The Allocation
On July 22, 2026, Indonesia's Ministry of Communication and Digital (Komdigi) finalized the selection of winners for the 700 MHz and 2.6 GHz spectrum bands, closing a process that began with an April 2026 auction and ran through an objection period that closed without a single formal challenge. Telkomsel, Indosat, and XLSmart — the operator formed from the 2025 merger of XL Axiata and Smartfren — each walked away with spectrum in both bands (Komdigi/Postel).
In the 700 MHz band, XLSmart secured the largest block at 2x15 MHz for a winning bid of Rp1.06 trillion, while Telkomsel and Indosat each took 2x10 MHz for Rp642.5 billion and Rp507.48 billion respectively. In 2.6 GHz, the ranking flips: Telkomsel took the largest slice at 80 MHz (Rp545.84 billion), Indosat 60 MHz (Rp372 billion), and XLSmart 50 MHz (Rp231.6 billion) (Liputan6).
Why These Bands, and Why Now
700 MHz is Indonesia's "digital dividend" — spectrum freed by the analog-to-digital television transition — prized for its ability to carry a signal long distances and through solid obstacles, which is exactly what's needed to reach the archipelago's more remote geography. 2.6 GHz sits at the other end of the trade-off: shorter range, but far more capacity, the workhorse band for dense urban 5G traffic. Regulators structured this selection under Ministerial Decrees No. 175 and 176 of 2026 specifically to pair a coverage band with a capacity band in the same award (Komdigi Jatim). That pairing matters because Indonesia has, until now, run its mobile networks almost entirely on 2100 MHz and 2300 MHz — bands already saturated by 4G traffic, leaving little room for 5G to actually differentiate itself from existing service.
Attached to the licenses are binding, not aspirational, obligations: 4G service must reach 538 villages currently without it, 5G must cover at least 51% of the national population by the fifth year of the license, and average mobile broadband speeds must hit 100 Mbps by 2029 (Postel; CNN Indonesia).
The Case for Mandates
The strongest argument for hard-wiring coverage targets into the license itself, rather than trusting operators to build out rural networks on commercial logic alone, is that Indonesia has already tried the alternative. Left to ordinary return-on-capital math, three competing carriers will always prioritize Jakarta, Surabaya, and Medan over a village in East Nusa Tenggara, because the payback period on a rural tower can run a decade or more against dense-urban buildout measured in months. A binding 538-village obligation, tied to a license an operator cannot walk away from without losing spectrum it paid for, closes that gap in a way that competition alone has not. Regulators are not wrong that market incentives and universal-service goals frequently diverge, and a five-year clock creates real accountability that a voluntary rural-coverage pledge would not.
The Revenue Problem
Where the design gets harder to defend is the state's own cut. Komdigi has projected roughly Rp6 trillion in initial non-tax state revenue (PNBP) from this selection, rising to Rp29.9 trillion over ten years (CNN Indonesia). Combined with the roughly Rp3.36 trillion the three operators bid upfront across both bands, that is capital extracted from the same balance sheets that are simultaneously obligated to fund towers, backhaul, and base stations in 538 villages. Every rupiah an operator sends to the treasury as spectrum rent is a rupiah not spent laying fiber to a village that the license now legally requires it to reach. Regulators can mandate coverage or they can maximize auction revenue; treating both as costless is the policy sleight of hand worth watching over the next five years.
The rigidity of the 100 Mbps-by-2029 target is a second concern. Locking a specific speed figure into a license written in 2026 assumes today's radio technology, spectrum efficiency, and demand patterns hold for half a decade — in a sector where standards bodies, chipset vendors, and even satellite-terrestrial integration are moving faster than that. A target set too low becomes irrelevant; one set too high becomes an excuse for retroactive relief negotiations that undermine the credibility of the obligation altogether.
A Better Balance
None of this argues against spectrum obligations as a tool — used narrowly, they are one of the few levers that reliably gets service to villages the market ignores. But Komdigi would serve both goals better by treating the PNBP take as a variable to be minimized rather than maximized within a fixed-obligation framework, and by pairing hard coverage counts (villages connected) with technology-neutral speed floors rather than a single fixed Mbps figure five years out. Indonesia's operators now have the spectrum they were starved of. Whether the state's own revenue appetite becomes the next bottleneck is the story to watch through 2029.