“Dark fiber” might sound dramatic, but the term simply refers to (mostly) fiber optic cable that has been laid in the ground but doesn’t carry data—in other words, it isn’t “lit.” There are miles of such unlit cable all over the United States, much of it laid during the 1990s telecommunication boom but never connected to anything. It’s not unlike a subdivision with the streets laid out and the sewers and utilities in place but with no houses, because no developer ever took the opportunity to build.
Nevertheless, many companies are hard at work laying new fiber. Yet there are significant differences—business and economic, not technological—between the 1990s’ boom and today’s renewed activity.
Dark Fiber yesterday: boom and bust
The first fiber ‘land rush’ was part of the telecom boom that accompanied the late-1990s Internet boom. “The 90s saw an explosion in the number of competitive local exchange carriers laying all-new fiber throughout the whole country,” according to Doug Barnett, senior analyst at consulting and research firm Atlantic-ACM. “The expectation was when this Internet thing took off, bandwidth usage was going to explode.” The driving philosophy behind laying new optical fiber was “build it and they will come,” in Barnett’s words.
Paul Sazill, senior vice president of product management for communications firm Level 3, paints the same picture. “In the original days of fiber construction—in the late ‘90s—there were a number of nationwide long-distance, intercity fiber networks built by a lot of different companies.”
All this frenzied activity led The Wall Street Journal in 2002 to describe the late-90s telecom industry as a “gigantic poker game.” At the time of that article, the tech bubble had burst, and the telecom sector had seen more than 60 bankruptcies. That ushered in a bust period that lasted for most of the last decade, during which time comparatively little new fiber was laid. Much of the existing capacity remains unlit, though nobody has a clear idea of how much.
Dark Fiber today: cautious growth
The cable-laying industry is seeing growth once again. In a widely quoted statistic, the CRU Group research firm estimated 19 million miles of optical fiber were installed in the U.S. in 2011, the most since 2000. “We haven’t returned to the days of a bunch of companies building nationwide fiber networks, but we’re past the days of nobody building anything,” says Sazill.
One driver of this new construction is new demand from new users. The original boom centered on the obvious places with high bandwidth demands: New York, the San Francisco Bay Area, and so on. Sazill cites the current efforts of Google to roll out Google Fiber high-bandwidth service to cities in Kansas, Missouri, and Texas.
Even in the areas that got connected during the first boom, there are new structures to be served. “The new stuff that we’re building, we’re building it to new places, to new buildings,” says Sazill. “Fiber still only reaches a relatively small percentage of all the buildings in the country. The buildout process is ongoing.” This process can be seen in the continual increase in the percentage of academic institutions with dark fiber connections to their ISPs (see Table 1).
Table 1: Academic institutions with dark fiber, by type of institution: FY 2005–11 (Percent)
SOURCE: National Science Foundation, National Center for Science and Engineering Statistics, Survey of Science and Engineering Research Facilities.
Overall, however, “there’s not a lot of laying of new dark fiber,” says Barnett, who goes on to say that the “build it and they will come” attitude is “almost nonexistent.” Most new builds, he says, are to meet a concrete business-case need. Sazill explains, “Private companies like Level 3 and competitors build out fiber proactively where we see that opportunity, but we also do a lot of construction that is based on customer orders.”
One brake on the construction of dark fiber is the unwillingness of some carriers to be identified as primarily infrastructure companies. Atlantic-ACM carries out annual customer surveys asking respondents to rate quality and price competitiveness across service providers. About five years ago, “we started taking it out of the research because most of the providers were uncomfortable being listed as an active dark fiber provider,” says Barnett. “The carriers wanted to focus on marketing their lit services rather than just selling a commodity.”
An April 2013 article by Fedor Smith on the Atlantic-ACM website points out a corollary to the commodity issue: “Growing bandwidth demand means higher revenue from an existing customer in the future, and management of that bandwidth delivery has margin in its own right, all of which is foregone when the carrier turns over the fiber to the customer.” In other words, while a long-term lease on fiber is a nice guaranteed revenue stream, it foregoes the ability to take advantage of price rises in the future, as well as management fees.
In addition to the potential loss of revenue, another drawback to leasing dark fiber is that a company could wind up enabling a competitor. Most cable-laying companies these days also offer their own lit services, as Level 3 does, that they wouldn’t want to undercut by letting the competition piggyback on their infrastructure. “We think about it,” says Sazill, “but there are a lot of customers we can sell fiber to that are not competitive with us. There are many enterprises that need fiber, and that group is growing, so we sell to them. In regards to other competitive carriers that want to buy fiber that we have laid, we take those scenarios on a case-by-case basis.”
Despite these checks on dark fiber construction, Barnett says, “we’ve seen the buying activity of dark fiber pick up from both a wholesale and retail perspective.” On the 2013 Atlantic-ACM US Long Haul Wholesale Carrier Report Card, 27% of the participating wholesale buyers reported buying dark fiber, and 57% of those said they expected to spend more on dark fiber over the next year. As an example of the kind of customer driving that market, Barnett points to wireless companies who need Ethernet connectivity from their cell sites to their distribution points. “We know from our interviews that there’s at least one of the majors purchasing dark fiber for these backhaul services, and several others are looking at it.”
The “build it and they will come” approach may be dead, but they’re still coming, so people are still building it.
Logan Kugler is a freelance technology writer based in Silicon Valley. He has written for over 60 major publications.
No entries found