Lately a certain kind of ad keeps showing up, the kind that follows you around Instagram for a week afterward: a company nobody's ever heard of, promising symmetrical gigabit speeds at a price well under whatever the local cable or DSL provider charges, aimed at one specific neighborhood. It has the shape of a scam. Usually it isn't. Real fiber ISPs building real networks advertise almost exactly like this, and the industry has a specific word for what that company actually is.

A recent post here covered UTOPIA, a network built on sharing one wire across many competing ISPs. This is the opposite bet: instead of sharing infrastructure, a company builds an entirely new fiber network in a market an incumbent already serves, usually over cable or DSL rather than fiber, and tries to win customers away on price and service alone. That company is called an overbuilder, and the word itself turned out to be worth a second look.

A term coined by the side it competes against

"Overbuilder" isn't a neutral industry label that grew up on its own. It comes from cable and telecom incumbents describing a competitor who shows up and builds a new network directly on top of territory the incumbent already considers its own patch. The word dates back to the cable competition fights of the late 1990s, when it was mostly used the way you'd expect from that origin: as a complaint. Someone is overbuilding my market.

What's stuck around is that the term never got replaced with something neutral once the challengers themselves started using it. Regulators use it. Trade press uses it. The overbuilders use it about themselves. A word coined by an incumbent to describe an unwelcome intruder is now the standard, unremarkable term the intruder uses on its own homepage. Nobody renamed it into something that sounds like what it actually does, which is introduce price competition into a market that would otherwise be a monopoly or duopoly.

What "my market" is usually built on

The word carries an assumption worth pulling apart: that the incumbent owns something here beyond a customer list it hasn't lost yet. Worth checking what that territory is actually made of before accepting the framing.

In a lot of these markets it's copper telephone wire, in some cases decades old, running DSL that the FCC's own 2024 broadband standard (100/20 Mbps) disqualifies by design rather than by neglect: every ADSL deployment and most VDSL2 deployments at normal loop lengths fall under that line regardless of how well-maintained the line is. About 18% of US homes still have DSL or satellite as their only wired option, concentrated in rural areas, lower-income areas, and places the incumbent hasn't gotten around to retiring the copper.

CenturyLink's own DSL pricing puts a number on it: roughly $55/month for a connection that tops out anywhere from 10 to 100 Mbps download depending on distance from the nearest hub, upload a fraction of that. Jio Fiber, serving a market usually filed under "developing," sells symmetric 100 Mbps fiber, same speed both directions, for ₹699 a month, a little over $8. This blog's own rural broadband series already covered a house in non-metro India running that exact class of plan. The gap isn't a technology-availability problem. It's that one side of this comparison had a captive customer base and no cheap way to lose it.

That's the actual backdrop behind "someone is overbuilding my market": a network the incumbent had every year and every dollar of monopoly-era profit available to upgrade, sitting untouched until a second company with fiber and a lower bill showed up in the same neighborhood.

Picking which towns get a second option

Overbuilders don't pick markets by throwing a dart at a map of underserved America. The selection criteria that show up across industry material are closer to a scorecard:

• Incumbent customer satisfaction scores, specifically weak ones. Net Promoter Score data is part of the actual investment case.
• Whether the incumbent is still running cable or DSL rather than fiber, since that's the easiest speed and reliability gap to win on.
• Household income and poverty rates in the area, because a fiber build has to be financeable, not just technically justified.
• Housing density and pole infrastructure quality, since aerial fiber on existing poles is far cheaper to build than trenching new conduit underground.

The genuinely striking part is what happens before the overbuilder even finishes construction. Multiple industry reports describe incumbents improving service, speed, or pricing in a market the moment an overbuilder announces plans there, sometimes months before a single new customer is actually connected. The mechanism doing the work isn't the new fiber itself. It's the threat of a second option existing at all.

The names behind those ads

That's the mechanism behind the ad from the top of this post: the whole selection process above is built around finding neighborhoods where an unfamiliar name and a too-good price land hardest against a weak incumbent. A handful of names that show up doing this repeatedly, market by market:

• Brightspeed, Apollo-backed, spun out of Lumen/CenturyLink's copper footprint and now building fiber across a 20-state footprint.
• Metronet and Lumos, both used as regional fiber platforms in a joint venture between T-Mobile, KKR, and EQT, the same firms named above.
• Ziply Fiber, Pacific Northwest, itself a Searchlight Capital/BCI-backed overbuild that reached enough scale to get acquired outright by Bell Canada (BCE) in 2025 rather than the other way around.
• GoNetSpeed, an independent build concentrated in the Northeast, expanding town by town rather than through a national brand.
• Ripple Fiber, the actual ad that started this post: a 100 percent fiber-optic operator now being acquired by Eaton Fiber, backed by a $1.5 billion investment from Bain Capital and Tillman Global Holdings. Who's actually calling the shots behind it is the interesting part. Eaton Fiber funds, builds, and operates the network; Verizon is the exclusive retail brand under a commercial agreement, handling sales, marketing, and support. Right now a Ripple build still shows up to a customer as Ripple Fiber, the way it does today in Ocean Shores, Washington. Once Eaton's acquisition of Ripple closes, expected before the end of 2026, existing Ripple customers move onto Verizon's platform, in a market Verizon never served itself, over fiber Verizon never built.

None of that guarantees good service, and none of it is an endorsement. It's mostly that "never heard of them, and their price looks wrong" is the normal shape of a fiber overbuild pitch, right up until the name on the truck is one you already trust for reasons that have nothing to do with who actually built the line.

Where the money behind this actually comes from

The financing side of overbuilding has shifted a lot over the past few years, and private equity is most of that shift.

Fiber-to-the-home networks are now described in PE materials explicitly as a "competitive moat": once a market has fiber in the ground, a third competitor duplicating that build a second time is usually not worth the capital, so whoever wins the overbuild race in a given town tends to hold that position for a long time. That's an attractive property to an investor, independent of who the ISP actually is or how they treat customers.

Federal BEAD funding, roughly $42.5 billion aimed at unserved and underserved areas, pulled a new wave of PE capital into the sector, since firms can co-invest alongside public subsidy money. Firms like EQT, Stonepeak, KKR, and Northleaf Capital Partners are now regular names in fiber overbuild financing, not just in already-competitive suburban markets but in the rural buildouts BEAD is meant to support.

Google Fiber is the clearest single example of the arc this can take. It started as Alphabet's own description of it, a deliberate market experiment meant to pressure incumbents into building faster networks. As of this year, Alphabet is selling the majority stake to Stonepeak and merging the business with Astound Broadband, producing a combined entity spanning around 7.1 million locations, with Alphabet retaining only a minority stake and Stonepeak holding operational control.

Which part the incumbent actually takes over

The Eaton/Verizon structure raises a specific version of a question worth asking about all of this: when a brand with the track record described earlier in this post starts acquiring or licensing a network someone else built, does the good network survive contact with the company that let its own copper rot?

On the evidence of this specific deal, the construction and operations side mostly doesn't change hands. Eaton Fiber keeps funding, building, and maintaining the network after the Ripple acquisition closes. It's Bain Capital and Tillman Global Holdings running that fiber, not Verizon's own legacy operating model, and their moat thesis, the same one described above, depends on that fiber staying good rather than degrading. That looks less like an incumbent absorbing a good asset into itself and more like an incumbent buying access to construction it can't do in-house at the same pace or cost, while paying someone else to keep building and running it.

The part that does change hands directly is retail: billing, pricing, support, upsells, the exact layer the DSL comparison earlier in this post was actually about. AT&T's acquisition and eventual write-down of DirecTV, and Verizon's own acquisition and later sale of AOL and Yahoo, are the precedent for what a big carrier's in-house operating model does to something it fully absorbs. This deal isn't a full absorption on the network side. The retail layer, not the fiber itself, is where Verizon's name actually ends up on the bill.

Worth being precise about scope here, because this is the arm's-length structure of one specific deal, not a description of how every acquisition on the list above works. It's closer to the gig-economy pattern of keeping the brand and the customer relationship while paying a separate company to own the harder, more capital-intensive part underneath. Ziply Fiber's sale to Bell Canada and Brightspeed's ownership by Apollo are full acquisitions, one company holding both the network and the retail layer, not a licensing split. The good-network-survives-because-someone-else-still-owns-it read applies to Eaton/Verizon specifically. It isn't a general law about what happens when an incumbent buys an overbuilder.

The risk in how overbuilders are financed

Overbuilders exist because incumbents, left alone, don't have much reason to improve. A second wire is what forces the first company to compete on price and service instead of just holding a captive customer base.

Trade press covering this shift keeps flagging the same pattern: the financial playbook overbuilders were founded to disrupt is now funding a lot of them directly. Segmented tier pricing, equipment leases, data caps that creep back in gradually after being marketed as a point of difference, these are moves PE-backed telecom operators have already run in other parts of the industry. None of that is guaranteed for any specific overbuilder. But once a fiber overbuild wins its market and becomes the moat an investor described, nothing in that ownership structure requires the company to keep behaving the way it did to win the market in the first place.

Summary

The industry borrowed a complaint and turned it into a category. An overbuilder is, by definition, the thing an incumbent didn't want to happen to its market, and the whole reason the category matters is that incumbents mostly won't improve without one showing up. What's less settled is whether the companies now funding overbuilders at scale are optimizing for that same pressure, or just for the moat it eventually creates.