Who Pays for the Sky

There is a romantic version of agricultural technology in which a smallholder opens an app and a satellite tells her when to irrigate. It is a good demo and a bad business. The farmer with two hectares cannot pay a price that sustains an Earth-observation pipeline, and charging her would be the wrong place to put the meter anyway. The value of seeing a field from orbit does not accrue to the field. It accrues to whoever carries its risk.

The buyer is the balance sheet, not the field

Behind every farm sits a stack of institutions exposed to it. An insurer underwriting a crop policy needs to verify a loss without sending an adjuster down a dirt road. A bank financing the harvest needs to know the collateral is still growing. An exporter shipping to Europe needs deforestation evidence for every parcel in the supply chain. A ministry running a subsidy needs to know it reached real fields. None of them are farmers. All of them have a budget for certainty.

The economics follow from a single fact: the imagery is free. Sentinel-2 and Landsat are public, paid for by taxpayers and re-flown every few days. Our stack computes ten-plus indices — vegetation, water, chlorophyll, surface temperature — from that open data, so the marginal cost of one more report is close to zero. The pipeline is built once and amortizes across every field it ever reads. That is why the unit you sell is a decision-grade report, priced per report, and never the pixels underneath it.

Free imagery is not a discount; it is a different business. When the data costs nothing, you cannot sell pixels — you can only sell the decision the pixels make possible.

What a report has to earn

An institution does not pay for a green map. It pays for a document it can act on and defend. So three things must travel with every index: provenance — which satellite, which dates, which method; confidence — how sure, and where the cloud cover hid the answer; and a recommended action tied to the decision that commissioned it. We sell the analysis and the judgement, not financial advice — the loss is verified, the parcel is flagged, the credit file is informed. What the institution does with its capital remains the institution's call.

We proved the shape on ourselves before we sold it. A full farmland intelligence report for Taiwan's Chishang rice valley — computed from Sentinel-2 by the same pipeline we deploy for insurers, banks and exporters — exists because the marginal cost let it exist. That is the whole thesis of this firm in one report: build the system once, point it at a field, and hand the answer to the balance sheet that needed it. The farmer keeps farming. The sky, finally, has someone to bill. Intelligence, harvested.