I’ve been asked by several ISPs if they should consider operating on an open access network. Like with any decision of this magnitude, there are no easy right or wrong answer, but instead a lot of pluses and minuses to consider. Following are a few of the most important factors to consider about operating on an open-access network.
Capital Expenditures. One of the primary reasons to think about using somebody else’s network is the savings from not having to fund and built a new network. For small ISPs without a lot of borrowing capacity, an open-access network might be one of the easiest ways to get more customers.
But there is a flip-side to not spending money on capital. If the ISP plans on eventually selling the business, there is a lot more long-term value created by owning a network than by riding somebody else’s. There might be far more corporate value created by building a small fiber network with a few thousand customers than by serving 10,000 customers on an open-access network. If you pursue open-access, it has to strictly be about the cash flow generated today rather than about the value created in the future.
Economy-of-scale. Another reason to consider operating on somebody else’s network is that anything that makes your ISP larger adds to economy-of-scale. There is a big benefit to spreading the costs of overheads like OSS/BSS systems and corporate staff costs over as many customers as possible. Any costs you can shuffle off to an open-access expansion should make your other markets more profitable.
But economy-of-scale savings can diluted if you decide to tackle a open-access network that is far away from your existing operations. It’s never as cost-efficient to open and operate in a new distant market compared to one that is next door.
Trust. One of the scary parts of being on an open access network is being captive to the processes and prices charged by the network owner. An ISP is taking a leap of faith that the network owner will always perform as promised.
If the network owner decides to increase wholesale rates, there is no option but to go along. While a rate increase would also apply to the other ISPs on the open-access network, it’s possible for rates to get too high compared to other competitors like the cable company.
It can also be a problem having to rely on somebody else’s processes. For example, in most open-access networks, the network owner builds new fiber drops and installs customer electronics. It can be devastating for a marketing plan if the network owner can’t deliver customer installations on time, or decides it has a limited budget to add more customers. An ISP is also a captive of all other processes of the network owner like trouble reporting and resolution, timely network upgrades, network monitoring, and the ordering process. Even if the processes are good today, the network owner can change the way they do anything. This is possibly the biggest reservation for an ISP that is used to working on its own network.
No Technology Advantage. It’s an odd situation for an ISP to be operating on a fiber network and yet have no technology advantages over many of your competitors. Every ISP on the open-access network has the identical capabilities as you. This means that an ISP on an open-access network must distinguish themselves through either price or customer service. Open-access can turn into a race to the bottom if one of the ISPs on the network decides to deeply slash prices.
This blog wasn’t meant to scare ISPs away from working on an open-access network. There are ISPs that are thriving in this environment. But it’s not for everybody, particularly for ISPs that want to control the customer experience from beginning to end.
10 replies on “Should ISPs Consider Open-Access?”
Middle mile/dark fiber is a solution to most of your objections.
I have a slight contrarian view: municipalities should all have open access backbones and require that be the only backbone used to serve their households, including cable-based internet.
The big ISPs have “natural monopolies” which prevents competition anyplace other than the fringes (and then, the smaller ISPs are forced to use less powerful technologies like fixed wireless).
The big ISPs, having saturated their markets, don’t have any mechanisms for natural growth — so the idea that they’re healthy, competitive, capitalist entities is not correct. They’re tumors — you can’t get rid of them and you can’t introduce new competitive entities into the market which would force them to hone their business (except maybe something like Starlink, which is going to hit scaling limits pretty fast).
Forcing the ISPs onto the last 100′ would get rid of the unreasonable barrier to entry, a bunch of local ISPs could spring up and get back to competing on price and service. Having a full muni backbone could leave it to the entity protecting the citizens to ensure that low income households aren’t left behind.
So, I’d say the question ought to be, how do we make muni backbones pervasive and awesome?
Internet ought to be like running water, where everyone is entitled to access. We can’t get there from where we’re starting, something radical is needed.
Then, the next question is, how could we use something like SDN to do something equivalent at the muni level for cell coverage?
Our company has just been hired to design/build a municipal dark fiber middle mile network in Erie County, NY, funded by the feds. Hopefully, we will be able to IRU fibers to all. Google Erienet for more information. Here’s a an overview: https://www3.erie.gov/economicdevelopment/sites/www3.erie.gov.economicdevelopment/files/2021-10/erienet-broadband-services.pdf
I disagree with the assertion that municipalities should all have open access backbones. When serving rural America, you need to get every dollar to stretch as far as possible. If, at the same time that you are trying to get a network to be ubiquitous in a region, you are only going to sell to wholesalers, then you are effectually increasing the costs to the end-users. A total provider can leverage their assets to keep costs as low as possible. If the municipality is required to only provide the middle mile then the ISPs must mark up the costs from the municipality to make any margin; this markup is their own overhead for service, equipment, materials, labor, etc.. If the municipality is only focused on sustainability, then it doesn’t need to take in a specific margin. Additionality, municipal networks are focused on the needs and best interests of the community. I would agree that some municipal entities do not understand the full retail market however, there are those that do and providing low-costs high-quality end-to-end services at a cost point that is sustainable will stretch each dollar and can bridge the digital divide sooner.
Trying to parse some dense text here — are you saying, Robert, that municipalities shouldn’t be restricted to middle mile, and should be able to deliver end to end service for their residents?
If so…sure, you bet, if there’s a not-for-profit middle mile and municipalities can deliver the last 100′ economically, ab-so-lutely! There are lots of places in rural America that we’ve already had overwhelming proof that the big ISPs won’t touch because they’re too low density and low profit for their taste. They are actively dropping those customers like hot rocks and municipalities, counties… whatever… ought to pick up the slack.
I’m mostly saying, kick the big ISPs out of the position of being the sole owners of the municipal backbone. It’s not doing them any good as an asset in a growth-oriented public company since…municipalities aren’t growing enough new subscription services to satisfy their shareholders. And, they refuse to compete on price. And, they have a stranglehold on entry to the market.
So, it’s the government’s *duty* to step in and take this component of a fundamental service away from them.
Many, many, many places don’t have enough critical mass for big ISPs to take any kind of care of and they’re the ones that are truly suffering.
Left to their own devices, the whole damn internet ifrastructure is going to be like copper DSL or payphones, left to be picked up by the low bid loser when the big ISPs realize there’s no growth left and that it’s time to divest to private equity. For the benefit of our society we need to change the model before capitalism discards something we care about because it’s no longer a “growth area.”
All you need to do is to look through the list of landline service providers to figure out what the future of the internet is, if left to the growth-obligated hands of public internet companies.
It doesn’t really matter if the internet continues to be of general interest. These companies need to figure out ways to take more money out of the system to satisfy their growth obligations. Whether that be picking the pockets of Facebook and Google by getting rid of net neutrality, or trying to play out the 5g fantasy of getting more money for faster mobile connections (or, somehow, tapping into the lucrative network slicing market??!??), they’re needing the money… and that’s the bottom line.
And, delivering more of the same to a fixed sized audience isn’t going to cut it.
Who remembers when CenturyTel ran payphones? Anybody here still on copper DSL? — hurry up and get your starlink subscription soon…
Doug et al: “Open access” as a telecom concept dates back (at least) to the old copper telephone days and the effort in the 70’s and 80’s to introduce competition into that world. The “pro-competitive” folks (which, for much of that period, included the FCC) pushed for telcos to make their infrastructure available to competitors like MCI. Not surprisingly, the telcos resisted fiercely–and, largely,successfully. Thus, in those “bad old days” a customer of, say, Pacific Bell, could not buy voice telephone service, security services or any other service that PacTel offered over the latter’s copper network. Ditto for cablecos: you could not buy any video services from, say, Comcast, other than those offered by “your own” cable network. (The result was the huge growth in rental video discs.) In today’s broadband market, those days and restrictions are long gone: all networks that offer Internet access are, effectively, “open access” in the old meaning of that term, i.e: any purchaser of Internet access, no matter where or from whom the latter is purchased, can access any service that is available over the net–including, voice telephone, security, video etc etc etc–regardless of whether those services come from the owner of the underlying internet infrastructure or not.
Thus, in the historical (and most meainingfurl) sense of “open access” all networks exhibit that characteristic.
The last remaining “frontier” is the actual Internet connectivity as distinct from a fiber or radio system that carries the internet connectivity. “Open Access” in that sense is technically difficult to do and offers little benefit to final customers–indeed, in many senses, it reduces the value and increases the price and inconvenience of the underlying infrastructure. Think Railroads!! All RR’s, as “common carriers” are required to transport not only freight but actual freight cars from other RR’s over their infrastructure provided the latter have the right size wheel trucks, connectors, signalling capability, size and weight characteristics etc. I do not (and never have) seen a push to permit other parties to join in the actual traffic management of any particular RR. Ditto for pipelines.
Frankly: this should be the same for telecom as well (esp fiber which does not have the frequency issues that radio has), In fact, fiber BB mostly does act that way in practice.
In short, the current fad for “open access BB networks” is a product of economic, technical and policy confusion…..and should be abandoned. BB networks should provide open access to all services–which, in fact, they pretty much do already–and let the owner/builder of the underlying network manage the whole thing. That said, as an old “New Dealer”, I do think that BB networks should be officially classified as “common carriers” under Title II of the 1934 communications Act–and regulated as such.
Indeed, that should have been done years ago. As a very old player in this field, (I am now 81 and have been in this field for over 50 years) I was a senior Congressional staffer on both the Senate and House Commerce Committees during the late 70’s and 80’s when this fight was fought. “My team” in this battle (which wanted to bring all major telecom companies under Title II) lost the war comprehensively–and the USA is a poorer country because of it. (NB: most other advanced countries did bring cable and BB into their general structure for telecom regulation–though to differing degrees and manner. As a result, their policy fights have, by-and-large–been less destructive and their telecom sectors more rational ever since.
Hi Timothy,
Interesting stuff. I’m probably just drowning in terminology a bit. BB is “broadband?” And, when you say that BB networks provide open access to all services, are you saying something like, “everyone on Comcast can access all internet services regardless of who hosts them (basically, re-affirming internetwork routing and common network standards like TCP/IP)?
But, of course, they don’t let you run your own IP service from their pole connection to somebody’s house.
What’s missing is a framework that provides local competition. The problem is the massive majority of Americans who have access to only one or two ISPs, the effective local monopolies of cable providers, etc. If Comcast has unique access to the cable going to your house and all of the cable/fiber all the way to their colo connection with the greater internet, you, the consumer, aren’t likely to see any relief; someone has to run a parallel physically overlapping network on the scarce pole real estate, which is logistically challenging and very expensive, especially since there’s a perfectly good Comcast (or whatever) that you’ll be competing with.
If, on the other hand, there was a lit, routed muni fiber network with a standard interface to connect local residences (which Comcast or whomever could take advantage of), it would seem feasible that local ISPs could move in and provide competitive (generic internet) service offerings.
Or, if the muni provided lots and lots of dark fiber runs that local ISPs could bid to light up and run their redundant network on, that would be another way to eliminate the stranglehold incumbents have on pole real estate, and the unfair advantage they have, given the cost of running fiber throughout a municipality.
So, I’m hoping I’m not just misunderstanding your point… please do give me some more details if I’m pointing in the wrong direction. (What I want will never happen because we’ve lost the will to mold private industry in ways that benefit local consumers, if the private industry is big enough and has enough lobbying clout… but, fun to speculate about.)
As you say, in most open access networks there’s not much involvement for the retail ISP other than marketing, billing and customer service. Now, customer service is very important, especially technical support, which is where an ISP gets most of their calls. That is one way to differentiate yourself. It would be a nice feature for an open access operator to provide the retail ISPs some level of visibility into the fiber network equipment and CPE so the retail provider can field those calls and do some level of troubleshooting before escalating the call to the network operator. Another area where an ISP can differentiate themselves on an open access network is the additional services they offer over that connection. They can offer a VoIP product (residential and commercial), an IPTV product, and other ancillary services that will improve ARPU and stickiness. Otherwise, like you say, it is a race to the bottom on price, and subscribers will switch back and forth based on that.