Can Open Access Work?

MLGW_Substation_Whitehaven_Memphis_TN_2013-01-06_006Today I am meeting with the Public Utility Districts (PUDs) in Washington State and they have been gracious enough to invite me to be the keynote speaker at their convention. These are the rural electric companies that serve much of the state outside of the major cities.

Whenever there is any listing of the fastest Internet speeds in the country the areas served by some of these PUDs show up among the fastest places because many of the PUDs have invested heavily in fiber. But they have a unique business plan because there is a legal restriction in the state that prohibits PUDs from being in the retail telecom business. This has forced them into operating open access networks where they build the fiber network and let other companies provide services.

No two of the PUDs have gone about this wholesale business in exactly the same way, and so together they provide multiple experiments on ways to operate a wholesale open access network. I know several of the PUDs well and they have one universal problem – no large, well-financed service provider has agreed to offer service on their networks. No big cable companies or telcos or anybody you ever heard of wants to serve the many customers on these fiber networks. There are a handful of connections sold to companies that serve large businesses, like Zayo and Sprint, but no bit company that wants to serve smaller customers.

What is lacking is vigorous competition on their networks from multiple companies willing to serve residents and small businesses. And that is what open access is supposed to bring. Instead, most of the retail service on these networks is provided by local ISPs who took advantage of the opportunity to reach more customers. In many cases the local ISPs were so small and undercapitalized that the PUDs had to assist them to expand onto their networks.

There are not many other open access networks in the US. One of the largest ones was in Provo, which struggled with the model and eventually sold their network to Google. I was privy to Provo’s books and they could not find a business plan model that would make their business model cash flow. But if we look outside the US there is another great example of how open access can work if done right. Europe has a number of cities that have built fiber networks and invited ISPs and others to serve customers. In Europe this has been a big success because numerous service provider show up to provide service. Some of these providers were the former state monopoly companies that were unleashed to compete after the formation of the European Union. But there are also new competitors there akin to our CLECs and ISPs.

The big difference between US and Europe is that here none of the incumbent competitors are willing to operate on somebody else’s network. I can’t think of one example in the US of large cable companies competing against another one. And there is very little competition between the big telcos other than some fierce competition for some giant government and business accounts. Here in the US the PUDs have only been able to attract small local ISPs to operate on their networks. For the most part these ISPs do a good job, but they are small and have the problems that all small telecom companies have.

Many of the PUDs are in the uncomfortable position of only have one real service provider on their network. Should the owner of that business die or just go out of business a PUD could see most of their network go dark and all of the residents and businesses in their towns lose their fast Internet.

Anybody who understands telco finances instantly understands why this model is so hard to make work. A company must spend a lot of money to build a fiber network and then can charge only relatively small fees to others that use it. A typical revenue for wholesale access to a fiber network is in the range of $30 per customer per month and that is really not enough revenue to pay for building and operating the fiber network. By comparison, most triple play providers have an average revenue per customer north of $130 per customer per month.

The PUDs built the fiber because they are in rural areas where nobody else was going to do it. Their communities have already benefitted tremendously from the fiber. But they have their work cut out to keep this going, and I am sure they will figure out a way to do so.

2 thoughts on “Can Open Access Work?

  1. Doug,
    You are right on in your analysis. Open Access is extraordinarily difficult due to the capital intensive nature of the business offset by relatively low ARPU for the network owner. I will say that one model worth watching is the “new” UTOPIA plan with Macquarie’s capital (stories adding more color were out last week). Macquarie is (supposedly) committed to build past 100% of the homes in ALL UTOPIA cities (all the cities had to agree) in exchange for a “utility” fee charged to all homes – regardless of whether they subscribed to the service or not. [That is my reading of the articles I have seen and not a statement of fact].

    I would say that there is a more fundamental flaw with the traditional open access model, however. A key question is: Can (or will) an operator who has not risked such a substantial sum of their own capital, operate in a “rational” manner?

    Once the capital is deployed, the operators attracted are small (under-funded)players – as you correctly point out – that may or may not be viable competitors. Refer back to the days of LD competition, early CLEC (i.e. resale & unbundled loops), etc and we saw countless competitors without assets attempt to build a customer base with a plan to “flip” it down the road (without regard) for whether they were making money.

    In those scenarios, the new entrant sells on price as their business plan has a very short term focus – build a customer base. If / when their plan fails, they simply close the doors. The network owner is left to service the customers in one way shape or form (either themselves or finding another viable operator and transferring service). The surviving players, including the network owner, are responsible for the fallout – not the least of which is adjusting those subscribers to what is rational, profitable, pricing.

    In a true open access market, it is not entirely unlikely that it is more than a single player that is under funded and goes down prior to stabilization in the market. In these scenarios, is the customer truly better off?

    I recently became a cord-shaver because I simply couldn’t rationalize a $200 monthly bill for my triple play – a large sum of which is cable. However, I am close enough to the economics to know that it costs A LOT of money to build and operate these networks. The key is market-share. A Bernstein study was announced today that Google Fiber is taking 75% market share in key neighborhoods and 30% even in low income neighborhoods. Google may be able to make money with these figures (and far lower opex with a different video distribution model). The “100%” model Macquarie is pursuing in UT may also change the economics. But, I have been close enough to the overbuild model to know that “just because you build it, doesn’t mean they will come”. Furthermore, when they come, their allegiance is short-lived. Customers quickly adopt the mindset we all had in the days of LD competition (80s & 90s) of “what have you done for me lately?”

    When operators start handing out $300 or an HD TV to change providers, no one is making money.

    Best of luck with your presentation.

    • It’s not quite so bad that people are building customer bases to flip, because in these rural areas there is more of a challenge in finding somebody to sell to. The real challenge for the service providers on the wholesale network is to make a long-term profit on a network you don’t own, having to work in conjunction with the network owner to diagnose network problems, and always wondering if the rules of engagement will change on you.

Leave a Reply