SpaceX Reaches Orbit. UK Modular Struggles to Break Ground. Why?
What SpaceX can teach us about an industry that keeps getting in its own way.
George Palmer, 17/03/2026
Image credit: SpaceX
The controversial actions of SpaceX's CEO aside, the technical and commercial achievements of the company are genuinely impressive. In 2025, SpaceX put more kilograms of payload into orbit than every other space agency on earth, both public and private combined. And it beat the global collective by a factor of five.
Max Olson has spent years trying to answer a deceptively simple question about how: if the methods are public, the principles openly discussed, and the engineering philosophy explained in interviews and factory tours, why has nobody copied it?
Until October 2025, I worked inside a UK modular housebuilder. When we closed, people lost money and people lost jobs. Reading Olson's essay Atoms are Cheap, Process is Pricey(written as an introduction to his forthcoming book on the same subject) I couldn't stop drawing parallels. Not between rockets and housing in any literal sense. But between the systemic failure modes of two industries trying, and largely struggling, to build hard things at scale. Parts of the comparison are uncomfortable. I think they are ultimately useful.
The insight that changes everything
Before SpaceX existed, its founder Elon Musk travelled to Moscow to buy intercontinental ballistic missiles, hoping to repurpose them for a Mars mission. The Russians quoted him prices that made the whole venture unworkable. Leaving empty-handed, he was forced to reframe the problem entirely. The raw materials in a typical rocket: the aluminium, the titanium, the carbon fibre, amount to roughly 2% of what you pay for the finished product. The other 98% is process: supplier markups stacking through contract layers, custom designs that never reach manufacturing scale, and single-use hardware thrown away after every launch.
The response was not to negotiate harder. It was to ask a different question entirely. Not "what do rockets cost?" but "what should rockets cost?"
Now substitute aerospace aluminium for timber, steel, and plasterboard. A modular home is not a complicated object. The materials are cheap, widely available, and well understood. A factory-built home should, in theory, cost less than a site-built one: tighter tolerances, less waste, faster build times, lower labour costs. This is not a controversial claim. It is the founding premise of the entire modern methods of construction movement.
And yet. Ilke Homes collapsed in June 2023 owing £320 million, with over 1,000 staff made redundant. Legal & General's modular division accumulated losses of £279 million over eight years before halting production entirely. House by Urban Splash went into administration in 2022. The government's own MMC Taskforce, allocated £10 million to advance data and standards across the sector, never once convened a single meeting.
The subsequent House of Lords inquiry concluded, with some understatement, that the UK government needed "a more coherent and determined approach." Public money had been committed to individual companies without a coherent strategy, without measurable objectives, and without anybody systematically asking why modular homes consistently cost more to deliver than the theory said they should. The subsidies went in. The inquiry reports came out. The underlying cost problem was never properly addressed.
The atoms were not the problem. The absence of a system-wide approach was.
The flywheel that SpaceX built and most of modular housing missed
Olson's most important insight is not about any single SpaceX innovation. It is about the system they built, and how its parts reinforce each other.
First-principles thinking identified the waste. Vertical integration gave SpaceX control over the process to eliminate it. Standardisation drove costs low enough that customers came in volume, and that volume made the economics of vertical integration viable. Volume then enabled rapid iteration and tight feedback loops, which made rocket reusability achievable, fundamentally transforming the cost structure of an industry that had thrown away its hardware after every flight for sixty years.
Each priority supported the others. Effort applied to one increased the speed and ease with which the others were achieved. Olson calls this a flywheel, the same virtuous cycle that sits at the heart of Amazon's operating culture. It is also, in different vocabulary, what the UK modular housebuilding industry has needed and rarely found. The firms that have come closest to building it are, not coincidentally, still standing.
Take Tide Construction and its manufacturing arm, Vision Modular Systems. Tide operates as the developer and main contractor; Vision handles offsite manufacturing, primarily for the build-to-rent and student accommodation markets. By owning both functions under one roof, they cut multiple links out of the supply chain along with the compounding mark-ups each supplier would otherwise extract. There is no negotiation between parties whose incentives only partially align. The process is controlled end to end, and the efficiency gains stay in the system rather than being captured by individual contractors at each handover point.
The results follow logically and the benefits are not purely financial. When the whole project lifecycle sits within one organisation, innovation benefits everyone rather than being hoarded for competitive advantage at a single tier. Vision and Tide report a 45% reduction in carbon emissions, construction time cut in half, and an 80% reduction in site waste. They have delivered some of the world's tallest modular buildings. They are still building.
A different expression of the same logic appears in Reds10, which has built its business around vertical integration across public sector frameworks covering education, defence, and justice. Demand that runs on political cycles rather than pure market forces provides a more predictable pipeline and gives the manufacturer the security to invest properly in industrialised processes. Pivoting across different framework categories demands operational agility from the entire supply chain, and when that supply chain reports to the same organisation, production planning can be holistic and group-level resistance is buttressed.
Image credit: Andrew McCarthy
Now compare that with what failed. The companies that collapsed were not short of ambition or capital. Ilke had £100 million in fresh investment as recently as December 2022. L&G had the backing of one of Britain's largest institutional investors and Tophat were backed by Goldman Sachs. It is worth acknowledging that all modular housebuilders, like conventional housebuilders, are exposed to macroeconomic forces of market slowdowns but carry a heavier overhead of a manufacturing facility.
But what the all failed companies lacked was a self-supporting flywheel. Large factories were built and optimised for scale production before the pipeline of work existed to fill them. Without volume, prices were marginally higher than traditional construction rather than meaningfully lower. Without a lower cost to offer clients in exchange for standardised design, most projects remained unique. Vast amounts of money and time were spent innovating and optimising the modular structure, then attempting to wedge it into the commercial and contractual ecosystem of the conventional housebuilding industry. A system built for an entirely different way of working.
Bespoke designs were offered where standard products should have been the starting point. Fragmented supply chains meant risk was priced in at every tier, with no mechanism for taking a holistic view of where that risk sat or how to manage it efficiently. The learning from one project did not compound into the next.
As Bill Hughes, chair of Legal & General Modular Homes, acknowledged when announcing the factory closure: "Without the necessary scale of pipeline, it is not sustainable to continue producing more modules." He was right. But pipeline scarcity was not simply bad luck or a hostile market. It was, in significant part, a consequence of never solving the cost problem that would have generated the demand.
SpaceX broke this cycle by using its lower price point to persuade its principal client, NASA, to adapt to the rocket rather than demand the rocket adapt to them. The 5-metre fairing became an industry standard not because it was optimal for every payload, but because SpaceX made it the default and told the market to adjust. Most did.
UK modular builders faced the opposite pressure. Every client wants something specific to their project. Planning departments carried semi-justified anxieties about cookie-cutter housing. Housing associations wanted particular unit mixes. Local authorities wanted schemes that respected local character. Every concession to customisation was a vote against the standardisation that would have made the economics work. And without the lower costs that standardisation enables, there is no compelling case to put to the client in exchange for their flexibility.
This is the trap. Modular housebuilding will always struggle to beat conventional construction when it demands standardisation but cannot offer meaningfully lower costs to justify it. The only other route is the kind of end-to-end automation that Japan's Sekisui House (the world's largest volumetric modular housebuilder) has achieved over decades of patient capital investment. But that level of commitment also requires volume. Which you only generate through lower cost. Which requires standardisation and vertical integration. The flywheel, turning in a different direction, but the same flywheel.
You can optimise factory efficiency until the cows come home, but if you do not have system-wide transparency you cannot guarantee that the improvements you make to the modular structure are not simultaneously multiplying the complexity of the groundworks. Without vertical integration, system-wide efficiency remains elusive because every layer of the supply chain is pricing in its own margin and risk. More importantly, without full commercial transparency across the supply chain, it is impossible to build a culture where mistakes and inefficient design are addressed openly. In the fragmented model, the contractor who raises a problem frequently absorbs the cost of fixing it while the rest of the chain moves on. The rational response is silence.
Failing in public and what it costs not to
One of Olson's more striking observations concerns SpaceX's relationship with failure. The company published compilation videos of its rockets crashing into droneships, set to music. This was not bravado. It was a signal that visible failure is acceptable, even necessary, when you extract the lessons and move fast on the next iteration.
Image credit: US Launch Report
I find this section of the essay the most personally pointed.
The culture of the UK construction industry, modular manufacturers included, does not treat failure as data. Failure is contra-charged and sometimes litigated, and rarely analysed. Lessons are absorbed, at best, by the business that paid for them, or more often quietly set aside as the industry moves on to the next project with fingers crossed. When something new is attempted and does not quite work, the instinct is not to iterate. It is to conclude that new things are dangerous and return to bricks and mortar. The industry's relationship with transparency is, to put it charitably, complicated.
SpaceX's engineering philosophy rests on a simple premise: your model is always wrong in ways you do not yet know. The only way to discover where it breaks is to test it against reality, often and cheaply. A high production rate is not just good for manufacturing economics. It is a learning rate. More units means more data. More data means better designs. Better designs mean simpler, cheaper units. The cycle compounds.
The lean manufacturing culture native to many module manufacturers, where a problem spotted on the line stops production, and the team swarms to find the root cause, this approach seldom translates to the construction site. A building project is typically delivered by a collection of subcontractors, each completing their scope and moving on, that collection of contractors may never work alongside each other again. That structure creates a powerful incentive not to raise problems. Better to leave an issue for the next trade to inherit than to flag it and own the delay. The feedback loop does not just slow down. It inverts. Problems are hidden rather than surfaced, and the system gets no smarter.
This is the cultural gap that vertical integration is partly designed to close. When the people spotting problems and the people fixing them work for the same organisation, the incentive structure changes. Failure becomes information rather than liability.
What the flywheel actually requires
Olson is careful to point out that strategy alone is not enough. SpaceX's methods are learnable. The culture is the harder part to transplant.
In modular housing, the strategic logic has been understood for years. Volume enables investment in the factory. Factory investment reduces unit costs. Lower unit costs enable more competitive pricing. Competitive pricing generates more volume. The flywheel is not a secret. It has been described in industry reports, parliamentary inquiries, and investor presentations for the better part of fifteen years.
What it requires, and what has proved elusive, is the discipline to defend standardisation under commercial pressure, to place price reduction (not just cost reduction) at the centre of every initiative, the patience to build volume before optimising margin, and the cultural disposition to treat problems as opportunities to improve the process, regardless of where in the supply chain they emerge.
The SpaceX lesson is the underlying logic: when process is the cost, you have to control the process. When volume is the enabler, you must target the cost that secures the volume. When standardisation facilitates the volumel, you cannot let the market customise it to death. When failure provides critical data points; fail quickly, cheaply and publicly.
The atoms really are cheap. A modular home does not require aerospace-grade materials or precision-machined components. The materials are ordinary. The design, for the most part, is not experimental.
The fragmented, convoluted process is the problem. The companies that optimised for end-to-end control are still standing. The ones that did not are the subject of parliamentary inquiries and administrator reports.
Rockets and houses are not the same thing. But the lesson is.
