How does this study measure the value of Earth observation, and why use square kilometres, revisits and decisions rather than market-size forecasts?
The short answer. EO is sold in three units, and each tells a different story. Data is sold per square kilometre, at a resolution and a delivery speed. Access is sold per tasking request, per revisit or as capacity on a satellite. Answers are sold per asset monitored, per alert or per decision supported. Setcoin prices each unit from public price lists, contracts and company filings, then uses five models to connect them: how often a constellation sees a place, what a square kilometre costs to produce, what a buyer pays for monitoring, how much of an analytics subscription reaches the imagery supplier, and how exposed a company is to one government customer.
Three units of value
| Unit | What the buyer gets | Typical pricing | Who sells it |
|---|---|---|---|
| Data | Images or radar scenes of an area | $ per km², by resolution and age | Satellite operators, resellers |
| Access | The right to point a satellite where and when needed | Tasking fees, capacity reservations, satellites-as-a-service | Operators |
| Answers | Detected changes, counts, measurements, alerts | $ per asset or site per year, per alert, enterprise licences | Analytics firms, operators with analytics |
Why not start from market forecasts
Market-size reports for EO in 2025 range from about $3.8 billion to about $8 billion, and their 2034 forecasts from about $6 billion to over $15 billion, because they define the market differently: some count only satellite data, some include analytics and geographic information services, some include manufacturing (Chapter 15). Setcoin uses forecasts only to bound the market and builds its view bottom-up from company revenues and contracts.
Evidence used
| Evidence | Examples | Weight |
|---|---|---|
| Company filings | Planet, BlackSky, Satellogic 8-Ks and results | High |
| Government contracts | NRO EOCL, NGA Luno, European defence awards | High (values as announced) |
| Price lists | Planet and Satellogic public prices; reseller rate cards | Medium (list prices, discounts common) |
| Satellite catalogue | Setcoin space engine (CelesTrak SATCAT, 20 Oct 2026) | High for fleet counts; names can undercount |
| Physics | Orbit speed, swath, field of regard, daylight | Standard values |
| Setcoin calculations | Five models | Labelled, formulas shown |
The five models
| Model | Inputs | Output | Chapter |
|---|---|---|---|
| Revisit | Satellites, field of regard, daylight factor | Hours between looks at mid-latitudes | Chapter 9 |
| Constellation cost | Satellites, unit and launch cost, life, operations, km² collected, share sold | Cost per km² sold | Chapter 50 |
| Value chain | Assets, price per asset, images, km² per image, $/km², compute, platform fee | Supplier share and analytics margin | Chapter 52 |
| Buyer's data cost | Area, revisits, price per km², minimum order, processing, analytics fee | Annual cost | Chapter 65 |
| Government stress | Revenue, government share, US intelligence share, cut, growth elsewhere | Revenue next year | Chapter 67 |
Definitions
| Term | Meaning in this study |
|---|---|
| Resolution | Ground sample distance (GSD): the ground size of one pixel |
| Revisit | Average time between opportunities to image a point |
| Latency | Time from collection to delivery |
| Archive | Imagery already collected, sold from a catalogue |
| Tasking | A request to collect a new image of a place and time |
| Sovereign deal | A government buys dedicated capacity, a satellite or a constellation |
Limits
List prices overstate what large buyers pay; government contract values are ceilings; company revenue mixes data, analytics and satellite sales; catalogue names undercount some fleets (notably China's Jilin-1). Ranges are given where the evidence is thin.