Business Model Analysis

The Venture That Looks Commercial Until You Look at It Twice

A pitch deck lands in front of an investor. The company builds satellites. It has revenue. It has customers, apparently several. It speaks the language of platforms, recurring revenue, and capital-light scaling. Six months of diligence later, something deflates: the “several customers” are one government agency and two of its subsidiaries; the recurring revenue is a multi-year cost-plus contract renamed as subscription; the platform is a bespoke system with a single integrator. The venture is not fraudulent. It has simply been described in a vocabulary that obscures what it actually is.

This is not a rare pattern in space. It is a structural one. A sector that spent decades inside cost-reimbursable procurement is now populated by companies describing themselves in commercial terms, often sincerely, without the underlying architecture having changed. The practitioner’s problem is not moral judgment. It is analytical: how do you distinguish a business whose parts fit together from one whose parts merely coexist? How do you tell whether a model is a system or an aggregation? Business Model Analysis, applied with discipline, is the method that answers this question — when it is applied as an analysis rather than a description.

From Strategic Management to the Space Ledger

The modern form of the method is the Business Model Canvas, formalized by Alexander Osterwalder and Yves Pigneur in Business Model Generation (2010). The Canvas consolidated a conversation that had been building through the 1990s and 2000s in the strategic management literature — work on strategic alignment, on revenue architecture, on how firms differ in the way they turn capability into money. Osterwalder and Pigneur’s contribution was to render the conversation visually operable: nine building blocks arranged so that a team could see the whole at once and argue productively about its parts.

The intellectual move predates the Canvas. Michael Porter’s work on activity systems had already argued that competitive advantage was less about discrete choices than about how they fit together. Henry Chesbrough’s writing on open innovation and business models (2003, 2006) had insisted that the same technology monetized through different models produces radically different outcomes. What Osterwalder and Pigneur did was give this tradition a shareable artifact — the Canvas — and, in doing so, made business-model thinking available to teams who would never read the underlying theory.

The space sector adopted the vocabulary late and unevenly. For most of its history, the space industry did not need business models in the commercial sense: capability existed, budgets existed, prime contractors absorbed risk through cost-reimbursement, and the word “customer” meant an agency whose budget cycle shaped every other decision. The commercialization wave of the last fifteen years — launch services sold by the ride, Earth observation sold as imagery subscriptions, communications sold as capacity-on-demand — brought the language of business models into space strategy. It also brought the confusions. Models that evolved inside program offices were re-described as platforms. Anchor-customer dependencies were re-described as reference accounts. The analytical task is to look past the language.

The Canvas Is a Coherence Test, Not a Checklist

The critical misuse of the method is to treat it as an inventory — nine blocks, fill them in, hand in the artifact. This produces descriptions. The analytical move is different: the Canvas is a coherence test, and its value lies in the linkages between blocks rather than the contents of any single one.

What the method actually sees is how the pieces reinforce each other — or fail to. A value proposition is coherent with a customer segment only if the segment is the one that values the specific differentiation offered. A revenue model is coherent with a cost structure only if unit economics improve, rather than degrade, at scale. A set of key resources is coherent with key activities only if the activities can be performed at the cost and cadence the revenue model requires. Each adjacency is a testable claim, and each claim can fail.

The Map Exposes What the Narrative Hides

Consider a commercial in-orbit servicing venture whose pitch is satellite life extension. The value proposition is articulated cleanly: GEO operators facing end-of-fuel events can extend their assets’ revenue-generating life by contracting a servicing mission. The customer segment is defined: established GEO operators with high-value platforms and long replacement cycles. The revenue model is a per-mission fee with optional long-term service contracts.

Translated into the Canvas and examined as a system rather than a list, the model resolves into a specific architecture. The key resource is proximity-operations capability — the autonomy, sensors, and propulsion to approach and dock with an uncooperative satellite safely. The key activity is mission execution: each servicing job is a dedicated campaign with its own vehicle or payload. The key partnership is the launch provider, because each mission depends on getting the servicing vehicle to orbit on schedule and within budget. The cost structure is dominated by per-mission economics: servicing vehicle production, integration, launch, and operations.

Now read the linkages. The revenue model is per-mission. The cost structure is per-mission. The partnership with the launch provider sets the single largest variable cost. A change in launch price flows directly through to unit economics, which flows directly to the viability of the value proposition — because the customer is paying to extend an asset with a known replacement cost, and the servicing price must sit below that threshold to be attractive.

The map has produced a deliverable the narrative could not. In prose, the venture’s risk sounds diffuse: technology risk, execution risk, market risk. In the map, the risk concentrates on a single linkage. The venture’s viability rides on launch pricing staying within a window that keeps per-mission economics below customer replacement cost. If launch prices rise, the value proposition collapses first at the margin and eventually wholesale. If launch prices fall further, the value proposition strengthens — but so does the alternative (replacement over life extension), because cheaper replacement satellites undermine the premise that extending the current asset is the best option.

The non-obvious insight is that the venture is not, structurally, a servicing company. It is a bet on a particular launch-price corridor. An analyst without the map would not phrase it that way. The same analyst with the map cannot phrase it any other way. This is the characteristic deliverable of the method: a reframing of what the business actually is, produced by tracing linkages rather than describing blocks.

Where the Canvas Works, and Where It Misleads

The method’s strength is architectural. It forces the analyst to treat the business as a system, expose the linkages between components, and locate viability in the coherence of the whole rather than the excellence of any part. In space, where technology narratives frequently substitute for business analysis, this discipline is corrective.

Its weaknesses are as specific as its strengths. The Canvas is a snapshot. Early-stage ventures pivot; a model accurate in one quarter may be obsolete in the next. The analyst must revisit the artifact at major inflection points rather than treating it as permanent. Data is often thin for private companies; the analyst must flag confidence levels and resist the temptation to treat inferred components as established facts.

The Canvas can oversimplify multi-sided platforms. A business that creates value by intermediating between two distinct user groups — data buyers and data sellers, launch providers and payload owners, downstream services and constellation operators — has a cross-side value proposition that a single Canvas cannot hold. In these cases, the method should be supplemented with platform-ecosystem analysis that renders both sides and their interactions explicitly. Treating a platform as a one-sided business produces a coherent-looking Canvas and an analytically hollow one.

The method does not capture competitive dynamics. It describes the model; it does not evaluate the pressures bearing on it. Pairing with Porter’s Five Forces or with disruption-theory readings is the correction. Without that pairing, the Canvas can tell a story about a viable architecture sitting inside a market that is silently eroding it.

Finally, government-dependent models have unique dynamics that the framework underweights. Procurement cycles, political risk, and the strategic premium governments place on domestic capability all shape revenue and cost in ways that generic Canvas blocks do not surface. Pair with policy-cycle analysis and with institutional analysis when government dependence is structural, not incidental.

Within the library, business model analysis is a connective tissue method. It feeds market structure work, interacts tightly with disruption theory, and provides the architectural baseline against which regulatory exposure and supply-chain dependency can be traced. Treated as an end in itself, it produces descriptions. Used as a connector to other methods, it produces judgments that hold up under stress.

A Note for the Practitioner