Opinion: FTTH is (relatively) cheap

The debate around next-generation access in general and Fiber to the Home in particular is dominated by one single overbearing concept: deploying a new fiber infrastructure is prohibitively expensive. We all have this cost figure of around $1000 per home connected as a baseline, and it has become something close to a fact. In this post I won’t address the fact that most of the players actually deploying the infrastructure discover (often to their surprise) that they rarely get even close to that number since Herman is working on a report on that very topic. Let’s assume for a minute that $1000 per home connected is a real cost that applies to say 75% of the population in the developed world.

What I want to address here is the notion itself that FTTH is “expensive”. In the world we live in, there is no such thing as an absolute cost, it’s all relative. Choosing where to spend one’s money, whether you are an individual or an organization is always the result of arbitration. And once you start examining the relative price of FTTH, a number of interesting things emerge. There are essentially three areas of comparison that I find interesting: at state level, at real estate level and at the consumer’s individual level.

At state level, the “relative” view of the investment lies around other infrastructure arbitrations, especially ones involving public subsidies or participation in one way or another. A local government that wishes to accelerate the deployment of quality broadband in its territory might be faced with arbitrating between giving a nudge to next-generation access or giving it to a new road, bridge or railway line. It’s actually fairly easy to compare because most of these large infrastructure projects are public. Here’s an example from a few months ago that really got me thinking. The French high-speed train that connects Paris to Bordeaux will be upgraded in the next few years between Tours and Bordeaux (in French). The line between these two cities is not as modern as it could be and the trains are therefore constrained on speed. Instead of the three hours it currently takes from Paris to Bordeaux, the upgrade of the line could allow trains to get to Bordeaux in 2 hours and 20 minutes by 2017. The cost of the project is 7.8bn €, 44% of which will be subsidized by national or local governments.

Considering that the fibering up of the whole of France (including all the rural areas) has been valued at around 30bn €, the comparison basis is this: French public entities can build a viable project to reduce the travel time between two major cities by 40mn for about a quarter of what it would take to get next-generation broadband to the whole country. On this basis it seems hard to believe that a viable project could not be built for national FTTH. The fact is that deploying next-generation infrastructure, in the grand scheme of things, is much cheaper than roads, bridges, railways and airports. If we had none of these, the arbitration of course should be in favor of them. When it comes to upgrading them or building more, the arbitration becomes viable and NGA should look comparatively better. The main reason it doesn’t, in my opinion, is a lack of understanding of its importance by ageing politicians and established routes to financing for all of the legacy infrastructure projects that the NGA ecosystem still lacks.

At real estate level, the outlook is different: we’re talking private money and well-oiled ecosystems where decision makers focus on what’s most viable for them. It has been established in a number of markets (Sweden and North America among them) that fibering up a property has positive impacts on its lease price, sale price and time to find a tenant. So the cost of fibering up a property should in turn be put in relation to (say) the price of rent. In NYC, an average one bedroom flat rents for $2500 per month, and the prices only go up with size (obviously). That means that if fibering up a flat helped eliminate a few days of lag between tenants, the $1000 investment for FTTH (assuming it was that expensive in NYC, which is doubtful) would be paid for. Of course, NYC is one of the most expensive cities in the world, but it’s also a place where FTTH deployment is happening very slowly and without great involvement from the real-estate owners. So again, why isn’t it happening? Simply because of misaligned understanding. Real-estate people by and large do not understand the benefits for them because it hasn’t been articulated with their goals in mind.

Finally, even at the individual consumer’s level, the cost is relative: the new ipad has sold millions of copies in one week-end, at a price point of roughly $750 per unit. Interestingly, the new retina display on this device is such that the amount of bandwidth needed to stream content to it will be multiplied by 4 compared to traditional HD. There’s now an increasing number of reports that show that tablets in general (and ipads in particular) exist in multiples in more and more households. At what point does the arbitration become “I want decent broadband to use these devices optimally”?

These comparisons are interesting to me because they demonstrate that subscribing to the meme that FTTH is “expensive” is not very productive. In many markets I see service providers hesitating as a consequence of that with little to no thought about how it is not only possible but comparatively cheap to build investment models that make sense for this new infrastructure by reaching accross the ecosystems to those who can benefit directly from it.

FTTH is (relatively) cheap when you look at it this way.