TerraWatch Pro ยท ยท 6 min read

Earth Observation Weekly Briefing - April 20, 2026

Three Ways EO Companies Solve the Commercial Go-to-Market Problem

Welcome to another subscriber-exclusive, Pro edition of the TerraWatch newsletter. If you know someone who would benefit from this, please encourage them to subscribe so they can receive the newsletter every week.

If you want to receive the monthly deep-dives and get access to the archive, upgrade to a Premium subscription.


๐Ÿ“ˆ EO Market Signals

๐Ÿ’ฐ Deals & Funding

๐ŸŽฏ Moves & Strategy

๐Ÿ’ก
The opportunity: NASA is inviting EO companies to build their own instruments to fly on government science missions. If the data rights are structured so that industry can use the data commercially, this is effectively subsidised constellation infrastructure. The cost of getting to orbit gets shared; the commercial upside stays private. That is a fundamentally different relationship from buying commercial data.

The risk: Earth science is surviving politically by attaching itself to exploration โ€” instruments designed for vegetation monitoring on Earth are being pitched as lunar resource identifiers. That is clever politics, but it means instrument design gets optimised for dual-use convenience rather than scientific precision. If that foundational layer degrades because instruments are designed for the Moon first and Earth second, everyone downstream feels it.

๐Ÿ›ฐ๏ธ Other Signals


๐Ÿ’ก TerraWatch Insights

Three Ways EO Companies Solve the Commercial Go-to-Market Problem

The EO industry keeps borrowing the SaaS playbook: build a platform, offer an API, let customers self-serve. But EO data doesn't behave like SaaS. A hospital can configure a CRM tool, they understand the workflow, they know what they need. An agribusiness looking at a vegetation index doesn't know what to do with it. What threshold triggers action? How does this relate to their crop models? What does a change in NDVI mean for soybeans in this soil at this growth stage?

The knowledge to make EO data useful doesn't live inside the commercial customer. It lives in the gap between the EO company and the customer. Government buyers don't have this problem: defence and intelligence agencies have trained imagery analysts, established GIS workflows, and decades of experience with satellite data. In other words, they can self-serve. Commercial buyers such as insurers, ag cooperatives or utility companies cannot.

This isn't theoretical. Planet's commercial revenue share dropped from 45% to 18% over four years as government and defence surged. Iceye's โ‚ฌ250M revenues year was driven almost entirely by sovereign contracts. The EO companies making real progress commercially are the ones that solved the domain access problem, and how they solved it tells you a lot about where commercial EO is actually heading.

That asymmetry is the core go-to-market problem in commercial EO, and no amount of better sensors or smarter AI solves it on its own. Every EO company generating meaningful commercial revenue has had to find a way around it. The mechanisms differ, but they fall into three categories: partner with domain experts who own the customer relationship, co-build products with them, or acquire the domain expertise outright. Each comes with different tradeoffs around control, defensibility, and scale.

Distribution Partnership

This is the most common model: a market leader in a specific vertical that already sits inside the customer's workflow integrates EO data into its own platform. The end buyer may never know satellite data is involved. The EO company gets reach it could never build alone and the domain partner gets a data layer that differentiates its existing product.

What makes this work is that the domain partner handles everything the EO company cannot such as translating the data into the buyer's language, embedding it in operational workflows, navigating industry-specific regulations, and carrying the commercial relationship. The EO company provides the data under a licensing or integration agreement and stays upstream.

Iceye has built the most systematic version of this in insurance. Its SAR-derived flood data sits inside Munich Re's risk platform, where underwriters access it through Munich Re's interface. Munich Re also resells Iceye's full product suite (flood, hurricane, wildfire) to its global client network. Iceye has replicated this across the insurance value chain for companies such as AXA, Aon, Swiss Re, MAPFRE, Insurity, Juniper Re. Planet follows the same pattern in agriculture as seen from Planet data embedded into Syngenta's Cropwise and Bayer's global operations as well as in insurance through AXA's claims and risk platform.

Tellingly, Planet's own sales teams acknowledged that they are now "primarily focused on government customers" while using partners to serve commercial sectors. In other words, one of the most prominent EO companies acknowledged that the direct-to-commercial model doesn't scale without domain partners.

Product Partnership

This is a rarer model, but becomes more defensible. Imagine both sides invest in building a jointly defined, sellable product โ€“ not just data embedded into an existing workflow โ€“ but an actual new product that requires both the EO capability and the domain capability to exist. Replacing one side means redesigning the product, not swapping a data feed.

The distinction matters because it changes the power dynamic. In a distribution partnership, the domain company can theoretically switch EO providers. In a product partnership, the EO company's data and the domain company's analytics are fused into something neither can deliver independently, creating mutual dependency.

Planet and SynMax jointly offer Theia, a maritime intelligence product that fuses Planet's daily imagery with SynMax's AI-driven vessel detection, classification, and dark ship tracking. Each side markets the other's capabilities, and they jointly win government contracts. Planet can't build maritime AI and SynMax does not operate an EO constellation. In short, the product exists because they collaborate.

The Vantor and Windward partnership follow the same logic: fusing satellite persistent monitoring with maritime behavioural AI to detect and track dark vessels. EarthDaily and Liberty Mutual Reinsurance are co-building parametric insurance products where satellite-derived triggers automate payouts โ€” a defined product that requires both EO data and actuarial design.

It is worth noting where product partnerships happen where the output is a clearly definable product with measurable triggers and delivery promises. Agriculture and traditional insurance remain dominated by distribution partnerships because the output is harder to package as a discrete, jointly owned product. The vertical has to be ready for a co-built solution, not just for data.

Acquired Expertise

This is the most risky model and obviously the least common. Rather than partnering to gain domain expertise, some EO companies acquire it outright โ€“ buying the teams, the customer relationships, and the vertical expertise they need to reach commercial markets directly.

The logic is clear: if domain expertise is the bottleneck, owning it removes the dependency on partners who could switch EO providers or renegotiate terms. It lets you capture the full margin from pixel to decision. And it lets you move faster: no partnership negotiations, no joint product alignment and no split incentives.

EarthDaily is the most interesting example. They acquired: SkyForest, which brought forestry and wildfire expertise, Descartes Labs which brought insurance and mining expertise. Combined with EarthDaily Agro's 35 years of agricultural analytics, they now own the domain expertise across five verticals, alongside the satellites and the platform.

As you can imagine, this is also the most capital-intensive path and the riskiest. You are betting you picked the right verticals and the right companies. If insurance or mining doesn't scale as an EO market, you own assets that are not generating returns. And you need to retain the domain teams you acquired; their expertise is the whole point, and it walks out the door if they leave. Most EO companies can't afford this strategy. EarthDaily can because Antarctica Capital backs it. But that is not a model everyone can replicate โ€“ which is why distribution and product partnerships remain the more common paths.

Main Takeaway

There appears to be three paths to the same destination: partner for domain access (Iceye, Planet), or co-build products with domain experts (SynMax, Windward), or if you can, acquire the expertise entirely (EarthDaily). Every EO company generating meaningful commercial revenue seems to have chosen one or more of these paths. More interestingly, not many seems to have had success by going direct to commercial buyers with a platform and an API alone.

The implication for evaluating EO companies is straightforward: look at the partnership and acquisition portfolio, not the sensor or the algorithm. The depth and breadth of domain relationships (whether partnered or owned) is the leading indicator of commercial trajectory. The companies without them are the most dependent on government contracts, and the most exposed when those contracts cycle.


Until next time,

Aravind.

Read next