Optus's regional 5G buildout is where the coverage math gets costly

Optus has switched on 5G in Moree, the agricultural and thermal-spring town in New South Wales, extending a network that now reaches 94.5 percent of Australia's population across more than 6,000 sites and furnishing a regional tally in the same announcement: more than 100 locations upgraded across regional NSW in the past 12 months, and another 100 NSW sites due by the end of next year.

The announcement does not carry a price—no cost for Moree, no per-site economics for what comes next, no equipment or financing detail—and that absence leaves the coverage percentage doing work a capital figure normally does. Population coverage is the easiest number in telecom to publish and the hardest to underwrite, because it pools the dense urban sites Optus has been building since its 2019 city-first launch with the sparse regional ones now coming online, and says nothing about the revenue behind either. It also defines the remainder precisely: 5.5 percent of the population, the rural and remote tail that costs the most to reach and is the least likely to headline a press release.

Optus's own account puts growth of the regional program at 82 percent in 16 months, with 2,444 regional sites expected to be upgraded by 2030. The towns those sites serve are smaller than the cities already counted inside the 94.5 percent, which suggests revenue per new site is thinner than the average the coverage number implies, and that cost per covered person climbs as coverage approaches completeness. Moree, described by the carrier as historically agricultural and known for its thermal springs, is a fair illustration of the customer base the next phase chases. Whether the residual pays is a subscriber question rather than an infrastructure one, and a local general manager's line about connectivity helping communities thrive is as close as the announcement comes to putting a rationale on it.

Front-end costs and approval queues, not capital, now set the AI data center calendar, and mobile substitutes density for permitting as the binding constraint, keeping the lesson intact: money is rarely the thing that holds a buildout back, and the marginal site is the one that has to be justified rather than assumed.

Anchor-contracted digital assets are the only ones that earn infrastructure pricing; everything else fights for capital. A regional 5G site sits at the far end of that second bucket—no anchor tenant, no contracted offtake, revenue from retail subscribers in towns whose headcount is pooled with Sydney's in the headline statistic. Carriers can announce percentages instead of dollars because these assets never have to be marked to a buyer, and a mobile network, unlike a data center campus or a fiber route, has no comparable sale that forces the number out. Optus has told the market how many sites and what share of the population; the 100 NSW sites due by the end of next year are the next occasion to say what each one costs.