Cable vs Fiber: Nine US Broadband Market Types Across 115 Million Addresses

Sort every address by the operators that can serve it at 100/20 Mbps, count each operator once whatever technology it uses, and few combinations occur. Cable is the starting position nearly everywhere. What varies is the challenger.

The largest single market in the country is cable against a fiber network at 40.0% of addresses. Cable with no qualifying competitor at all still covers 16.9%. Cable facing only fixed wireless covers a further 14.3%, and two or more fiber networks now overlap at 10.9%, a larger share than most operators assume.

Cite this table: G2M Insights, "Cable vs Fiber: Nine US Broadband Market Types Across 115 Million Addresses," G2M Communications Research, edition archetypes-2025-12-v1, g2m.ai/research/communications/cable-vs-fiber-nine-broadband-market-types.

The types separate sharply on setting. Cable against fiber sits at 1,189 people per square kilometer with a median household income of $80,625. The four thinnest types sit between 10 and 23 people per square kilometer. Competition has gone where the households are, which is the same place the revenue is.

Name the market type for every part of your footprint before setting a plan for it. A cable-only area and a cable-versus-fiber area sit four points apart on take-up and need opposite postures: one is a defense of an installed base, the other is a contest for it.

No metro is automatically one market. Across the 189 metros with at least 100,000 addresses, the median metro contains 6 of the nine types, 3 of them holding at least 5 percent of its addresses, and the largest type covers 54%.

How far that goes varies enormously, and it sorts metros into three groups you can act on. In 73 metros a single type covers 60 percent or more of the addresses and one plan is enough. In 83 metros the largest type covers between 40 and 59 percent, so a second plan is needed for a substantial minority of the footprint. In 33 metros no type reaches 40 percent at all.

Cite this table: G2M Insights, "Cable vs Fiber: Nine US Broadband Market Types Across 115 Million Addresses," G2M Communications Research, edition archetypes-2025-12-v1, g2m.ai/research/communications/cable-vs-fiber-nine-broadband-market-types.

Cite this chart: G2M Insights, "Cable vs Fiber: Nine US Broadband Market Types Across 115 Million Addresses," G2M Communications Research, edition archetypes-2025-12-v1, g2m.ai/research/communications/cable-vs-fiber-nine-broadband-market-types.

Philadelphia is 93% one type and can be planned as one market. Phoenix has no type above 25%, six types above 5 percent, and its most common type is a cable monopoly while its second is cable against fiber, which call for opposite postures. Chicago, Detroit, San Francisco and San Diego sit in the same group. A metro-level plan is right for Philadelphia by luck and wrong for Phoenix by construction.

Check which of the three groups each of your metros falls into before deciding how much local variation it needs. A single plan is defensible in Philadelphia or Houston and indefensible in Phoenix or Chicago, where a metro-wide offer is set for a quarter to a third of the footprint and applied to all of it.

The taxonomy reads differently from inside a footprint. Take every address a brand can serve at 100/20 Mbps and sort it by the structure it sits in, and each operator turns out to be running a different number of contests. That is the number a plan has to match.

Cite this table: G2M Insights, "Cable vs Fiber: Nine US Broadband Market Types Across 115 Million Addresses," G2M Communications Research, edition archetypes-2025-12-v1, g2m.ai/research/communications/cable-vs-fiber-nine-broadband-market-types.

The national operators run the narrowest mix, averaging 2.7 effective market types. Regional cable runs the widest at 3.7, and the extremes are wider still: Armstrong faces 5.1 and altafiber faces 1.3. A national brand can plan against one dominant contest and be roughly right. A regional operator that does the same is wrong across most of its footprint.

The shape of a row tells you which business you are in. Fixed wireless operators and rural cable hold the uncontested and unserved columns. Competitive fiber builders hold almost none of them, because they build where a contest already exists. Between those two is where most regional operators sit, running three or four contests at once.

Count your own contests before you count your competitors. If your footprint spans three or four market types, a single retention offer, a single price change and a single sales target are each right for one of them and wrong for the rest, and the cost of that shows up as churn you cannot attribute.

Market type also decides the scale of the company on the other side of the contest. Where cable faces fiber, the smallest operator a household can pick still reaches about 1.0M addresses. Where a single fiber operator serves, the smallest reaches 26k. That is a 40-fold difference in the kind of competitor being faced, inside the same country and often the same state.

The operator population behind this is steeply concentrated. Ranking all 1,568 operators that offer qualifying service by footprint gives a clean power law: between ranks 10 and 1,000 a tenfold drop in rank corresponds to roughly a thirtyfold drop in size, and the fit explains 98.7% of the variation. The market has a small number of very large operators and a long tail of small ones, with little in between.

Cite this chart: G2M Insights, "Cable vs Fiber: Nine US Broadband Market Types Across 115 Million Addresses," G2M Communications Research, edition archetypes-2025-12-v1, g2m.ai/research/communications/cable-vs-fiber-nine-broadband-market-types.

Size does not translate into local pricing power, and the filings show why. Xfinity files a single list price for each of its 7 speed tiers across all 24 states where it reports rates, and 10 of the 14 multi-state rate cards in the survey are uniform in the same way. A competitor holding a few tens of thousands of addresses carries no such constraint and can price street by street.

Cite this table: G2M Insights, "Cable vs Fiber: Nine US Broadband Market Types Across 115 Million Addresses," G2M Communications Research, edition archetypes-2025-12-v1, g2m.ai/research/communications/cable-vs-fiber-nine-broadband-market-types.

Read scale as a constraint on the larger operator, not an advantage. A small challenger can put a sharp offer on a few thousand addresses and lose nothing elsewhere, because it has no national rate card to protect. Matching that across a twenty-state price list is expensive enough that it usually will not happen, which is why targeted local offers work against national operators. If you are the larger operator, the answer has to be a defined local exception with its own approval route, because the standard rate card cannot produce one.

The 6.4% of addresses with no qualifying wireline option are the only places where demand is genuinely unmet, and the shortfall is smaller than the wireline numbers suggest. Wireline take-up there runs between 46% and 50% against 79% where cable faces fiber. Broadband adoption of any kind runs at about 86%, against 93%.

The difference shows up as substitution. Households relying on a cellular connection alone reach 16.5% where no qualifying service exists and 18.3% where only fixed wireless is sold, against 9.8% where cable faces fiber. These are households already paying for connectivity and buying the only product available to them.

Cite this chart: G2M Insights, "Cable vs Fiber: Nine US Broadband Market Types Across 115 Million Addresses," G2M Communications Research, edition archetypes-2025-12-v1, g2m.ai/research/communications/cable-vs-fiber-nine-broadband-market-types.

Treat these areas as conversion, not acquisition. The demand is already spending, so a build case here rests on displacing a cellular connection at a countable household base rather than on persuading households to buy broadband for the first time.

The four findings above resolve into one sequence. Demand comes first and comes from demographics, so the household count in any part of your footprint is knowable before a competitive analysis begins and should not be re-forecast market by market. Structure comes second and decides the contest, so the plan that follows is about share.

How far below the metro that planning has to go is itself a question with an answer per metro. In 73 of the 189 metros a single type covers 60 percent or more of the addresses and one plan is enough. In 33 metros no type reaches 40 percent, and those need three or four. Sorting your own footprint into those groups is a cheaper first step than localizing everything at once.

The lever that matters most is the one the largest operators have given up. A national rate card carrying one price per speed tier across twenty-odd states cannot answer a targeted offer on a few thousand addresses. Small challengers exploit that, and they are effective doing it. Whichever side of the asymmetry you are on, the response is the same in form: a defined local exception, priced for a named market type, with an approval route that does not run through the national card.

The one place where the demand assumption changes is the 6.4% of addresses with no wireline option sold. There the households are already buying connectivity in another form, so a build case rests on conversion at a countable base rather than on market creation.