Ten years ago, I routinely told clients that nobody builds fiber to overbuild an established fiber network. And at the time it was true except for a few well-known examples. For example, when Google Fiber went to neighborhoods in the Research Triangle in North Carolina and in Austin, Texas, both the incumbent telco and cable company both reacted by building fiber. I know a few folks that live in those areas today, and they have the choice of three fiber providers.
Apparently, a lot of fiber overbuilders are now overbuilding existing fiber networks. RVA LLC recently published a report that said that there were 10.4 million new fiber passings built in the country in the year ending September 2023, but that only 8.4 million were to homes that didn’t already have fiber. That means there were 2 million fiber passings built in 2024 that are competing with another fiber provider. And I assume most of these passings also compete with a cable company, except for the cases where the existing fiber provider is a cable company.
Anybody building a second fiber clearly thinks they can be profitable. I’ve never looked at a business plan for somebody who is going to compete against both fiber and a cable company, but I have to assume they have a reasonably low breakeven penetration goal in mind in a market with three or more providers.
The chance to be competitive and successful with a new network comes down to cost per passing. In neighborhoods with aerial utilities, the incumbent telco and cable company have a huge cost advantage over anybody else because of the ability to overlash fiber onto existing wires. AT&T cited numbers to investors for years of spending $600 or less on fiber per new passing – a number that a new overbuilder can’t touch.
Incumbents also have an another advantage because they already have customers to move to fiber. Folks are often surprised when our surveys often find that telcos often still have a 10% – 15% penetration of DSL in many urban areas. Those DSL customers are likely thrilled to be moved to a faster product – usually with little or no change in price.
In the same survey cited above, RVA LLC says that cable companies typically lose about 33% market share to a fiber overbuilder. The math doesn’t seem to leave a lot of market share for a third overbuilder, unless they have some major advantage like really low prices. A second fiber provider has no speed advantage.
It’s also likely that much of the new national fiber passings come from passing MDUs, where building fiber on one street passes a lot of potential customers. A decade ago, a lot of fiber overbuilders avoided MDUs because the cost to string fiber in older buildings was exorbitant. But over time the industry has developed low-cost techniques for stringing fiber in hallways, and MDUs are now sought after instead of avoided.
I’d be curious to hear from somebody who knows how the market penetration rates settle out in a 3-provider market. Can each provider make money in the long run, particularly since this kind of market is likely to see fierce price competition?
It’s clearly no longer true that nobody overbuilds fiber, with 20% of fiber passings in 2023 built over existing fiber. I still have a hard time promoting the idea of overbuilding fiber. The ISPs I work with tend to be smaller and are leery of any new market with extra risk. But maybe overbuilding fiber isn’t a bad strategy for an ISP who is confident of its ability to sell. I would love to hear from somebody who thinks overbuilding fiber with fiber is a winning strategy.
I see it everywhere. I live in a neighborhood with a 20 year old, underground Frontier fiber plant and Spectrum. AT&T is in the process of overbuilding it. A few years ago, my company built a neighborhood almost simultaneously with Vexus and then AT&T overbuilt it a year later. We have a 25% penetration on that neighborhood and that works for us. We are building in front of another fiber provider in another city where we had built a fiber network connecting the schools. In that case, we still have fast mover advantage but can make it work with a 20% penetration. That experience gives us the confidence to move forward with a planned project to overbuild a neighborhood that already has Spectrum and AT&T but where we can keep the cost per passing low. It will require aggressive pricing and a strong ground game (door to door), and needs to work with 15% to 20% market penetration. Cost per passing is the key which is why I’m surprised the build in my neighborhood – it will be an expensive build. I tell my industry peers that, in any neighborhood they serve that has good demographics and density and can be built for a lower than average cost per passing, expect to be overbuilt at some point.
It really just comes down to the economics of building in that particular area. From permitting resistence, density, poles and ground condition, and home value and a few other things you could toss in there.
Where I’ve argued that fiber to the farm makes no economical sense at all, when you increase the density it flips.
I’m in a market with the cablco is building fiber outside of their copper plant and *2* overbuilds to their existing network are concurrently being built where the density makes sense. They are being partially displaced by fttx companies and they are becoming a fttx company.
If your home or business is in a position that you might have to call the neighbor’s party in for being too loud AND you’re above ‘low income’, you can probably support 2-4 services in that density with the competition building at their own expense.
Good column – and an issue with important implications about regulation. I don’t have an answer to the question asked. But I do think it’s increasingly complicated these days because ISPs often are competing on the bundle, not just on the fiber/home broadband connection. So there are likely puts and takes, especially as linear TV fades away, cable is 4th largest mobile provider (18 M total subs now), etc.
any competition that doesn’t have the government putting a thumb on the scale finds an appropriate price structure. competition and choice allows the end user to choose on price, quality, service, etc.
We’re seeing networks built for free (or near free), breaking the competative scale by allowing those companies to undercut. They can effectively drive everyone else out of business because they’re selling at what would be a loss had they not got government money.
So it’s really a ‘race to monopoly’ when there’s government involvement.
And when there’s not, multiple vendors (in higher density areas) overbuild the area to try to take their 20% of customers.
I harp on this a lot, but the big holdup for competition is usually access to carrier grade fiber into communities. The monopoly hold a lot of providers have is that they own the backhaul and they get to choose who competes and what the competition pays for backhaul, often on a network that was built on giften money, and often in right-of-ways they have exceptional access to vs others.
As a consumer in a dual fiber provider city, I can report that FTTH providers aren’t all the same. One uses aerial assets while the other is a microtrencher. One’s step up plan is 2 GBs and other’s is 3. One uses its own ONTs and routers, and the other doesn’t. And customer service varies wildly.
If the first FTTH is a lax operator there’s definitely room for smartly run business to enter the fray and do well. And nobody likes the cable company at all.
And as a Wisp operator I can pick off 10% of the fiber ISP’s clients for one reason or another and make a decent living doing it. The lower client count gives us the opportunity to make sure their managed LAN is 100% operational and respond fast to any trouble tickets they may generate. We come out looking like the elitist top tier provider that only the rich people can afford. Winning… 🙂
this is why I argue about this fiber as an endgame solution. consumer choice should be the endgame. a competative marketplace should be the endgame. We peel off plenty of customers from cable and fiber because we have comperable products in the consumer’s price point and good customer service.
Imagine if you were in a one fiber vendor market, everyone else was pushed out, and you got one of the ‘bad’ providers? do you sit at home to poor service happy because you have ‘fiber’?
What is the competitive biz model for those overbuilding existing fiber? Short term lower pricing vs. Brand X fiber? Short term loss leader pricing to dislodge first movers and “own” the customer prem for the long term? Then incumbent Brand X fiber lowers its pricing and it’s a race to the bottom.