This article was written with the assistance of AI and is part of [[The Translation Series]]. Orginally published in [Linkedin](https://www.linkedin.com/pulse/what-australias-business-model-elliot-duff-v70zc/) on April 4, 2026 from the original in [Obsidian](https://publish.obsidian.md/elliotduff/Published/Articles/What+is+Australia's+Business+Model%3F)  > _And has anyone ever written it down?_ There has been much discussion that the failure of innovation in Australia is a translation problem - a failure to commercialise. I have made this argument myself. It is the leading argument of the SERD. But I would now argue that the translation problem is a symptom, not the disease. The underlying problem is that Australia does not have a business model for the nation. You cannot translate to a destination that has not been defined. You cannot commercialise into a market that has never been described. And you cannot fix a pipeline when no one has agreed what should come out the other end. The translation problem is real. But it sits inside a larger failure — the failure to ask what Australia is actually building, for whom, and how it intends to capture the value. That is a business model question. And it is the question this article attempts to answer. If you were to fill in Australia's current innovation canvas honestly, it would look something like this (see image at top): - **Value Proposition** — research outputs, publications, trained graduates, grant programs - **Customer** — the research sector itself, universities, research institutes - **Channel** — peer review, grant committees, government programs - **Revenue** — government funding, tuition fees, indirect cost recovery - **Key Partners** — other universities, ARC, NHMRC, Department of Education - **Key Activities** — research, publication, grant writing, reporting - **Customer Relationship** — compliance-based, reporting to funders not to market The canvas is internally consistent. The problem is the customer is wrong. And when the customer is wrong, everything that follows — the value proposition, the channels, the revenue model, the key activities — is optimised for the wrong outcome. That is where thirty years of innovation policy has left us. Not through malice or incompetence, but through the accumulated logic of a canvas no one has ever explicitly drawn. ## What is a Business Model Every person has a business model. Not in the corporate sense — but in the fundamental sense. How you spend your time. What skills you invest in. What relationships you build. What you are willing to trade, and what you are not. These choices, made consciously or not, constitute a model for extracting value from the resources you have, in the situation you find yourself. Most people never articulate it. They live it instead — accumulating experience, adjusting as they go, occasionally stepping back to ask whether the model is still working. Some never ask at all. Organisations are no different. Every company, every institution, every government department is operating a business model — a set of assumptions about who the customer is, what value is being created, and how that value is captured. The model exists whether or not anyone has written it down. The **Business Model Canvas** — developed by Alexander Osterwalder — is a tool for making that model visible. Nine boxes. One page. It forces clarity on the questions that determine whether a model succeeds or fails. It does not create the business model. It articulates and communicates the one that already exists — and in doing so, makes it possible to examine, challenge, and improve it. An unexamined business model is one you cannot improve. One you cannot communicate. One you cannot hold anyone accountable to. I was introduced to the Business Model Canvas in Silicon Valley more than a decade ago, as I embarked on the commercialisation of my own research. It was a revelation — not because it was complicated, but because it wasn't. A single page that forced clarity on questions I had never been asked to answer in thirty-six years of publicly funded research. Who is the customer? What problem are you solving for them? How do you capture the value you create? Nobody had ever asked me those questions. Not at CSIRO. Not in the grant applications. Not in the performance reviews. The system I worked in was extraordinarily good at producing knowledge. It had no language for what happened next. That is not a personal failing. It is a system design. And it is the design that needs to change. Which brings us to Australia. ### The Four Sectors of the Economy Every modern economy operates across four sectors. Each has its own business model logic. Each creates value in a different way, for different customers, through different mechanisms. And each of Australia's four sector business models is broken — in a different way, for a different reason, at a different point in the value chain. ### 1. Primary Industries — The Royalty Model Under Stress The primary sector business model is the simplest to articulate. You have a resource — say iron ore. You find the customer who wants it. You work out how to extract it, how to deliver it, and you receive payment. For a company this is straightforward. They hold a licence to the resource, extract it, sell it, and return value to shareholders. For a nation it gets more complicated. It is the nation that actually owns the resources. What the nation controls is the licence. What it gets back is a royalty, a tax, or employment. That has been the Australian model for decades — and for decades it has worked well enough that no one examined whether it was the right model, only whether the royalty rate was set correctly. But we are now seeing the stress fractures. The gas sector is the most visible example. Australia is one of the largest exporters of liquefied natural gas in the world. And yet the return to the nation — through the Petroleum Resource Rent Tax — has been so poorly structured that some of the largest gas projects in Australian history have paid minimal tax for years. The resource is Australian. The extraction is happening in Australian waters. The wealth is being captured offshore. The primary sector business model is not just underperforming. In gas it is breaking. And the response has been to argue about the royalty rate rather than to ask whether the model itself needs to be redesigned. ### 2. Manufacturing — A Business Model in Search of a Market The manufacturing business model is straightforward in principle. You identify a customer. You work out how to make the product. You establish supply chains, manufacture the good, and sell it. At the level of an individual company, this can work. At the level of a nation, Australia has been in trouble for a generation. The reasons are structural, not incidental. They will not be fixed by policy alone, because they are not policy failures. They are business model failures. The first is supply chains. Australian manufacturing depends on inputs that travel vast distances across complex global logistics networks. That model works when the world is stable, freight is cheap, and borders are open. COVID demonstrated what happens when it isn't. But the response to that fragility has been to rebuild the same supply chains rather than redesign them. The model survived the crisis intact. The lesson was not learned. The second is scale. Australia has twenty-six million people. That is not enough to sustain competitive manufacturing at global scale across a broad range of goods. This is not a failure of ambition or policy. It is geography and demography. The business model that works for Germany — deep industrial specialisation, export-led, embedded in continental supply chains — does not transfer to a nation of Australia's size and isolation. Pretending otherwise is not a strategy. It is a wish. The third is the government response. Future Made in Australia, the National Reconstruction Fund, the critical minerals processing push — these are serious attempts to rebuild Australian manufacturing capability. But they share a common assumption: that if you subsidise enough of the cost structure, the old business model becomes viable again. It does not. Subsidised production that cannot survive without ongoing government support is not a manufacturing sector. It is a life support system for a business model that has already failed. The question that has not been asked — and that must be asked before another dollar is committed — is what a viable Australian manufacturing business model actually looks like. It is probably not volume. It is probably not commodity goods competing on price with nations whose cost structures Australia cannot match. It is more likely to be found in high value, sovereign-critical production — defence, critical minerals processing, medical technology — where proximity is a strategic advantage rather than a cost liability. Or deeply embedded in the primary sector supply chain, where Australian manufacturers serve Australian producers with the kind of specialised, responsive capability that distance makes impossible for offshore competitors. That is a different business model. It requires a different canvas. And it will not emerge from a funding program designed around the assumptions of the model it is meant to replace. ### 3. Services — The Wealth That Walks Out the Door The services business model is simple in principle. You have a skill. You find a customer who needs it. You deliver the service and receive payment. At the level of an individual company this can work extraordinarily well. At the level of a nation it presents a challenge that Australia has not yet found an answer to. The problem is mobility. Most services today are digital. And digital services are not anchored to the place where they are created. The code can be written anywhere. The platform can be hosted anywhere. The customer can be anywhere. And critically — the wealth created can move anywhere. Australia is good at building digital services companies. It is not good at keeping them. Wotif is the example that every Australian founder should study. Born in Brisbane in 2000, founded by Graeme Wood, it solved a real problem — last minute travel bookings — with elegant simplicity. It grew to dominate its market. It was genuinely world class. And in 2014 it was acquired by Expedia for approximately $703 million. Where is the wealth creation now? In Seattle. The intellectual property, the platform, the revenue stream — captured offshore. Australia got the sale price. Expedia got the business. Wotif is not an outlier. It is a pattern. Australian digital services companies are built here, scaled here, and then acquired — by American platforms, by private equity, by global players who understand that the value is in the IP and the customer base, not in the geography where it was created. The national business model for services has no structural answer to this. There is no mechanism that keeps the value onshore once the acquisition offer arrives. And in a digital economy where services are the fastest growing source of economic value, that is not a minor gap in the canvas. It is a hole through the middle of it. ### 4. Knowledge — When We Treat IP Like Wheat The knowledge economy is the most complex of the four sectors — and the one Australia understands least well. In the knowledge economy, what is being traded is not a physical good, not a service, not a resource extracted from the ground. It is intellectual property. IP can live in people — in the skills, experience, and insight that walk around in human heads. It can live in patents — legally protected expressions of novel ideas. Or it can live in the companies that hold and deploy those patents, building competitive advantage from knowledge that others cannot easily replicate. The challenge is that Australia does not treat knowledge this way. We treat it like wheat. We grow it — through universities, research institutes, and decades of public investment in science and research. We harvest it — through publications, patents, and graduating researchers. And then we sell it — or more precisely, we allow it to leave. The PhD takes their skills offshore. The patent gets licensed to a multinational. The research breakthrough gets published, cited globally, and commercialised somewhere else. We are primary producers of knowledge in an economy that rewards those who process and own it. This is not a metaphor. It is a business model. And it is the wrong one. Knowledge is not wheat. Wheat is a rival good — if I have it, you don't. Knowledge is non-rival — if I share it, we both have it, and properly managed, the value multiplies rather than divides. The business model that works for a commodity — grow, harvest, sell — actively destroys the property that makes knowledge valuable. Every time Australia exports raw knowledge and imports the processed product back as technology, software, or licensed IP, we run a terms-of-trade deficit in the most important economy of the twenty-first century. Let me make this personal. Over a thirty-six year career at CSIRO I worked on some of Australia's most significant robotics and autonomous systems research. Projects that pushed the frontier. Work that was genuinely world class. The intellectual property that emerged from that career — including Zebedee, the world's first handheld SLAM-based mobile mapping device, and Minegem, an early GPS-denied underground navigation system — is now owned by offshore companies. My business model worked. I had a career. I was paid. I contributed to science and to Australia's research reputation. But here is the question I now ask: how did the country benefit? This research was created with public funding. Australian taxpayers invested in the infrastructure, the salaries, the institutions, and the decades of accumulated capability that made it possible. The return on that investment should flow back to Australia — in sovereign capability, in industrial application, in economic value retained onshore. Instead the IP walked out the door. Legally. Legitimately. Through the normal operation of a system that was never designed to keep it. My case is not unique. It is the standard operating model of Australian research commercialisation. We fund the knowledge creation. We celebrate the research excellence. And then we watch the value get captured elsewhere, by companies with the business model sophistication to know what they are buying. The ROI on Australian public investment in knowledge is being collected in California, in Tokyo, in London. Not in Brisbane, or Melbourne, or Canberra. That is not a personal failure. It is a national business model failure. And until we design a canvas that captures the value of knowledge rather than exporting it like a raw commodity, it will continue. A couple of years ago the Federal Government released the Critical Technologies List — an attempt, I assumed, to identify and protect Australian IP in strategically important domains. I was mistaken. It was a list. Carefully researched, thoughtfully compiled, and published with appropriate fanfare. But without a business model to act on it — without a mechanism for retaining, deploying, or capturing the value of those technologies — it was wheat without a mill. We named what mattered. We built nothing around it. ### Universities — Straddling Two Business Models No institution in the Australian innovation system carries a heavier burden — or faces a more fundamental internal contradiction — than the university. The university sits across two entirely different business models simultaneously. On one side it is a services business. It provides education, issues credentials, and charges fees. The customer is the student. The product is the degree. The revenue is the tuition. This model is well understood, commercially measurable, and under enormous pressure — as the post-COVID collapse of international student revenue made brutally clear. On the other side the university is a knowledge business. It conducts research, creates new knowledge, and produces the IP and trained minds that are supposed to drive national innovation. The customer is theoretically society. The product is discovery. The revenue is grants, block funding, and indirect cost recovery. This model operates on entirely different logic — long time horizons, uncertain returns, non-rival outputs whose value is realised elsewhere, often decades later. These two models do not sit comfortably together. They compete for the same resources — the same people, the same buildings, the same institutional attention, the same prestige economy. And when they compete, the services model wins. Not because university leaders choose it cynically, but because it generates immediate, measurable revenue in an environment of chronic underfunding. Teaching loads crowd out research time. Grant writing crowds out translation work. Publication metrics crowd out commercialisation effort. The researcher who spends two years taking a discovery to market is two years behind on citations. In the current incentive structure that is a career penalty, not a reward. But there is a third problem that runs deeper than the competition between teaching and research. The knowledge model that universities operate is not really a knowledge economy model at all. It is a primary industry model in disguise. Knowledge is grown through research investment, harvested through publication and patenting, and then exported — to multinational licensees, to offshore companies with the commercial sophistication to know what they are buying, to researchers who take their skills to better-resourced institutions elsewhere. The university captures the grant. The researcher captures the citation. The IP captures a flight to San Francisco. Australia then imports the processed product back — as licensed technology, as proprietary software, as consulting fees paid to the offshore firms that commercialised what Australian public funding created. We are running a**knowledge terms of trade deficit**, and the university system is the primary point at which it originates. This is not a criticism of universities or the people within them. It is a criticism of the business model they have been asked to operate — one that was never designed to capture the value of knowledge, only to create it. A wheat farm does not become a flour mill by growing better wheat. It requires a different business model, different infrastructure, and a different understanding of where the value actually lies. Until universities are given — and held accountable to — a business model that captures knowledge value rather than simply producing it, they will continue to do what the current canvas rewards. Teach students. Publish papers. Watch the IP leave. ### Australia's Business Model — A Nation That Has Never Filled In the Canvas The transition from one sector to another is natural. It is a feature of every modern economy. Nations move from primary production to manufacturing, from manufacturing to services, from services to knowledge. This is not disruption. It is evolution. And Australia has been talking about making that transition for forty years. The Clever Country. The Knowledge Nation. The Innovation Economy. The rhetoric has been consistent, bipartisan, and sustained. What has been missing — consistently, across every government, every review, every white paper — is the business model. Without a business model a business will fail. Without a business model for the nation, a nation will fail. Not suddenly. Not dramatically. Slowly, quietly, and in ways that are easy to attribute to other causes — until the canvas is so empty that the failure can no longer be disguised. Four sectors. Four business models. Each broken in a different way, for a different reason, at a different point in the value chain. Primary industries extract and export — and the nation captures a royalty while the processing margin, the manufacturing value, and the IP flow offshore. The gas sector is showing us in real time what happens when even the royalty model stops working. Manufacturing tries to compete on cost and scale against nations whose cost structures and market sizes Australia cannot match. Government subsidises the gap rather than redesigning the model. The canvas keeps producing the same answer to a question the market stopped asking a generation ago. Services creates genuine world-class companies — and then watches the value walk out the door the moment an acquisition offer arrives from offshore. Wotif went to Seattle. The next Wotif is being built in a Brisbane garage right now, and without a different national business model, it will follow the same path. Knowledge — the sector that should be Australia's greatest competitive advantage in the twenty-first century — is treated like a primary commodity. Grown with public funding. Harvested as publications and patents. Exported as raw IP. And imported back as licensed technology at a price that captures none of the value Australia created. And sitting across the services and knowledge sectors simultaneously, universities are asked to be both a fee-for-service education business and a knowledge creation engine — with incentive structures that reward the former, punish the latter, and leave the translation gap between research and market permanently unfilled. This is Australia's business model. Not as we aspire to it. As it actually operates. No one designed it this way. It accumulated — through decades of individual decisions, each rational in isolation, none examined as a system. No one filled in the canvas. No one asked: given the resources we have, and the situation we find ourselves in, how do we extract the maximum value for Australia — not just for the company, not just for the university, not just for the researcher — but for the nation? That is the question a national Business Model Canvas forces you to ask. And it is the question Australia has never seriously answered. A healthy national canvas would look different in every box. The customer would be Australian industry and Australian society — not the research sector, not the grant committee, not the offshore acquirer. The value proposition would be sovereign capability, retained wealth, and industrial competitiveness — not publications, not royalties on raw commodities, not tuition fees from international students. The channels would be demand-side institutions with genuine power — RDC-style bodies across every major sector, a NIC with a mandate to broker between knowledge and market, procurement policy that pulls innovation rather than waiting for it to be pushed. The revenue model would capture value at the point where knowledge becomes product — not at the point where the raw knowledge leaves the building. The key partners would be universities and research institutes as suppliers, not governors. Global companies operating in Australia as co-investors, not extractors. State governments as implementation partners, not bystanders. And the key metric — the number that sits above every other number on the canvas — would not be research output, or citation count, or GDP growth. It would be value retained. Wealth that was created in Australia and stayed in Australia. That is the canvas Australia needs to fill in. Not for a company. Not for a sector. For the nation. The National Innovation Centre should be tasked with doing exactly that — writing Australia's business model, sector by sector, with the customer in the room, and presenting the canvas publicly. So that for the first time every Australian can see what we are building, who it is for, and whether it is working. > A nation that cannot describe its business model cannot change it. --- ## Postscript — A Business Model is Not Enough A business model is necessary. It is not sufficient. A nation is more than an economy. It is a complex system of interdependent models, each shaping the others, each requiring its own canvas. The business model described in this article is one layer. There are others that Australia has equally failed to articulate. A defence model — how the nation protects its sovereign interests, its people, its territory, and its critical infrastructure. AUKUS is an attempt to redesign this model. But as the absence of Defence from the SERD submissions showed, it remains disconnected from the innovation system that should be feeding it. Sovereign capability cannot be built in silos. A social model — how the nation invests in its people. Health, education, housing, welfare. The return on this investment is human capital, productivity, and social cohesion. Australia's social model is under stress on every dimension simultaneously, and the connection between social investment and economic performance is rarely made explicit in policy. An environmental model — how the nation manages its natural capital. Water, land, biodiversity, climate. Australia is drawing down on this balance sheet faster than it is replenishing it. Natural capital does not appear on the national canvas because we have no language for assets that are not priced by the market. A governance model — how decisions are made, who is accountable, and how the system corrects itself when it fails. The federation is itself a governance model — and the story of Australian innovation policy is partly a story of how poorly that model functions across Commonwealth and state boundaries. A cultural model — how the nation understands and narrates itself. Its identity, its values, its place in the world. Soft power is a real strategic asset. Australia has it and largely does not deploy it deliberately. These models are not independent. They interact, constrain, and enable each other. A business model that ignores the social model will eventually hollow out the human capital it depends on. A defence model disconnected from the innovation system will always be buying yesterday's capability at tomorrow's price. An environmental model treated as a cost rather than an asset will eventually foreclose the options the business model needs. Australia needs all of these canvases. Filled in honestly. With the right people in the room. And connected to each other in a way that reflects how the system actually works. The NIC cannot do all of this alone. But it could start the conversation. And starting the conversation — naming the models, filling in the boxes, making the assumptions visible — is the precondition for everything else. ## Postscript — The Fourth Industrial Revolution is a Business Model Problem I was involved in the introduction of the Fourth Industrial Revolution into Australia — travelling with government around the country, conducting interviews with companies large and small, writing articles. What I observed then shapes what I argue now. There has been much discussion of the Fourth Industrial Revolution — the convergence of digital, physical, and biological systems that is reshaping how goods are made, how services are delivered, and how value is created and captured. The discussion has largely focused on the technology. The robots. The artificial intelligence. The sensors, the data, the automation. These are real and consequential. But they are not the revolution. The Fourth Industrial Revolution is fundamentally a business model transformation. Cyber-physical systems — the integration of digital intelligence with physical processes — do not just make existing business models more efficient. They make entirely new business models possible, and render existing ones obsolete. Consider what connected supply chains actually do. They collapse the distance between design and manufacture. They make mass customisation economically viable. They shift value from physical production to data, software, and service. They enable manufacturers to sell outcomes rather than products — a jet engine sold by the hour rather than the unit, a mining truck sold by tonnes moved rather than the machine itself. These are not technology changes. They are business model changes, made possible by technology. Australia has largely missed this distinction. The national conversation about Industry 4.0 has been dominated by the technology layer — which robots to buy, which AI platforms to adopt, which digital skills to train. The business model layer — what new value propositions become possible, which customers can now be served in ways that were previously uneconomic, how value capture shifts along the chain — has received almost no serious policy attention. The result is predictable. Australian firms adopt the technology. The business model transformation happens offshore, in the companies that designed the technology and understand what it makes possible. Australia buys the robot. Someone else captures the value the robot creates. This is the wheat farming model applied to industrial transformation. We adopt the tool. We do not redesign the canvas. The Fourth Industrial Revolution will not wait for Australia to catch up. The window for business model transformation — the period in which new models can be built before incumbent models harden into new orthodoxies — is open now. The NIC, if it is serious about its mandate, should be asking not what technologies Australia should adopt, but what business models those technologies make possible — and how Australia can own them. > Only the correct business model that will drive translation.