Satellite broadband is most valuable where India’s other networks struggle to reach

The most valuable Indian satellite broadband connection may serve a clinic, school or relay site that has few reliable alternatives. Selling a terminal to a well-connected household is a different proposition. TRAI’s 5 November recommendations on authorisation and spectrum assignment for satellite communication network providers offer a chance to distinguish those markets. The case for easier entry is strongest where new network capacity has a clear cost or reliability advantage over extending the terrestrial network. TRAI, November satellite network recommendations · TRAI, April 2026 consultation paper

Capacity, access and the retail bill

A satellite network provider operates infrastructure. The business that sells a connection to a school or household may use that infrastructure under a wholesale arrangement. The distinction matters because licensing the upstream provider can increase physical capacity while leaving a single downstream supplier in a location. A policy that makes multiple networks technically possible can still produce limited effective competition if switching requires a costly terminal replacement or a new installation contract. Publishing comparable coverage and contract terms helps customers assess which option is real.

Capacity also has a geography and a clock. A shared radio beam serving several customers has a finite throughput at peak times. The marginal cost of adding a customer can be low when traffic is light and appreciable when the beam is congested. This creates different incentives for operators serving dispersed sites and operators selling high-volume urban subscriptions. Average speed in a brochure will be a poor test of the service an institution receives during the busiest hour.

Compare the whole service

Take a hypothetical settlement of 80 households. A reliable terrestrial link and local distribution would cost ₹24 lakh upfront and ₹3 lakh a year to operate. A shared satellite installation costs ₹4 lakh upfront and ₹6 lakh a year. Ignoring discounting and equipment replacement, the five-year totals are ₹39 lakh and ₹34 lakh. By year seven the totals are ₹45 lakh and ₹46 lakh. The ranking changes over a plausible project horizon even with no change in unit prices. These are illustrative inputs. A serious comparison must adjust for uptime, capacity, latency, replacement costs and the chance of additional customers.

The example also shows why household affordability differs from network cost. Dividing ₹34 lakh by 80 households and five years gives an average of ₹8,500 a household per year under full take-up. If only 40 households subscribe and the shared cost is unchanged, it doubles. Early demand can be too thin to justify separate retail connections. An anchor customer, such as a health centre buying a service with a clear uptime target, can support a local access point. A contract with portable equipment or defined transition terms can protect the community if fibre subsequently becomes cheaper.

An additional test is the cost of keeping a backup connection available. A low-volume backup can prevent substantial service interruptions while generating little monthly traffic. A procurement based on cost per gigabyte would undervalue that option. Availability during an outage and the time needed to restore service are better measures for this use.

Put the subsidy where access is scarce

The strongest case for a wider rollout is that uniform offerings simplify contracts and help firms learn quickly. Terrestrial networks also fail in places that appear well covered on a map, and a satellite backup can have value there. That argues for evaluating the service an area actually receives. Where public funds are involved, an open comparison across technology and shared procurement for remote public institutions could direct spending towards unmet demand without predetermining the network.

The authorisation debate should therefore be judged by wholesale terms, repeat orders from remote sites, installation delays, busy-hour performance and total cost per functioning connection. Evidence that many customers in underserved areas can switch providers without replacing expensive hardware would make a stronger case for relying on ordinary retail competition. Persistent concentration and low take-up at full price would support competitive procurement for specific services. The aim is dependable connectivity in places where geography makes it expensive to provide, with the provider paid for the service it delivers.