Spend enough time around the Earth observation industry and you notice a pattern: every few months it produces a new theory of why commercial demand has not arrived, each more sophisticated than the last. Three are in circulation now. That the prize is no longer the best imagery but ownership of the reasoning layer built on top of every geospatial signal. That the markets are ready and what persists is an “understanding gap,” buyers who have not yet been shown the value. That the industry needs a mindset change: stop chasing the latest satellite capability, start thinking about the value it can generate for customers.
Three theories, all conceding the thing the industry spent a decade refusing to concede: capability is not the constraint. That concession is correct, and overdue. Then each of them reaches for the wrong cause.
EO is never the only thing. It is one source among many, and not even the only kind that senses: radar, optical and hyperspectral from orbit; the same instruments on aircraft and drones; ground stations, gauges, samples; and then the models, the market data, the buyer’s own records. A satellite product is one input into a decision that is already drawing on a dozen others, and it has been for as long as anyone has built a product on it. So the value is not what the satellite can do. It is what the observation adds to the job the buyer is actually doing.
And the buyer is not doing one job. Sometimes the job is to see that something has changed: a coastline that has moved, an asset degrading against its own history. Sometimes it is to explain it: whose plume this is, which actor cleared the forest. Sometimes it is to act on it: size the reserve, order the evacuation. Sometimes it is to vouch for it: that a shipment really is deforestation-free, that emissions really are below the line. And at the far end, to set the rule others are held to: the threshold itself. These are different jobs, and an EO product is worth a different amount in each, because the worth tracks how much rides on the answer. At one end sits awareness nobody has to act on. At the other sits a figure a bank or a regulator will hold you to. The product is the same. The worth is not.
But doing the job well is not the point either. A monitoring feed nobody acts on, a forecast that changes no plan, a dashboard admired and never used: the job was done, and nothing moved. An EO product is worth what it moves. Where nothing moves, the picture can be flawless and it is still scenery.
This is why the understanding gap is the most comfortable of the three answers. You cannot explain a budget owner into existence. When a buyer is on the hook for a number, nobody has to teach them to want a better one; they find it, they integrate it, they pay, because the job is already theirs. The places where EO sits unbought are rarely places where the buyer failed to understand. They are places where nobody had to act on the answer. “Showing the value” carries an assumption underneath it, that the value is already there, fully formed and merely badly explained. Often there is nothing to show, because there is no job with consequences for it to attach to. The understanding gap is the supply side’s flattering story for missing demand, the way slow traction becomes “an early market.”
It is also why the call for a mindset change, sincere as it is, stops one step short. “Use what we have better, generate more value for customers” keeps the EO company as the agent and the satellite at the centre. The chair never moves. That is optimisation: squeeze more from the existing model. But the assumption that broke is that the value is something the industry produces. Optimise against a broken assumption and the position gets worse: you grow more skilled at selling a number nobody is on the hook for. The change of mind that would matter is the one an EO company finds hardest to make: the value is not something we produce and hand over; it is generated only when the observation moves a job the customer is accountable for, and sometimes that job does not need a satellite at all. Saying so, this is not a job for us, is the line between selling a thing and doing a job. A project shop takes every opportunity that comes, never says no, bends however far it has to, never fires a customer. The point is to sell the deliverable and book it, whether or not anything moves. Data, model, insight: the name on the invoice changes, the reflex does not. A product company declines the work that is not its own and turns down buyers it cannot serve, because the point is the job, not the sale, and the job is not done until something moves. It is the most ordinary product discipline there is, and the industry treats it as unthinkable.
And I know there will be someone who objects. That education is sometimes the real blocker: a buyer with an accountable job and a budget who has simply not connected it to EO. Or that capability has to exist before demand can find it, because you cannot want what is not built. Both are sometimes true. But strategy is a loop: find the gap, take your bet, make a coherent plan, execute, and let the result send you back to the gap. And each objection is a claim about the gap. One gap is a buyer who has the job and has not connected it. Another is a capability that does not exist yet. A third is that there is no accountable job at all, and it is the one the industry keeps mistaking for the other two, because those two it can close by explaining or building. One question tells the three apart: is there a job someone must answer for, and does this move it. When the bet does not pay, a working loop asks it again. This one does not; it reaches for a new theory and bets on the same misread. Which is why the theories keep coming, every few months a new one.
The buyers, when you listen to them, are already there. A senior climate-capital allocator described the gap to me without once mentioning data quality: geospatial has no unit of transaction. The intelligence is good, and it sits on a dashboard, looked at from time to time, because there is nothing to underwrite against it. A founder building exactly the geospatial model the market prizes is quietly moving his own money out of the model layer and into the physical assets his satellites watch. Soft asset to hard asset, because that is where the value he can hold sits. Capabilities get built and then shelved, not for want of accuracy but for want of a buyer: a monitoring tool that worked, abandoned because no one had to report what it measured, so no one paid to see it.
Which leaves the part none of the three theories will say, because it is the one that costs a supplier a sale. The expertise in EO is the judgement to tell which job carries consequences and which does not, and to say when the honest answer is that it adds nothing. Methane a satellite reads on a clear day and loses under cloud is a confidence interval nobody is forced to act on; a blind spot you can predict is not an enforcement instrument. A reading that lands on a dashboard and changes no decision anyone answers for is decoration at high resolution. An EO product is worth what someone will stake a decision on, and the most useful thing I can tell a buyer is, sometimes, not this.
So the awakening is half-finished. The industry has seen that the data was never the point, and gone looking for the point in the next layer up: the reasoning, the messaging, the mindset of the seller. It sits one layer down instead, and on the other side of the table. Beyond the ladder of worth sits one more move, a different kind of thing: whoever sets the rule everyone else is held to owns what counts as a right answer. That is power over the frame, not a larger move inside it, and whether anyone can hold that power privately or it always ends up public is a piece of its own. Everywhere below, the unasked question is the same: who is accountable for the difference the observation shows, and whether, this time, the answer needs a satellite at all.