The Promise of Fiber (I): To the end-user

Looking back on this last couple of years, I can’t help but be surprised at how fast the status of the FTTH subject shifted from science-fiction to work-in-progress. Considering the level of necessary investments, it’s only fair to question the validity of such investments. Strangely enough, when you ask fiber-oriented ISPs or operators about it, you get unconvincing responses.

At a conference last november, a panel of such operators was asked a tricky question: "what’s the killer app’ of fiber?" The overall response was as vague as it was darwinian: "we trust the customers to figure it out by themselves". One of the panelists even made an interesting – if flawed – parallel. "Look at SMS, he said? No mobile operator would have invested a penny on SMS and nowadays it's their most profitable product. The customer generated the usage, not the operator." I would have pointed out to him that SMS has the distinct advantage of leaving a usage trace that can form the basis of billing, something that internet usage doesn't do. The key is finding services that can be monetized, after all.

This got me thinking about the promise of fiber. Obviously, fiber doesn't hold the same promise whether you're an end-user, an operator, an incumbent, a local government, etc. So I thought I'd try and examine critically the announced promises for each of these players in the value chain. This first installment focuses on the end-user.

The most obvious promise to the customer is descending bandwidth. While ADSL2+ peaks at 20M download rates (and for that you have to practically live inside the CO), current fiber offers flirt with 100M dowload rates. The fact that customer perception focuses on descending bandwidth though is mostly a result of how broadband has been marketed so far. Beyond the higher descending bandwidth rates, fiber promises much higher - potentially symetric - ascending rates which, while not necessarily understood by most users today, could be the key to a more in-depth evolution of usage. Higher upload and/or symetry allow for more interactivity between the user and the network. If I was a journalist, I'd talk about broadband 2.0 but I'm not, and I'm sure you get the point.

Bandwidth in itself is only a small part of the picture though. Filling 20M of download capacity is already pretty tricky if you don't do eMule, so imagine filling 100M! What's more significant than the width of the band (if you'll pardon the pun) is the fact that it allows for faster download and upload, not just more. A lot of time-delayed usage (download on day one and enjoy on day two) will become instantaneous, or very near so. I believe this can also spell significant change in customer usage.

Another implied promise is quality of service. DSL is, by nature, an unstable solution, and the quality of services provided over DSL is often inadequate. Anyone who regularly watches TV over DSL knows that freeze frames, glitches, sound compression, etc. are daily occurences. And I'm not even talking about voice services. Fiber as a technology is more stable and should therefore provide better quality of service (not to mention that higher bandwidth means more services can comfortably share the link with little or no degradation). As anyone who works in telecoms can tell you though, the technical aspect of QoS is often not the most important. The heart of QoS is process, and that's a field in which operators and ISPs are notoriously undercompetent. So I'd say that while fiber has the potential for better QoS, it will only be realised if providers put in the efforts to actually make it so.

Finally, there's a theoretical promise which I believe to be false, even though many operators have used that argument in their presentations on fiber. DSL is, by construction, not capable of covering 100% of the population. The fact that the signal degrades the further you are from the CO means that in many areas DSL simply is not available to the end customer. Fiber has no such limitation, and as such could be the vector of a wider coverage, especially of areas currently devoid of any broadband offer. The thing is, when the business plan of urban coverage is already extremely uncertain, which operator in their right mind would go and deploy in rural areas, where the costs are higher and the density much lower? While extended coverage might be a promise of the technology, it certainly isn't a promise any of the players currently in the game can keep...

Overall, I'd say that the most significant of these promises is probably speed. Did I miss anything?