From Moonshots to Platforms: Why Repeatable Space Programmes Outperform One-Off Flagships
Part 2/2: Operating models, contracts and metrics that scale
A platform-first posture is ultimately an operating model choice. It changes how programmes are scoped, contracted, measured, and evolved. The contrast with moonshot-style flagships is not ideology; it is the cumulative evidence on cost, speed, and reliability—and, crucially, on the ability to maintain capability without funding shocks or service gaps.
Starting with the scope, platform portfolios decompose big goals into modular, repeatable services such as launch to multiple orbits, cargo and crew transport, on-orbit tugs, comms relays, and cislunar logistics. Each service rides a common industrial base—even across providers—so improvements compound. The Shuttle was, by design, an integrated system doing many things that today are better unbundled; the result then was high fixed cost and limited flexibility, whereas today’s platform stack scales with demand. GAO’s Shuttle cost work and subsequent OIG/GAO reporting around Commercial Cargo and Commercial Crew underline how unbundling plus milestone contracts produced more predictable cost profiles.
Contracts set the tone. Cost-plus models allocate most risks to the government, reward inputs over outcomes, and reduce pressure to converge on a standard product. Fixed-price, milestone contracts push suppliers to iterate, reuse, and drive down unit costs. The Ansar-Flyvbjerg dataset quantifies the consequence: platform portfolios show near-zero average cost overrun, while bespoke portfolios systematically underestimate and then slip. That variance in outcomes is as important as the point estimates, because finance officials must plan around ranges, not wish-cases.
Cadence is the compounding mechanism. A service that flies monthly will see defects faster, implement fixes sooner, and amortise tooling over more units than a system that flies annually. This is why the “ten times cheaper, twice as fast” finding in the Oxford study is not a one-off quirk; it is how learning curves work when the product is truly repeatable. The same logic explains why NASA’s commercial frameworks, coupled with aggressive pricing by new suppliers and the rise of small satellites, changed the U.S. launch market structure in less than a decade.
Affordability and transparency are not optional in an era of multi-year Artemis budgets and competing national priorities. GAO’s recent SLS review emphasises the need for clearer cost baselines and long-term affordability plans. In a platform portfolio, cost-per-unit, cost-per-kilogram, and cost-per-mission-hour are visible and trendable. In a bespoke portfolio, unit economics are often not directly comparable across missions, and per-launch numbers can climb into headline-grabbing territory that constrains political support.
Talent and industrial base also respond to model choice. The Shuttle’s retirement produced a measurable employment and capability shock across segments, documented in industry and policy analyses. Platform portfolios hedge that cliff risk: suppliers invest in production lines, test stands, and workforces because the production run is open-ended rather than fixed at a handful of units. That continuity is the difference between a cottage industry and a scalable supply chain.
The counter-argument is familiar: some missions are inherently unique, and some national objectives call for sovereign systems. Both points can be true while still organising the bulk of a civil space portfolio around repeatable services. Even heavy-lift can live inside a platform strategy if the industrial approach and contracting model emphasise incremental blocks and common cores rather than clean-sheet hardware for each tranche. The aim is not to eliminate flagships, but to ensure they do not crowd out the platforms that keep the system affordable and resilient.
This reframing has a commercial echo. Investors—public or private—favour models where unit costs decline with volume and where schedule risk is bounded. The Oxford work puts numbers to what markets have already priced: portfolios built on standardised units and iterative improvement produce more launches, more variety, and lower risk of catastrophic overruns. When paired with modern procurement—fixed-price milestones and service buying—those portfolios also expand the supplier set, which is the surest way to keep prices honest.
The Shuttle and SLS are instructive rather than cautionary tales. They delivered real capability and, in the Shuttle’s case, extraordinary missions. The question for the next decade is portfolio balance. If the centre of gravity shifts towards platforms—launch services, cislunar transport, on-orbit servicing, and data backbones—the public sector can buy more science, more exploration, and more resilience for the same money. The evidence base around Shuttle costs, Commercial Crew outcomes, SLS affordability concerns, and platform learning curves points in one direction: repeatable beats bespoke when the mission is to scale.
