Due Diligence in the Dark: An Investor’s Framework for Evaluating Aerial Imagery and EO Companies
The capital influx into Earth Observation (EO) and Geospatial AI is accelerating, but for many private equity and venture firms, the sector remains a “black box.” When every company claims to have a proprietary constellation, 5cm resolution, and “revolutionary” AI extraction, how do you separate the generational winners from the pixel-peddlers?
At Sixty Carlton, we’ve audited the commercial and technical stacks of the industry’s biggest players. For investors evaluating a move into EO in 2026, here is the framework we use to de-risk the deal.
1. The TAM Trap: Distinguishing Between "Pixels" and "Intelligence"
Most companies pitch a massive Total Addressable Market based on the value of the entire geospatial industry. This is a red flag.
The Framework: A company selling raw imagery (pixels) is in a commodity race to the bottom on price. A company selling automated roof condition scores or urbanization indices (intelligence) is a high-margin software business.
The Diligence Question: Is the revenue coming from one-off government contracts for "flyovers," or is it recurring Imagery-as-a-Service (IaaS) integrated into an insurance carrier's underwriting workflow?
2. Hardware Reality Check: IP vs. White-Label
We are seeing a surge in companies claiming proprietary sensor technology that is actually off-the-shelf hardware with a custom shell.
The Framework: Proprietary hardware is only a moat if it leads to a proprietary data advantage (e.g., Vexcel’s multi-modal aerial stack or specific SAR capabilities). Furthermore, hardware alone cannot invent the past; true defensibility often lies in longitudinal data. Nearmap’s competitive advantage, for example, is anchored in its extensive historic products. By offering continuous, high-resolution aerial archives spanning over 15 years, they provide an invaluable temporal baseline for tracking urban evolution and environmental shifts. A new startup launching sensors today cannot retroactively capture a decade of development, making these historical archives an unassailable moat. If a company doesn't own the sensor IP or a vast historical library, they are at the mercy of their suppliers' roadmaps.
The Diligence Question: What is the specific data attribute they capture that Nearmap, Google, or Maxar cannot?
3. The GeoAI Defensibility Stress Test
In 2026, saying you have "AI extraction" is like saying you have "a website." It’s table stakes.
The Framework: Look for "Closed-Loop GeoAI." Does the system get smarter with every pixel it processes? The frontier of GeoAI has evolved from passive extraction to agentic, predictive workflows. EagleView’s April 2026 launch of Eagleview Horizon introduces an agentic AI engine functioning as a virtual geospatial analyst, integrating directly via Model Context Protocols (MCP) to proactively identify property risks and logistical challenges. Concurrently, Vexcel has aggressively optimized its massive data collection—extracting dozens of automated property and roadway attributes across millions of square kilometers—to directly fuel enterprise AI pipelines, proving that an AI model is only as intelligent as its foundational imagery.
The Diligence Question: Can they demonstrate "Semantic Search" capabilities—finding specific features across a continental-scale dataset in seconds—or is it just batch-processing static shapes?
4. Customer Concentration and Vertical Lock-in
EO companies often look healthy on paper because of one massive government or "Big Tech" contract.
The Framework: A healthy EO startup has diversified revenue across Insurance, Utilities, AEC (Architecture, Engineering, Construction), and Local Government.
The Diligence Question: If their top 3 customers left tomorrow, would the company survive? More importantly: Are their customers "using" the data, or is it just "sitting" in a GIS portal?
The Bottom Line
Evaluating an EO company requires more than a financial audit; it requires a structural audit of their data pipeline and GTM motion. The winners in 2026 aren't the ones with the most satellites or planes—they are the ones who have turned the physical world into a queryable, searchable database.
Thinking of investing in the geospatial sector? Let's talk about the diligence that matters.
Join the conversation on LinkedIn