Every satellite operator faces this question eventually — usually around the time they're finishing their CDR and someone in the finance team asks why the ground segment line item is bigger than the spacecraft itself. Build your own ground infrastructure, or use a shared service? The answer depends on a set of factors that are specific to your mission, and getting the analysis wrong in either direction is expensive. Building when you should have bought costs millions and months. Buying when you should have built creates operational dependencies and coverage limitations that compound over time.
I spent 14 months building a proprietary ground station before co-founding Orbitvein, so I have a reasonably direct view of what the build path actually costs and what it delivers. Let me walk through the structured analysis we use when operators ask us this question.
The True Cost of Building Proprietary Ground Infrastructure
The cost of a single custom ground station is not the antenna price. The antenna — a 3.7 m S/X-band parabolic with a motor controller and feed assembly — runs roughly $120,000-$180,000 for a commercial system. But that's maybe 15% of total project cost. The full build-out for a single operational station typically runs $1.4M-$1.8M when you account for everything:
| Cost Component | Typical Range | Notes |
|---|---|---|
| Antenna + feed + mount | $120K–$180K | Commercial off-the-shelf or custom for specific bands |
| RF chain (LNA, downconverter, SDR) | $80K–$150K | Varies significantly by band count and quality tier |
| Site infrastructure (power, shelter, HVAC) | $200K–$400K | Highly location-dependent; polar sites significantly more expensive |
| Backhaul connectivity | $50K–$120K upfront + $30K–$80K/year opex | Low-latency fiber at remote sites can be the most expensive single line item |
| Software integration (scheduling, decode chain, ops interface) | $200K–$400K | Typically 6-12 months of software engineer time |
| Frequency licensing and regulatory | $50K–$150K | Legal fees, FCC application, ITU coordination costs |
| Commissioning and integration testing | $100K–$200K | Travel, personnel time, first-pass testing |
Add it up and you get $800K-$1.6M for a well-executed single-site build. Then add the timeline. In our experience, the engineering procurement construction timeline for a new commercial ground station runs 12-18 months from contract award to first operational contact. And a single mid-latitude station typically covers 3-6 contact windows per day for a LEO satellite at 550 km — which means polar coverage gaps of 6+ hours at a time.
What the Service Model Actually Costs
A shared ground network charges per contact-minute of actual station time used. Current market pricing for S-band TT&C contacts runs roughly $8-$20 per contact-minute depending on station location, band, and volume commitment. X-band payload downlink at high data rates commands $25-$60 per contact-minute for the same reasons.
For a single LEO satellite with 4-6 contacts per day at 8 minutes average per contact, annual service cost at $15/contact-minute runs approximately $260,000-$440,000/year. That sounds significant — but compare it against the capital cost of the proprietary alternative. At $1.5M capital plus $100K/year opex (staff, power, maintenance), a proprietary single station pays back versus service in 4-5 years, assuming the station delivers equivalent coverage and availability. Most proprietary single-station builds do not deliver equivalent coverage because they cover only one orbital plane region.
The break-even math looks more favorable to the build case when you run 50+ satellites through a single station. At that scale, the per-satellite cost of owned infrastructure drops below what any shared service can offer. But early-stage operators with 1-20 satellites are nowhere near that scale, and the capital required to reach it comes long before the economics justify it.
Coverage Is the Hidden Variable
Capital cost comparison between build and buy is the easy part of the analysis. Coverage is where the math gets nuanced.
A single ground station at a mid-latitude location — say, El Segundo, California — provides excellent coverage for satellites with orbital inclinations below about 50 degrees. For a sun-synchronous orbit at 97.8 degrees inclination, the same station sees maybe 2-3 contacts per day per satellite, and only during windows when the orbital plane crosses the California longitude. A 6-hour communication blackout is entirely possible with a single mid-latitude station for a high-inclination orbit.
Operators building a sun-synchronous constellation for Earth observation need polar station coverage — Alaska, Svalbard, Antarctica, Tierra del Fuego. Building a polar station is materially more expensive than a mid-latitude station. Svalbard infrastructure costs run 2-3x the continental equivalent due to logistics and climate requirements. The coverage math for a 10-satellite SSO constellation that requires full daily revisit may require 3-4 stations at different latitudes, which means $4M-$6M in build capital before the first satellite launches.
This is the scenario where the service model's economics are most compelling. A shared network with 40+ stations distributed across all latitudes delivers the multi-site coverage that would require 3-4 proprietary builds, at a fraction of the capital cost and with none of the 12-18 month construction timeline.
Operational Risk: Build vs Buy
Capital cost and coverage are the two primary financial variables. Operational risk is the third, and it's the one that operators underestimate most consistently.
A proprietary ground station is a point of failure. When the LNA fails, the receive chain goes down. When the backhaul provider has an outage, the station is dark. When the motor controller firmware needs an update, you need an engineer on-site or a skilled remote technician. Station availability for a well-maintained proprietary installation typically runs 96-98% — which sounds good, but a 2% outage rate on a single-station operation means roughly 7 days per year of unavailability distributed unpredictably across your contact windows.
A shared network distributes this risk across multiple stations. When one station has a hardware issue, the scheduling system routes the contact to an alternate station with sufficient coverage. The operator sees the same contact window and the same data delivery — the network has absorbed the station failure internally. Network-level availability in a well-operated shared ground network is materially higher than single-station proprietary operations, because the redundancy is built into the architecture rather than requiring you to build it yourself.
When the Build Case Makes Sense
There are legitimate scenarios where proprietary ground infrastructure is the right answer:
- Government/classified missions that require COMSEC-accredited facilities, controlled physical access, and ITAR-classified handling procedures. A shared commercial network cannot meet these requirements.
- Very high data-rate missions with 50+ Mbps sustained downlink requirements that would generate $2M+/year in service fees. At that scale, owned infrastructure amortizes faster.
- Operators with 100+ satellites and mature constellation operations who have sufficient scale to staff and maintain a proprietary network efficiently, and whose orbital geometry concentrates passes over predictable locations.
- Missions requiring custom RF configurations not supported by commercial shared networks — unusual frequency allocations, non-standard modulations, or specialized antenna apertures for high-gain links.
For early-stage operators outside these categories — which describes most commercial NewSpace operators in the 1-50 satellite range — the build case requires a very specific set of conditions to be economically justified before the constellation reaches maturity. Most teams that have done the honest analysis find that service is the right answer for the first 3-5 years of operations, with a revisit of the build option once satellite count and data volume projections become more certain.
The decision framework is straightforward: define your coverage requirements, model your contact-minute demand across 1, 3, and 5-year scenarios, price both options on a total cost basis including capital, opex, and timeline risk, and add a risk-adjusted value for coverage flexibility. For most early-stage missions, the analysis comes out clearly in favor of starting with service and preserving capital for the spacecraft, software, and team — the parts of the mission that actually differentiate the business.
