Why Drug Development Is a Systems Problem, Not a Science Problem

When a drug fails, the instinct is to blame the science. Sometimes that is right, but more often the failure is systemic: a breakdown in how the many parts of development fit together. Drug development is a systems problem, not only a science problem, and treating it as the latter is itself a cause of failure. This essay makes the case. It is educational and is not investment advice.

The science is one component among many

Bringing a therapy to patients requires far more than a working molecule. It requires a regulatory strategy, a manufacturing process, a clinical operation, a reimbursement plan, a commercial organization, and the capital to sustain all of them over years. Each is a complex discipline, and the whole succeeds only if they are coordinated. A brilliant molecule embedded in a poorly integrated system fails just as surely as weak science, which is why development must be understood as a system rather than a sequence of independent steps.

Failures cascade across the system

The systemic nature of development shows up in how failures propagate. A manufacturing problem can derail a clinical trial, a poorly chosen endpoint can waste years of work, and a capital shortfall can force a company to cut the very activities that would have saved it. The base rates capture the aggregate result: only a low double-digit percentage of drugs entering testing reach approval (Wong, Siah, and Lo, 2019; Hay et al., 2014). Many of those failures are not failures of the underlying science but of the system around it, where one weak link breaks the chain.

The cost reflects the system, not just the molecule

The enormous capitalized cost of development, estimated at roughly 2.6 billion dollars per approved drug in 2013 terms, reflects the cost of running this entire system, including the many failures along the way (DiMasi, Grabowski, and Hansen, 2016). It is tempting to think of that figure as the cost of the science, but most of it is the cost of trials, manufacturing, regulatory work, and the attrition of the system as a whole. Reducing the cost of drugs, a common policy goal, therefore requires improving the system, not just the science.

Why a systems view changes decisions

Seeing development as a system changes how companies should be built and judged. It means resourcing regulatory, manufacturing, and commercial capabilities early rather than treating them as downstream of the science, a theme running through how to commercialize a biotech innovation and the analysis of the manufacturing moat. It means evaluating a company by the strength of its weakest link, not its strongest. And it means recognizing that the common mistakes that kill startups, catalogued in common FDA mistakes, are usually systemic failures rather than scientific ones.

Why the science gets the blame anyway

If most failures are systemic, why is the science usually blamed? Because the science is visible and definitive, a failed trial produces a clear result, while the systemic factors that led there are diffuse and harder to attribute. It is easier to say a drug did not work than to trace how a chain of decisions about endpoints, manufacturing, and capital set it up to fail. This misattribution matters, because it leads companies and investors to over-invest in science and under-invest in the system that determines whether science succeeds.

The pattern Most drug development failures involve systemic breakdowns, not only scientific ones. This is consistent with the base rates and the cost structure.

What a systems approach requires

Treating development as a system means integrating regulatory, manufacturing, clinical, commercial, and financial planning from the start, so that decisions in one area account for their effects on the others. It means building teams with the full range of expertise, not just scientific talent, and it means a kind of operational discipline that is less celebrated than scientific brilliance but more decisive. This is the discipline that distinguishes companies that reach the market from those that stall, a distinction connected to what makes a company investable.

The takeaway

Drug development is a systems problem, and the companies and investors who treat it that way, integrating the disciplines and judging by the weakest link, navigate it best. Treating it as a pure science problem, and blaming the science when it fails, misses where the real leverage and the real risk lie. For the investor's version of this insight, see the hidden risk most investors miss, and for how this is applied in practice, the advisory practice.

What a systems failure looks like in practice

It helps to make the abstract concrete. Picture a company with genuinely strong science that chose an endpoint without confirming it would satisfy regulators, designed its early process in a way that could not scale, and raised only enough capital to reach its first major readout. Each decision seemed reasonable in isolation and within its own discipline. Together they form a trap. The trial reads out positively on the chosen endpoint, but regulators signal it is insufficient; addressing this requires a new trial, which requires more capital; raising that capital requires demonstrating manufacturing readiness, which the company never built; and the financing market has tightened in the meantime. The science worked, and the company still fails, because the system around the science was never integrated. This pattern, in countless variations, is how most programs actually die, and it is invisible if one looks only at the final scientific result. Recognizing it requires viewing the company as a system from the outset and asking how the pieces interact, not just whether each is individually sound. That systems lens is precisely what separates durable companies from fragile ones, and it underlies the operational discipline described in how to commercialize a biotech innovation.