Rethinking Make-or-Buy: Strategic Manufacturing Decisions in Pharma
Lakshmi, Editorial Team, Pharma Focus Europe
For three decades, pharmaceutical manufacturers treated outsourcing as the default answer to capacity, cost and speed pressures. Supply disruptions, modality complexity and regulatory scrutiny have since exposed the limits of that reflex. This article reframes the make-or-buy question as a capability decision rather than a cost decision, offering senior leaders a structured way to judge where external partnership creates value and where internal control has become non-negotiable.
Introduction
Few strategic decisions in pharmaceutical operations carry as much long-term consequence as the choice of what to make and what to buy, and few are revisited as rarely. Network footprints agreed a decade ago, partners selected during a capacity crunch, technology transfers completed under launch pressure — these choices harden into architecture. They stop looking like decisions and start looking like the way things are done.
That inertia has become costly. The environment that made aggressive outsourcing self-evidently correct through the 1990s and 2000s has changed in almost every material respect. Capital is more expensive. Portfolios have shifted toward biologics, conjugates and advanced therapies whose processes are far less commoditised than a solid oral dose. Regulators across Europe now expect demonstrable oversight at every node of the supply chain, not only at the sponsor's own sites, and the Qualified Person carries personal accountability that cannot be contracted away. European policy has begun treating manufacturing location as a question of health security rather than commercial preference, with critical medicines availability and reshoring of essential production now explicit policy objectives. Pandemic-era disruption, shortages of critical starting materials and single-source dependencies have each, in turn, exposed how thin the margin for error had become.
None of this signals a wholesale return to vertical integration. Outsourcing remains the right answer for a large share of pharmaceutical activity, and the economics that justified it have not disappeared. What has disappeared is the licence to treat it as a default. The expectation now falling on senior leadership is more demanding: an ability to explain, asset by asset and activity by activity, why a given capability sits inside or outside the enterprise — and to defend that reasoning to a board, an inspector and a payer with equal confidence.
This article sets out why the traditional cost-led approach to make-or-buy analysis has become unreliable, what the full ledger of outsourcing actually contains, where internal ownership has genuinely earned its place, and how leaders can apply a consistent capability test rather than an institutional reflex.
The Reflex That Outlived Its Logic
The original argument for outsourcing was sound and remains partly valid. Fixed manufacturing assets are expensive, slow to build, and punishing when underutilised. Development attrition means most pipeline assets never reach commercial volume, so building dedicated capacity ahead of proof is a bet with poor odds. External partners aggregate demand across many sponsors, achieve higher utilisation, and convert a sponsor's fixed cost into a variable one.
Layered onto this was a strategic narrative about core competence. Innovation, clinical development, regulatory strategy and commercialisation were framed as the value-creating core; molecules, fill-finish, packaging and much of analytical testing were framed as executable by others. Balance sheets rewarded asset-light structures. Return on invested capital improved when the invested capital shrank.
For standardised small-molecule products in stable geographies, this reasoning held up well and continues to. The difficulty is that it was generalised far beyond the conditions that justified it. Activities were externalised not because rigorous analysis showed external execution to be superior, but because outsourcing had become the institutional reflex — and because the costs it generated were diffuse, delayed, and largely invisible to the models used to approve it.
The Ledger That Rarely Appears in the Business Case
Conventional make-or-buy analysis compares an internal cost per unit against an external price per unit, adjusts for capital avoidance, and stops. The comparison is incomplete in ways that consistently favour buying.
Coordination and oversight cost. Every external node requires quality agreements, person-in-plant coverage, audit programmes, deviation management across organisational boundaries, and change control that must be negotiated rather than directed. These costs are real, are typically carried in overhead rather than cost of goods, and scale with network complexity rather than with volume.
Response latency. When a process excursion occurs internally, the sponsor controls the investigation, the resourcing, and the priority. When it occurs at a partner site serving multiple sponsors, the sponsor controls none of these directly. The commercial cost of a four-week resolution versus a four-day one rarely appears in any comparative model, yet it can exceed years of unit-cost savings in a single event.
Knowledge erosion. Process understanding is not fully transferable through documentation. Tacit knowledge — why a parameter was set where it was, which deviations proved benign, how the process behaves at the edges — accumulates in the people who run the process. Sustained outsourcing of a platform gradually removes an organisation's ability to evaluate its own technology, negotiate credibly, or bring the activity back if circumstances demand.
Optionality forfeited. A partner's capacity is contracted, not owned. In a demand surge, an accelerated approval, or a competitor's supply failure that creates unexpected volume, the sponsor competes for allocation against other clients. Speed to respond becomes a function of contractual position rather than operational capability.
None of this argues that outsourcing is a mistake. It argues that the price quoted by an external provider is not the cost of using one, and that decisions made on price alone have been systematically biased for years.
Where Insourcing Has Genuinely Earned Its Place
The strongest case for internal execution is no longer cost. It is control over the things that determine competitive position.
Platform technologies. Where a manufacturing platform is used across multiple assets — a conjugation chemistry, a viral vector system, a lipid nanoparticle process, a continuous manufacturing line — internal ownership converts a per-product cost into an enterprise capability. Learning compounds across the portfolio rather than accruing to a supplier who also serves competitors.
Processes still in motion. Outsourcing works best when a process is locked. When the process itself is the subject of active development — as it is throughout advanced therapy manufacturing — the iteration loop between development and production becomes the rate-limiting factor. Every cycle that crosses a company boundary is slower, more expensive, and lower in information content.
Products where supply continuity is the value proposition. For assets with no therapeutic alternative, or where a stock-out carries regulatory and reputational consequences disproportionate to revenue, redundancy and control justify economics that would otherwise be unattractive. In Europe, where shortage notification obligations and national stockholding requirements have tightened considerably, this calculation has shifted further toward control.
Nodes with structural fragility. Concentration of a critical starting material, intermediate, or specialised finishing capability in a single geography or a single provider is a strategic exposure regardless of how well that provider performs. Insourcing here is not an efficiency decision; it is an insurance decision, and should be evaluated as one.
A Framework Worth Applying Before the Next Decision
Rather than a binary corporate posture, senior leaders benefit from a consistent test applied at the level of individual activities.
Is this activity a source of differentiation, or a source of parity? Activities that determine product performance, cost position or speed advantage warrant internal ownership. Activities where excellent execution is widely available and confers no advantage are candidates for partnership.
How stable is the underlying process? Mature, locked, well-characterised processes transfer cleanly. Processes under active development do not.
What is the cost of a failure, measured properly? Not the cost of remediation — the cost of the delay, the lost supply, the regulatory attention, and the erosion of prescriber confidence.
Can we still evaluate this if we stop doing it? An organisation must retain enough technical depth to specify requirements, assess partner capability and interpret data intelligently. Where outsourcing would eliminate that depth, the decision has consequences well beyond the activity itself.
What does this do to our optionality? Does the decision preserve the ability to change course in three years, or does it foreclose it?
The Answer Is Usually Neither, and Both
The most capable operating models emerging across the sector are not internal or external. They are deliberately hybrid, structured so that each mode compensates for the other's weakness.
Common patterns include internal ownership of the first commercial site with external capacity for volume beyond a defined threshold; internal development-scale manufacturing with external commercial supply once the process is locked; dual-source strategies where an internal site and an external partner both hold approved status for the same product; and retained internal analytical capability even where manufacturing is fully outsourced, preserving the ability to interrogate a partner's data independently.
These structures cost more than the cheapest available configuration. That is the point. They purchase response capability, negotiating position and regulatory resilience — assets that do not appear on a balance sheet but reliably determine which organisations absorb disruption and which are defined by it.
Governance: Treating This as a Decision, Not a Setting
The practical failure in most organisations is not analytical. It is procedural. Make-or-buy decisions are taken at the point of maximum pressure — a capacity shortfall, a launch timeline, a cost programme — by whichever function owns the immediate problem, and are then never formally revisited.
Three governance disciplines change this materially. First, a periodic portfolio-level review of the network, conducted on a defined cycle rather than in response to crisis, asking explicitly which activities would be placed differently if decided today. Second, a standing capability map identifying which internal competencies are being maintained, which are eroding and which have already been lost — reviewed with the same seriousness as pipeline attrition. Third, decision records that capture the assumptions behind each significant placement, so that when those assumptions change, the trigger to reconsider becomes visible rather than dependent on someone remembering.
Conclusion
The make-or-buy question has been asked in pharmaceutical operations for decades and answered, for most of that period, by default rather than by analysis. Outsourcing became the reflex because the conditions of an earlier era rewarded it and because the costs it generated were slow to surface. Those conditions have changed. Modality complexity, supply concentration, regulatory expectation and the strategic value of response speed have all shifted the balance — not uniformly toward insourcing, but decisively away from any single default.
What the current environment demands of senior leaders is not a new preference but a defensible method: a clear view of which capabilities differentiate the enterprise, an honest accounting of what external execution actually costs, and the governance discipline to revisit these placements before circumstances force the question. Organisations that can explain, activity by activity, why each sits where it does will meet the next disruption from a position of choice. Those that cannot will discover their architecture only when it fails.
Lakshmi is a science writer with a foundation in the laboratory. She earned her master's in biotechnology and trained through research internships at ICGEB (JNU) and DIPAS, DRDO, with her work appearing in the Egyptian Journal of Veterinary Sciences. Now APCRM-certified and part of the editorial team at Pharma Focus America and Pharma Focus Europe, she reports on pharmaceutical technology, research, and innovation — giving complex science a clear and confident voice for industry leaders.
