Much of the confusion around “dual-use” in the space economy comes from treating it as a customer mix problem rather than a design problem. Too often, dual-use is framed as a question of whether a system can serve both government and commercial customers. In practice, whether a capability ever becomes market-driven is largely determined much earlier — by architectural decisions made before demand is proven, before revenue materializes, and often before it is even clear who the long-term customers will be.
This distinction matters because many of the most expensive mistakes in the space economy do not stem from technical failure or lack of demand. They stem from systems that were designed in ways that quietly foreclosed market formation while assuming markets would emerge later.
Where Optionality Is Won or Lost: An Architectural Scorecard
Optionality means the freedom to follow real demand, wherever it actually emerges. Whether a space capability can transition from missions to markets is not determined by customer mix or go-to-market strategy. It is determined by a small number of architectural and structural decisions, typically made early, that either preserve or foreclose future options — often long before demand is visible.
Drawing on the design principles introduced in the Space-Based Data Centers analysis, the following scorecard makes explicit where optionality is actually created or destroyed. It is not a checklist for success, but a diagnostic tool for surfacing irreversible architectural tradeoffs that are often made implicitly.
Markets are not added later; they are either preserved early or quietly foreclosed.
The Optionality Scorecard
1. Multi-payer architecture
Core question: Can the system serve two or more materially different customer classes through configuration rather than redesign?
Optionality is preserved when modular architectures, overlapping requirements, and partial-utilization economics are designed in from the start. It is foreclosed when systems are bespoke-optimized for a single mission profile, or when the business case assumes that one anchor customer funds the full deployment.
2. Repeatability over perfection
Core question: Is the system designed to improve through iteration, or optimized for singular performance?
Architectures that favor repeatability explicitly model learning curves, standardize components, and accept early performance tradeoffs in exchange for scalability. Optionality erodes when manufacturing processes, cost structures, and operations assume that the first implementation is also the production model.
3. Robustness to substitutes
Core question: Does the value proposition remain viable if terrestrial or alternative solutions improve faster than expected?
Optionality is preserved when systems exploit non-substitutable orbital advantages and clearly articulate break-even conditions. It is foreclosed when viability depends on external constraints persisting, or on competitors failing to improve.
4. Execution risk allocation
Core question: Who captures the upside when costs fall or performance improves?
Architectures paired with fixed-price or performance-based models tend to drive learning and productization. Cost-plus or build-to-spec structures may make sense for early R&D, but they often mute the incentives required for markets to form.
5. Architectural flexibility
Core question: Do early, hard-to-reverse choices create future pathways or quietly close them off?
Security models, orbital configurations, power and data architectures, and interface decisions are often irreversible. Optionality is preserved when these choices allow segmentation, co-tenancy, and adaptation to unanticipated uses — and foreclosed when optimized narrowly for a single mission.
Taken together, these dimensions reveal three recurring outcomes across the space economy:
- Optionality-preserved systems, which retain the freedom to serve adjacent markets without major re-architecture.
- Optionality-naive systems, where early design choices unintentionally align the system with a narrow set of users while assuming markets will emerge later.
- Mission-locked systems, where optionality is intentionally foreclosed because mission requirements dominate by design.
Applying the Scorecard: Three Archetypes
The scorecard reveals patterns when applied to actual systems:
With that lens in place, Earth observation provides one of the clearest illustrations of how these dynamics play out in practice.
Optionality-Naive: Earth Observation and the Cost of Optimizing for the Wrong User
For more than a decade, the EO sector has operated under a widely shared assumption: that commercial markets in areas like precision agriculture, energy, insurance, logistics, and climate analytics would scale naturally as imagery quality improved. Higher resolution, better tasking, and more sophisticated sensors were treated as the primary unlocks for downstream demand.
That belief shaped a dominant architectural pattern.
Most commercial EO systems, including the WorldView constellation operated by Maxar Intelligence (now Vantor), DigitalGlobe's satellites prior to the 2017 Maxar merger, and Airbus Defence and Space's Pléiades constellation, were designed around narrow fields of view delivering sub-50cm ground sample distance (GSD), with WorldView-3 achieving 31cm panchromatic resolution. Tasking can be time-critical (e.g., collection within 24 hours under certain tasking options), with delivery often within ~48 hours of collection depending on product and workflow. These architectural choices make perfect sense for government and intelligence users, who value exquisite performance, control, and responsiveness — and who are willing to pay premium prices (often priced in the tens of dollars per km² for 30 cm-class products (varies widely by terms, product, and volume)).
Optionality dies in the architecture phase.
In practice, the opposite happened.
Commercial EO users generally do not need the highest possible resolution, nor do they want to manage bespoke tasking workflows. They need consistent coverage, predictable cadence, standardized products, and data that integrates cleanly into operational systems. When EO architectures are optimized for tasking and perfection rather than coverage and repeatability, commercial adoption becomes slow, fragmented, and expensive — even if the underlying technology is world‑class.
This is what optionality-naive looks like in EO: commercial demand is anticipated, but the system is optimized around the preferences of the users who pay earliest and most reliably. Over time, those early design choices quietly foreclose the very markets the industry expected to grow into.
Architectures optimized for the earliest paying users often foreclose the markets expected to emerge later.
The result is a familiar pattern. Governments become the dominant customers not because commercial demand is fictitious, but because the architecture increasingly fits government needs better than commercial workflows. What began as a market-seeking strategy drifts, unintentionally, into mission alignment — without ever being acknowledged as such.
By 2022, the last year of public reporting before Advent International acquired and later separated the company, the Earth-imagery business then operating as Maxar Intelligence (now Vantor) generated approximately $1.6 billion in revenue, heavily weighted toward government customers, primarily U.S. defense and intelligence agencies, despite more than two decades of effort to build commercial markets in precision agriculture, insurance, infrastructure monitoring, and energy sectors. The same architectural decisions that made government customers willing to pay premium prices for on-demand, high-resolution tasking made systematic coverage uneconomical, creating a structural barrier to commercial adoption where applications required frequent monitoring rather than occasional high-detail snapshots.
EO is not unusual because markets failed to appear. It is unusual because it so clearly demonstrates how optimizing too early for the wrong users determines who the system ultimately serves, regardless of intent.
Optionality Preserved: Planet as a Disciplined Counterfactual
That makes EO especially useful for examining a counterexample—not one that "proved the market," but one that preserved the ability to respond honestly to where demand actually emerged.
Planet Labs made a different set of tradeoffs early on. Rather than chasing maximum resolution and bespoke tasking, Planet prioritized broad field of view, high cadence, and standardized data products. These choices limited certain near-term opportunities, but they preserved something more important: the ability to serve multiple classes of users without redesigning the system.
Evaluating Planet Against the Optionality Framework
Multi-Payer Design: Planet's Dove constellation (3-5m resolution) and acquired SkySat assets (~50cm resolution) could serve both systematic monitoring users such as agriculture, infrastructure, and climate analytics, and government customers who valued cadence and global coverage over maximum resolution. The same satellite platforms and data products served commercial and government buyers without requiring different architectures or segregated constellations.
Repeatability: The Dove satellites were explicitly designed for mass production using the CubeSat form factor (3U) with commercial off-the-shelf (COTS) components where feasible. Manufacturing costs declined dramatically across generations as the company achieved production scale and manufacturing learning. The architecture anticipated building and operating 100+ satellites, treating iteration as the default rather than the exception.
Substitute Robustness: Planet's 3-5m resolution occupied a defensible value wedge between free government alternatives (Sentinel-2 at 10-20m, Landsat at 30m) and premium tasked imagery. For applications requiring daily or near-daily monitoring, such as crop health tracking, infrastructure change detection, disaster response, the systematic coverage model provided value that free lower-resolution data couldn't match, while remaining economical enough to support subscription-based pricing rather than per-image tasking fees.
Risk Allocation: Planet owned its constellation and captured the benefits of manufacturing improvements and operational efficiencies. Rather than building satellites to customer specifications under cost-plus contracts, the company sold data as a service through subscription models and API access. This structure internalized execution risk but also allowed Planet to benefit directly from cost reductions as production scaled.
Architectural Flexibility: The CubeSat platform architecture preserved pathways to serve different sensing modalities and mission profiles. Initially focused on optical imaging with the Dove constellation, Planet later integrated higher-resolution SkySat capabilities (acquired from Google/Terra Bella in 2017) and explored radar sensing. The standardized bus architecture and service-oriented delivery model allowed the company to add capabilities without redesigning core systems or delivery infrastructure.
Where Optionality Was Exercised
As commercial EO demand proved slower and thinner than early industry narratives suggested, Planet rationally leaned into defense and civil government customers. Crucially, that shift did not require a new constellation, segregated architectures, or mission-specific system redesign. Government customers — including NGA, NRO, and allied defense agencies — buy essentially the same products as commercial users: systematic coverage delivered through standardized APIs, analytics pipelines, and cloud-based platforms.
Planet's optionality-preserving architecture didn't eliminate government dependence. By recent estimates, government customers represent approximately 70-80% of the company's revenue. But the architecture enabled government demand without foreclosing commercial participation. The same platform serves both customer classes, allowing Planet to capture government stability while maintaining the ability to serve commercial users when applications mature or new use cases emerge.
Government demand can stabilize a commercial architecture — without turning it into a mission-locked system.
Planet's experience does not demonstrate that commercial EO markets are large or fast-growing. It demonstrates something more subtle and more important: that government demand can de-risk and stabilize a commercial architecture without turning it into a mission-locked one. The architecture preserved optionality; the market determined how much of it could be exercised.
Mission-Locked by Design: When Markets Aren’t the Objective
Preserving optionality should not be mistaken for insisting that every space capability must become a market. Some systems are mission-locked by design, and pretending otherwise obscures the real tradeoffs at play.
There are entire classes of space systems, particularly in national security, strategic sensing, and assured communications, where mission requirements dominate system design. These capabilities are optimized for performance, assurance, resilience, and control, often for a single or very small number of sovereign customers. Cost, scale, and flexibility are secondary considerations by necessity, not oversight.
Examples include strategic missile warning, highly classified ISR, protected military SATCOM, and sovereign PNT augmentation in contested environments. These are not failures of commercialization. They are mission-locked outcomes, and in many cases they are exactly what governments intend to build and sustain.
The problem arises when investors or policymakers price these systems as if market optionality exists despite architectures that explicitly foreclose it. In those cases, disappointment is not the result of poor execution. It is the result of mismatched expectations: confusing mission success with market formation.
Mission-locked systems can be durable, valuable businesses. They simply obey a different logic.
What This Means for Decisions Being Made Now
The most consequential decisions in the space economy are not about whether a capability is "dual-use" in theory. They are about whether its architecture preserves the freedom to respond when demand materializes unevenly — or locks that freedom out entirely.
Several categories of space infrastructure are at precisely this inflection point right now: the 12-24 month window where architectural decisions either preserve or foreclose market optionality:
Space-based data centers (often termed "orbital data centers" in industry discussions, though we use the broader term to include cislunar and beyond-orbital deployments): Multiple companies and sovereign investors are finalizing architectural choices around power systems, thermal management, compute modularity, security segmentation, and orbital configuration. The five scorecard dimensions map directly to whether these platforms will serve only their anchor customers or enable broader markets. Companies designing for single-mission optimization, such as classified government compute with no architectural pathway to commercial co-tenancy, are making explicit choices to foreclose optionality. Those preserving multi-payer design, repeatability, and architectural flexibility are betting that adjacent demand will materialize. The difference is being locked in now, in power budgets, thermal architectures, and security models that will be difficult or impossible to reverse once systems are deployed.
Proliferated LEO for national security: The U.S. Space Development Agency's Tranche program, proliferated missile warning and tracking architectures, and resilient military SATCOM initiatives represent multi-billion-dollars in government investment over the next decade. Some program structures emphasize open standards, commercial data rights, and vendor competition (optionality-preserved). Others optimize for single-program performance with proprietary interfaces, program-specific security architectures, and vendor lock-in (optionality-naive or mission-optimized by intent). The structural difference will determine which capabilities remain government-locked and which can serve intelligence community, allied, or eventually commercial customers through the same basic architecture.
Next-generation Earth observation: Synthetic aperture radar (SAR), hyperspectral, thermal imaging, and RF sensing constellations are being architected now by companies including Capella Space, Umbra, Pixxel, HawkEye 360, and others. Companies that preserve optionality across the five scorecard dimensions, designing for multi-payer use cases, building repeatability into manufacturing, and maintaining robustness against improving free alternatives, will be positioned to serve commercial markets if and when they achieve scale. Those that optimize exclusively for government intelligence workflows, bespoke tasking, and maximum performance will remain structurally dependent on defense and intelligence budgets regardless of technical excellence.
In-space logistics and servicing: Satellite servicing, orbital mobility, debris removal, and inspection capabilities from companies such as Northrop Grumman (MEV), Astroscale, Orbit Fab, and Starfish Space are early enough that architectural decisions haven't yet fully hardened. This is the rare case where optionality can be designed from first principles rather than recovered after deployment. The companies that preserve multi-payer architecture (government servicing, commercial life extension, debris removal), build for repeatability (standardized interfaces, learning curves), and maintain flexibility (multiple orbital regimes, diverse mission profiles) will define the market structure for the next two decades. Those that optimize for single anchor customers or bespoke mission profiles may succeed as contractors but will struggle to scale beyond their initial programs.
The pattern is consistent across these domains: optionality is cheapest to preserve early and prohibitively expensive to recover later. For companies and investors making long-lived infrastructure commitments over the next 12-24 months, the five scorecard questions — multi-payer design, repeatability, substitute robustness, risk allocation, architectural flexibility — should be asked explicitly before architectural decisions harden, not assumed to resolve themselves during go-to-market execution.
Some systems should be mission-optimized. That's often the right answer when performance requirements genuinely preclude flexibility, when government is the only plausible long-term customer, or when mission criticality justifies accepting vendor lock-in and bespoke architectures. But that should be a deliberate choice backed by clear strategic analysis, not an accident that becomes visible only years later when commercial markets fail to materialize despite early assumptions.
Confusion between dual-use intent and dual-use capability has misallocated billions across the space economy. The distinction lies in architecture, not aspiration. And architecture is determined in the decisions being made right now.
