The Economics of Overbuilding

There is something peculiar about the economics of broadband expansion.

Across many markets, operators continue investing billions to extend fiber networks, increase homes passed, and compete for subscribers. At the same time, other operators face mounting debt, refinancing pressure, consolidation, and questions about whether their original investment assumptions remain viable.

The demand for broadband is not disappearing. Quite the opposite.

But increasing demand for connectivity does not necessarily mean that every market can economically sustain another physical network.

And that raises a question worth examining: At what point does infrastructure competition become infrastructure duplication?

The Same Customer, Three Networks

Consider a neighborhood of 10,000 homes.

One provider already offers gigabit broadband. A second enters the market with a competing fiber network. Eventually, a third arrives, offering similar speeds and pricing.

Each operator incurs substantial capital expenditure to reach essentially the same addresses. Each must acquire customers, maintain its network, finance its investment, and generate an acceptable return.

Yet the arrival of the third network does not create another 10,000 households.

It simply introduces another competitor for the customers already living there.

From the consumer's perspective, the additional choice may be welcome. Prices may fall, service may improve, and incumbents may finally respond to complaints they previously had little incentive to address.

From the investor's perspective, however, the economics become more complicated.

The available revenue is divided among more providers, while much of the physical infrastructure expenditure has been duplicated.

Competition can improve the customer proposition without necessarily improving the economics of the industry supplying it.

The Market Is Already Testing the Limits

This is not merely a theoretical concern.

Ofcom's May 2026 Connected Nations planned network deployments report forecasts that, by the end of 2028, 73% of UK homes could have access to at least two gigabit-capable networks, while 36% could have access to at least three.

Those figures describe network availability, not subscriber distribution or profitability. Nevertheless, they reveal the scale of overlapping investment being contemplated.

In the United States, CostQuest's 2026 broadband research similarly identifies competitive overbuilding as an increasingly important driver of fiber expansion, rather than simply the extension of service into previously unserved locations.

Meanwhile, the commercial challenge is changing.

Passing another home is a construction achievement. Converting that passing into a paying customer is a different proposition altogether.

Where customers already have adequate broadband, a new entrant must persuade them to switch.

That may require lower prices, promotional incentives, superior customer support, better reliability, or attractive bundles combining broadband, mobile, television, and other services.

Every one of those strategies affects the investment model.

Lower prices compress revenue. Promotions increase acquisition costs. Better service may require additional operating expenditure. Bundling can improve retention, but requires the right commercial capabilities.

A network can be constructed on time, within budget, and to specification—and still struggle to earn an adequate return.

Competition Does Not Require Three Trenches

Much of the cost of fiber deployment lies in the physical work required to reach customers: rights-of-way, excavation, ducts, poles, restoration, and associated construction activities.

The UK's Department for Digital, Culture, Media and Sport previously noted estimates placing civil works at as much as 80% of deployment costs in some circumstances. It also cited Ofcom's estimate that reusing Openreach ducts and poles could reduce deployment costs by approximately half.

Those figures are context-dependent, but the underlying economics are straightforward.

If usable infrastructure already exists, constructing another physical route may be substantially more expensive than obtaining access to the existing one.

This is where wholesale and open-access models become interesting.

Openreach, created in 2006 following regulatory intervention, provides infrastructure and wholesale access to competing communications providers. Its model does not eliminate competing network construction, but it demonstrates how retail competition can operate without every provider independently owning the entire access path.

Other arrangements offer different degrees of sharing: joint trenching, access to ducts and poles, long-term fiber leases, indefeasible rights of use, and municipally owned networks supporting multiple retail providers.

In Willmar, Minnesota, for example, a city-owned open-access fiber project was under construction in October 2026, with three retail providers reportedly signed to deliver service over the planned network.

These models separate, to varying degrees, the economics of infrastructure ownership from the economics of customer acquisition.

They also raise legitimate questions.

Who determines access prices? Who funds upgrades? How are maintenance obligations allocated? What prevents a wholesale infrastructure owner from exercising excessive market power? And what happens when a shared physical failure affects every retail provider using the network?

Sharing infrastructure reduces some forms of duplication. It does not eliminate the need for sound governance.

More Bandwidth Does Not Always Mean More Fiber Routes

The next argument for continued infrastructure investment is familiar: artificial intelligence, future wireless generations, connected devices, and increasingly data-intensive applications will require more network capacity.

That is reasonable.

But the conclusion does not automatically follow that every neighborhood needs another independent access network.

Existing optical fiber can often support substantial capacity increases through changes to transmission equipment, wavelength utilization, PON technology, network architecture, and aggregation.

Of course, physical infrastructure has limits. Some routes lack available strands or duct capacity. Others require genuine path diversity, replacement, or expansion. New development also creates demand where no infrastructure exists.

But we should distinguish the need to increase capacity from the need to duplicate civil infrastructure.

The physical route may remain useful long after the electronics at either end have been replaced several times.

That makes the decision to construct another route a long-term capital allocation question, not merely a response to anticipated bandwidth growth.

When Overbuilding Makes Sense

There is a danger in taking this argument too far.

A single infrastructure owner can become complacent, charge excessive prices, delay upgrades, or restrict access. A competing physical network may be the most effective way to challenge that position.

Additional routes can also provide resilience, particularly when they are genuinely diverse rather than merely duplicating the same failure exposures.

In some markets, existing infrastructure may be inaccessible, commercially unreasonable to lease, technically unsuitable, or insufficient to meet demand.

In those circumstances, new construction may be justified.

The question is not whether a second or third network is inherently wasteful.

It is whether the incremental competition, capacity, independence, and resilience justify the incremental capital investment and physical disruption.

That assessment will differ between dense cities, established suburbs, new developments, and rural communities.

It will also depend on the terms under which existing infrastructure can be shared.

The Question Behind the Capital

When investors finance another fiber deployment, they are making assumptions about market share, customer acquisition, pricing, churn, operating costs, and the eventual value of the network.

Those assumptions deserve at least as much scrutiny as cost per passing.

A project can appear attractive because construction costs are competitive, yet become financially vulnerable when subscriber penetration falls short or competing providers respond aggressively.

Equally, an expensive network can prove valuable when it serves strong demand, earns durable revenue, and supports multiple commercial uses.

The real measure is not how many homes the operator can pass.

It is whether the business can attract and retain enough customers, at sustainable margins, to recover the full cost of its investment.

And there is a wider question beyond the individual investment.

A new network may generate an acceptable return for its owners by winning customers from existing providers, even if the market as a whole ends up supporting more physical infrastructure than it needs.

That may be a rational private investment without necessarily being the most efficient use of capital across the telecommunications system.

We should be able to examine both realities without treating competition itself as the problem.

A Different Way to Think About Broadband Expansion

Perhaps the next phase of broadband development should place greater emphasis on the relationship between infrastructure competition and service competition.

Where new physical routes are necessary, build them.

Where existing infrastructure can be upgraded, use it.

Where shared access can support genuine competition, make that access commercially and operationally workable.

And where investors propose another network in an already well-served market, ask what new value that infrastructure creates beyond redistributing existing customers.

Broadband demand will continue evolving. Networks will need investment, expansion, and modernization.

But the need for more broadband competition should not automatically be confused with the need for more broadband infrastructure.

The industry may have reached a point where the more difficult question is no longer how much fiber we can build.

It is how intelligently we use the infrastructure we already have.