that 40% are ‘wireless only” – when that number doesn’t include any data applications like DSL – but is used as an excuse to ‘sunset’ the PSTN .

  5. Excellent snapshot that provides a great insight into the actually bb experiences rather than the media and political debates that dominates this market.

  6. Esme,
    What a great article, spoken from someone who knows. I hung on your every word and had no surprise when you hung “Worst” on New Zealand. You and I have discussed it before. The challenge is never building a great network, but in finding a carrier to supply you a guaranteed level of service. Telecom NZ promises you nothing, excepting perhaps, huge broadband speeds, unachievable even at 3 a.m.

    I enjoyed Paul Budde’s comments as well, and I constantly wish there was a consumer protection agency which could actually “speed test” these networks and fine the slowpokes; essentially all of ’em. On that note, Esme, I would be very interested in how you came to “actual speeds?” A web application? I’d be curious. I won’t mention names, but a few old favorites are so compromised as to be worthless now.

    The network we built in New Zealand was wonderful, ran lots of cameras (within the network) and indeed, delivered internet. As you well-covered, the network is only as good as the pipe it’s hooked in to. Tahiti has a new cable coming from Hawaii and I’m hopeful it will land in NZ soon. We are already contemplating more projects there.

    Thanks again for a lovely article. You are always on the button!

  7. Bobby,

    I used Ookla’s I tried other applications to measure broadband speed and they’re much worse.

  8. bobby vassallo says ? If anyone has a preference, chime in.

    Thanks again, Esme. I always enjoy your gifts!

  9. Travel to Sweden.