We have more eyes on Earth than ever in human history. Thousands of satellites orbit the planet, capturing terabytes of data daily. They see through clouds. They image at night. They measure ground movement in millimetres. They track every forest, every coastline, every ice sheet, every city.
They show us aquifers depleting. Coastlines retreating. Growing zones shifting. Ice sheets collapsing. Land subsiding beneath cities built on the assumption it would stay still.
For the first time in history, we can observe physical reality changing in ways that will reshape where and how humanity can live.
And what do we do with this knowledge?
We ask: How do we make this data bankable? How do we fit it into existing investment criteria? How do we help insurers price risk better, help banks satisfy disclosure requirements, help supply chains optimise for disruption?
We try to make satellites profitable within a financial system designed for a planet that no longer exists.
That’s not a market failure. It’s a window into something deeper: our collective inability to act on what we know.
What Satellites Actually Show Us
When I was building solutions products at ICEYE, the question was always: Where in the process can we leverage Earth observation (EO) data?
That’s the wrong question. Every conversation was about fitting new information into existing frameworks. Finding a slot. Making it digestible for systems that already existed.
The assumption underneath: the frameworks are fine. They just need better data inputs.
But the data was telling us something different. The data was telling us the frameworks are built for a planet that no longer exists.
A 30-year mortgage assumes the land will be there in 30 years. Satellites show coastal erosion trajectories that say otherwise.
Infrastructure investment assumes stable baselines. Predictable flood patterns, consistent growing seasons, reliable water supply. Satellites show those baselines shifting faster than planning cycles.
Insurance pricing assumes you can model risk against historical patterns with some adjustment for trend. Satellites show discontinuities, not trends, but systems flipping from one state to another.
Supply chain optimisation assumes you can source from the same regions with logistical adjustments. Satellites show entire agricultural zones becoming unviable within a generation.
The insight isn’t here’s data to make your existing decisions slightly better. The insight is your decision architecture is built for a world that’s already gone.
That’s not an optimisation input. That’s a call to transformation.
The Language of Evasion
The Earth observation industry has developed a vocabulary that obscures this gap. Climate adaptation. Climate resilience. Climate risk. Companies use these terms interchangeably. Even the Intergovernmental Panel on Climate Change (IPCC) acknowledges inconsistent usage across its own reports.[1]
The vagueness is functional.
When a company claims to serve climate resilience, that could mean flood imagery for faster insurance claims. Or carbon credit verification. Or regulatory disclosure compliance. Or disaster response optimisation.
All of these help existing systems work slightly better. None of them questions whether those systems should continue.
Resilience sounds transformative. It implies bouncing back, adapting, thriving through change. But the products sold under that label mostly help the current system absorb shocks and continue. They’re resilience for the financial framework, not for communities or ecosystems.
Real resilience would mean some places acknowledging they need to be abandoned. Some industries acknowledging they’re incompatible with a livable planet. Some assets acknowledging they’re stranded regardless of what the balance sheet says.
That’s not what anyone’s selling. Because there’s no market for it.
The Optimisation Trap
The pressure to optimise rather than transform is immense. And rational.
Defence and intelligence represent 24-40% of Earth observation revenue.[2] Government overall accounts for roughly three-quarters of the market.[3] The commercial climate applications everyone talks about in pitch decks? Roughly 7-11%.[4]
Defence has clear budget owners, predictable procurement, and sophisticated buyers who know what they want. Climate transformation has none of that. No budget owner. No procurement category. No process for absorbing insight that says ‘your decision architecture is obsolete.’
So founders build what the system can purchase: optimisation tools.
First Street embeds climate risk scores in Zillow, Redfin, and Realtor.com listings.[5] Homebuyers see a flood risk rating and buy a slightly less exposed house. The system of 30-year mortgages on climate-vulnerable land continues unchanged.
Jupiter Intelligence serves three of the five largest US banks with climate analytics for regulatory disclosure.[6] The banks report their exposure with better precision. They keep lending to climate-exposed assets, just with improved footnotes.
GHGSat finds methane leaks so oil and gas operators can fix them.[7] The leaks get plugged, the waste gets reduced, the extraction continues.
These are commercial successes. The founders aren’t villains. They’re responding rationally to what the market will buy.
The market will buy optimisation. It won’t buy transformation.
Why Transformation Has No Buyer
The companies that tried to sell transformation failed.
Gro Intelligence raised over $125 million to show food and agriculture companies that entire sourcing regions would become unviable.[8] They made TIME’s 100 Most Influential Companies list.[9] By late May 2024, they’d shut down. Valuation collapsed from $850 million to under $25 million.[10]
Cervest raised $36 million to provide asset-level climate intelligence.[11] The UK government featured them as one of ten companies “at the forefront of” AI innovation.[12] Three months later, administration.[13]
Both had real technology. Both had genuine insight. Both had customers who found the information valuable. What they didn’t have: buyers who could act on transformation.
Nobody pays to be told they’re wrong.
A bank doesn’t want software that says ‘your mortgage book is built on a planet that doesn’t exist anymore.’ They want software that helps them keep lending with better risk footnotes.
A farmer doesn’t want intelligence showing their land becomes unviable in 15 years. They want precision agriculture that improves this season’s yield.
A real estate investor doesn’t want tools for managed divestment. They want risk scores that help them buy slightly less exposed properties while continuing to build.
And the beneficiaries of transformation, the communities that need to relocate, the workers who need to retrain, the ecosystems that need to recover, they can’t pay. That’s partly why they’re vulnerable in the first place.
There’s no budget line for ‘our entire decision architecture is wrong.’ There’s no request for proposal for ‘help us become a fundamentally different company.’ There’s no procurement category for managed decline.
Capitalism is structured around growth. Every budget assumes continuation and expansion. Ending things isn’t a line item.
We Have Everything Except Willingness
Here’s what we actually have:
We have the knowledge. Satellites track 745 million emission sources through Climate TRACE.[14] SERVIR provided climate data across 45+ countries before being defunded in 2025.[15] CLIMADA models risk for any location on Earth.[16] We know exactly what’s happening, where, and how fast.
We have planning tools. Buy-In Community Planning helps design managed retreat programs.[17] The Georgetown Climate Centre provides retreat planning guides. The Carbon Risk Real Estate Monitor (CRREM) shows exactly when buildings will become stranded assets.[18] The playbooks exist.
We have the technology to stop producing harm. Renewables are now cheaper than fossil fuels in most places. Heat pumps work. Electric vehicles work. Regenerative agriculture exists. We know how to build buildings that don’t leak energy. The technology to transition isn’t missing.
We even have the capital. Climate finance flows exceeded $2 trillion in 2024.[19] Trillions more sit in fossil fuel subsidies that could be redirected. The money exists.
So why isn’t transformation happening?
Look closer at those ‘transformation tools.’ They’re information tools. They tell you what’s happening. They don’t make change happen.
Climate TRACE shows where emissions are. It doesn’t stop them.
CRREM tells investors when buildings will strand. It doesn’t make them divest early.
Buy-In helps plan buyouts. It doesn’t fund them, and it doesn’t make people want to leave their homes.
The gap isn’t information. It’s action.
We’re not transforming because we don’t want to.
Transformation means loss. Losing homes. Losing jobs. Losing investments. Losing identities. Losing ways of life.
Homeowners don’t want to leave. Investors don’t want to take losses early. Workers don’t want to retrain. Fossil fuel companies don’t want to close. Politicians don’t want to mandate unpopular change. Everyone prefers to believe it won’t be that bad.
So we optimise. We squeeze a few more years out of the current system. We document the decline with ever-more-precise dashboards. We disclose the risks and continue taking them.
We have everything we need except the willingness to accept that some things must end.
Removal vs. Transition
Even when funding tries to do the right thing, it bends toward continuation.
Consider the Frontier coalition — Stripe, Alphabet, Meta, Shopify, McKinsey committing over $1 billion to purchase carbon removal through 2030.[20] The mechanism is genuinely innovative: advance purchase commitments for technology that does not yet exist but must. It worked. Charm Industrial has delivered over 6,200 tonnes of permanent carbon removal, more than any other company at the time, because these commitments made it viable.[21]
But look at who’s buying: companies running massive data centres, global logistics, cloud infrastructure. They’re pre-purchasing removal so they can claim progress toward ‘net zero’ while continuing to grow.
It’s sophisticated offsetting. Pay for removal here; keep emitting there. The carbon math might work on paper. But it’s not a transformation. It’s buying a license to continue.
Here’s the sharper distinction: removal lets the current system continue. Transition ends the harmful parts.
We already have technology to not emit in the first place. Renewables are cheaper than fossil fuels in most places. Electrification works. We know how to build without burning.
But transition means changing things. Retiring assets. Closing facilities. Accepting that some businesses need to end.
Removal is easier. Don’t change what you’re doing; add cleanup. That’s why we fund it.
Transition is harder. Stop doing this; do that instead. That’s why we avoid it.
The money flows to removal because removal doesn’t threaten the current system. Transition does.
The Allocation We’ve Chosen
Consider how we actually spend:
Fossil fuel subsidies: approximately $7 trillion per year.[22]
Global climate adaptation needs: approximately $215-387 billion per year.[23]
We could fund the entire global adaptation requirement many times over with what we currently spend subsidising the problem we need to solve.
This isn’t a trade-off with security or growth or any other competing priority. This is paying to make things worse. $7 trillion annually to keep fossil fuels artificially cheap, to keep extraction profitable, to keep the current system running past its expiration date.
EO companies pivot to government and defence contracts because those buyers have clear budget owners and procurement processes. That’s rational. The question isn’t why companies follow the money. The question is why, as a society, we actively fund our own destruction.
We have the capital for transformation. We’re choosing to spend it on continuation, and worse, on subsidising the very industries that make transformation necessary.
The satellites see all of this. They watch the emissions rise. They track the ice sheets calve. They measure the land subside. They document what $7 trillion in annual subsidies produces: a planet that’s changing faster than our institutions can respond.
The Obligation
No one is coming to save us.
Not markets. Markets optimise for existing preferences, and existing preferences are for continuation.
Not governments. Not fast enough, not with the current political economy.
Not technology. The technology exists. What’s missing isn’t innovation.
What’s missing is us.
The paths forward exist. They’re narrow. They require different capital structures, different timelines, different definitions of success. They require being honest about what’s actually optimisation dressed as transformation, and what’s actually transformation, which rarely looks like a startup.
It’s been done. risQ focused narrowly on climate risk for municipal bonds, positioned for regulatory tailwinds, and was acquired by Intercontinental Exchange (ICE) — not a unicorn outcome, but a path that worked.[24] The transformation tools that exist and work are mostly public goods: nonprofit infrastructure, government programs, open-source tools, coalition-funded initiatives.
The paths are real. They’re just not the paths most people are looking for.
They require willingness to accept loss. To fund endings, not just cleanups. To let go of what needs to go.
The satellites are showing us reality.
What we do with that knowledge, as founders, as investors, as citizens, as humans sharing the same thin atmosphere, is the only question that matters now.
References
[1] IPCC, “Guidance Note: The concept of risk in the IPCC Sixth Assessment Report,” 2021.
[2] Mordor Intelligence, “Satellite-based Earth Observation Market Report,” 2024; Novaspace/Euroconsult, “Earth Observation Data & Services Market Report,” 17th Edition, November 2024.
[3] Novaspace/ESA, “New Trends and Dynamics of the EO Market,” presentation at ESA EO Commercialisation Days, November 2024.
[4] Novaspace/ESA market segmentation, 2023.
[5] Zillow Group, investor press release, September 26, 2024.
[6] Jupiter Intelligence, company website and investor materials, 2024-2025.
[7] GHGSat, company website; EPA Alternative Test Method documentation.
[8] Semafor, “Gro Intelligence startup shuts down,” June 4, 2024.
[9] TIME Magazine, “TIME100 Most Influential Companies,” April 27, 2021.
[10] Semafor, June 4, 2024; Forge Global valuation data.
[11] Dealroom.co, Cervest company profile; TechCrunch, May 2021.
[12] UK Department for Science, Innovation and Technology, March 29, 2023.
[13] Evening Standard; Yahoo News, June 2023.
[14] Climate TRACE, coalition website and database, 2024.
[15] NASA Applied Sciences, “About SERVIR”; NASA Watch, “USAID Erasure Impact: NASA Halts SERVIR Solicitations,” March 2025.
[16] ETH Zürich, CLIMADA platform documentation.
[17] Buy-In Community Planning, company website; MIT Solve application.
[18] GRESB/CRREM, “Carbon Risk Real Estate Monitor,” 2024.
[19] Climate Policy Initiative, “Global Landscape of Climate Finance 2025.”
[20] Frontier Climate, coalition website.
[21] Charm Industrial, company reports and carbon ledger; CNBC, May 18, 2023.
[22] International Monetary Fund, “IMF Fossil Fuel Subsidies Data: 2023 Update,” August 2023.
[23] UNEP, “Adaptation Gap Report 2023” and “Adaptation Gap Report 2024.”
[24] Bond Buyer, “ICE expands muni reach,” December 9, 2021; Northeastern University, “risQ Acquired by Intercontinental Exchange.”