Most inventions that fail commercially do not fail in the laboratory. They fail between a working prototype and a paying customer. Organizations routinely underestimate that gap because they treat commercialization as a single activity rather than as a choice among distinct routes.

Consider EMI's CT scanner. Godfrey Hounsfield's invention earned a Nobel Prize and transformed medical diagnosis. EMI, a company better known for music and electronics, chose to manufacture and sell the scanners itself. Within a decade, GE and Siemens had captured the market, and EMI had exited the business. Those firms already had hospital sales forces, service networks, and regulatory experience. The technology succeeded. The pathway did not.

The lesson is not that EMI should have stayed out of the market. It is that how a technology reaches customers is a strategic decision, with consequences as large as the invention itself. This article maps the main commercialization pathways, the criteria for choosing among them, and the capabilities each one demands. The goal is to make pathway selection deliberate rather than accidental.

Commercialization pathways sit along a spectrum defined by how much of the value chain the innovator controls. At one end is direct commercialization: the organization builds, markets, sells, and supports the product itself. It keeps the full margin and the customer relationship. It also carries every risk, from manufacturing scale-up to distribution.

In the middle are partnership arrangements: joint ventures, co-development agreements, OEM supply relationships, and strategic alliances. The innovator contributes the technology. A partner contributes complementary assets such as manufacturing capacity, regulatory expertise, or access to sales channels. Value and risk are shared, and so is control.

Licensing sits at the far end. The innovator transfers the right to use the technology in exchange for fees and royalties, earning value without building the downstream organization. Pilkington licensed its float glass process to manufacturers worldwide from the 1960s onward. The returns far exceeded what Pilkington could have earned by producing all that glass itself. Qualcomm made a similar move in 1999, selling its handset and infrastructure businesses to concentrate on chips and patent licensing.

Spin-offs are a distinct option. The organization creates a new entity to commercialize a technology that does not fit its own strategy or structure. ARM emerged from Acorn Computers in 1990 as a joint venture with Apple and VLSI Technology, built around a licensing model Acorn itself would never have pursued. A spin-off gives a technology its own capital, management attention, and culture, at the cost of the parent's direct control. These options are not mutually exclusive. Sophisticated organizations often run several pathways at once across different markets or applications of the same technology.

Takeaway

Commercialization is a choice about how much of the value chain to own. Every pathway trades control for capital efficiency, and the right trade depends on what you already have.

Pathway selection starts with a question that is too often skipped: who controls the assets required to turn this technology into revenue? David Teece's 1986 analysis of why innovators so often lose to imitators remains the most useful framework here. It rests on two variables: the appropriability regime around the technology, and who owns the complementary assets.

Technology characteristics determine appropriability. Strong patents and well-documented knowledge favor licensing, because the innovator can transfer the technology while still enforcing its rights. Weak protection, or know-how that resists documentation, pushes toward direct commercialization or close partnership. In those cases the value lives in people and processes rather than in a contract. Breadth matters too. A platform technology with applications across many industries rarely suits a single in-house product line.

Organizational capabilities determine whether direct entry is credible. Suppose the firm already owns the manufacturing, distribution, and service infrastructure the target market requires. Then direct commercialization captures margin efficiently. If incumbents own those assets, entering alone means competing against the organizations that control the bottleneck. Hospital sales and service belonged to GE and Siemens, which is exactly where EMI struggled. In such cases, partnering with or licensing to the asset holders is usually the stronger position.

Market conditions shape timing and leverage. In young markets without a dominant design, the first innovator to reach scale may set the standard. That can justify the risk of direct entry, or aggressive licensing to speed adoption. In mature markets dominated by established players, bargaining power shifts to whoever controls access to customers. The best pathway is the one where the innovator's strongest assets meet the market's tightest constraint.

Takeaway

The question is not whether your technology is valuable, but who holds the assets needed to deliver that value. Profit tends to flow to whoever controls the bottleneck.

Each pathway requires a different organization. Direct commercialization calls for capabilities that research organizations rarely have: manufacturing scale-up, quality systems, sales channels, and customer support. It also requires the working capital to fund inventory and receivables before revenue arrives. Underestimating this transition is the most common failure. The capabilities that produce an invention look very little like those needed to sell it a million times.

Licensing looks lighter, but it requires its own discipline. A licensing business depends on four things: a carefully built patent portfolio, the legal capacity to enforce it, technical teams to help licensees implement the technology, and deal-making skill to structure royalties that reward both sides. Dolby built a lasting business on exactly this combination of rigorous intellectual property, demanding certification standards, and a brand licensees wanted on their products. Without the ability to enforce it, a license is merely a suggestion.

Partnerships require governance capability. The organization must define each party's contributions clearly and share enough to collaborate without giving away core intellectual property. It must also manage incentives as circumstances change and the partners' interests drift apart. Many alliances fail because of unresolved questions about who owns improvements and who controls the customer, rather than because of the technology.

Spin-offs require entrepreneurial management, independent capital, and a clean separation of intellectual property and people from the parent. A half-hearted spin-off that still depends on corporate approval processes inherits the parent's constraints without gaining its resources. Across all four pathways, the principle is the same. Each must be resourced as a distinct capability rather than treated as an afterthought once research is complete. Organizations that build commercialization skills deliberately, before the breakthrough arrives, move faster when it does.

Takeaway

Choosing a pathway is a commitment to build a specific kind of organization. If you are unwilling to build the capabilities a pathway requires, you have not actually chosen it.

Technology commercialization is a strategic design problem made up of three linked decisions: which pathway, chosen on what criteria, supported by which capabilities. Treating any of them as an afterthought hands value to someone else.

A practical discipline follows. First, assess appropriability: how well can the technology be protected and transferred? Second, map complementary assets: who owns the manufacturing, channels, and customer relationships the market requires? Third, honestly audit your internal capabilities against what each candidate pathway demands. Finally, choose the route where your advantages align with the market's constraints, and resource it fully.

Xerox PARC produced the laser printer, Ethernet, and the graphical user interface. Xerox profited enormously from the first and watched others build industries on the rest. Invention creates possibility. Pathway choice determines who collects.