Managing the Go/No-Go Decision: Stage-Gate Project Management in Drug Development

Introduction

Drug development is one of the most capital-intensive, risk-laden undertakings in any industry. Deloitte's 2024 analysis estimated an average cost of $2.23 billion per approved drug (link), and the attrition statistics behind that number are sobering (link). The average likelihood of approval for a new Phase I drug currently sits at just 6.7% — an all-time low — with Phase II remaining the most treacherous hurdle, where only 28% of programs successfully advance (link). Against this backdrop, the question of when to continue and when to stop is not a soft management judgment call. It is a strategic survival mechanism, and the way an organization structures its answer will define the health of its entire pipeline.

Stage-gate project management is the framework built to answer it. Used across industries since Robert Cooper formalized the concept in the 1980s, stage-gate thinking is now foundational to drug development programs at organizations of all sizes. Yet the rigor of its execution varies enormously — and that gap is where pipelines fail. This post breaks down what stage-gate really means in a biotech and pharma context, what a well-governed gate review looks like, and how project managers can own this process with the discipline it demands.

Go/No-Go decision checkpoint in drug development

What Stage-Gate Means — Through a PM Lens

Stage-gate is a structured lifecycle model that divides a project into stages — defined periods of work — separated by gates, which are formal decision points. At each gate, the team evaluates whether predefined criteria have been met before committing resources to the next stage. The simplicity of that description belies how difficult it is to execute well in a regulated, scientifically uncertain environment.

The Project Management Body of Knowledge (PMBOK) Guide frames this concept directly within its governance and project lifecycle language (link). A phase gate is a review point used to assess whether exit criteria for the current phase have been satisfied before the project proceeds. Critically, PMBOK identifies the full range of legitimate gate outcomes: proceed, proceed with modification, hold the project temporarily, repeat the phase, or terminate. In drug development, that last option — termination — is not a failure of the system. It is the system functioning exactly as intended.

The reason stage-gate matters so acutely in this industry is the asymmetry of late-stage failure. A program terminated at target validation costs a fraction of what it costs to fail in Phase III. The governance performance domain in PMBOK establishes that effective projects must have decision criteria aligned to value creation and signaling mechanisms that give leadership meaningful information at the right time. Stage-gate is the most concrete operational expression of that principle in life sciences.

Putting It Into Practice: How Ganvion Structures a Small Molecule Program

To understand what rigorous stage-gate execution looks like in practice, consider how Ganvion Biotech Solutions approaches a small molecule drug development program. When working with an early-stage biotech client, one of the first deliverables Ganvion establishes is a program architecture that maps every stage, every decision point, and every criterion that will govern advancement — before a single experiment is run. It is important to note here, that this architecture is not a rigid prescription. Established change control processes and iterative decision cycles are embedded throughout, allowing the science to direct the program at each stage rather than forcing data to fit a predetermined path.

The program is organized around six stages spanning the full preclinical and early clinical arc: Target Selection, Discovery of Active Compounds, Lead Selection and Optimization, Nominating Development Candidate(s), Compiling an IND Application, and Clinical Trials. Each stage is defined by its primary deliverable. Within each stage, individual decision points function as formal work packages, each carrying its own scope, resource requirements, and measurable criteria. Those criteria represent the specific activities and data outputs the team must produce before a gate decision can be made.

This three-level structure — stage as deliverable, decision as work package, criterion as activity — maps cleanly onto PMBOK project lifecycle principles and gives the program a four-digit organizational logic. Each activity carries a program number, a stage number, a decision number, and an activity number, creating full traceability from any individual data point back to the business case. When a client asks why a specific toxicology assay is on the plan, the answer is traceable to a gate criterion, which is traceable to a development candidate decision, which is traceable to the program's IND strategy.

Without this structure, advancement decisions get made informally, criteria get invented retroactively, and the program drifts toward the sunk-cost trap. A well-designed stage-gate architecture means a termination decision at Stage 2 can save tens of millions of dollars that would otherwise be committed to preclinical toxicology packages, manufacturing development, and IND-enabling studies for a compound that was never truly viable. The goal is not to make terminating programs easy — it is to make the decision defensible, documented, and timely.

What Goes Into a Gate Review Package

A gate review is not a status update. It is a formal decision document, and the project manager is responsible for assembling and owning it. The portfolio committee makes the decision — but only if the PM has given them an accurate, complete, and honestly framed package to evaluate.

A well-constructed gate package integrates several dimensions. The scientific summary presents data against predefined exit criteria — not a curated highlight reel, but an objective assessment including data that raises questions. The regulatory assessment addresses alignment with FDA or EMA expectations, IP position, and target product profile status. The commercial analysis frames the program's thesis against market potential and competitive landscape. The financial update compares actual spend against plan and includes an updated risk-adjusted NPV or probability of technical success assessment. A current risk register review rounds out the package, surfacing open and newly identified risks alongside mitigation status.

Critically, the package should close with an explicit go/no-go recommendation and rationale. Presenting data without a stated position is a governance failure — it removes the PM's accountability for owning the program's trajectory. In the Ganvion framework, because criteria are pre-specified before the stage begins, the gate question is never "did we do enough?" — it is "did we meet the criteria we agreed on?" That distinction matters enormously in practice.

The PM's Role vs. the Portfolio Committee

One of the most common governance failures in small biotechs is conflating the PM role with decision-making authority at gates. These must be structurally separated. The PM owns the completeness and accuracy of the gate package and surfaces data gaps proactively. The portfolio committee evaluates that package against strategic and financial context, makes the go/no-go decision, and documents it with rationale. This body should include scientific, regulatory, commercial, and financial representation — not because every decision requires consensus, but because a narrowly constituted committee will predictably miss dimensions that matter.

The PMBOK guide is clear that governance structures should be scaled to organizational context. A small virtual biotech needs a formal gate process just as much as a large pharma — it just looks different. But regardless of size, the PM and the decision-maker should not be the same individual. When they are, objectivity is compromised and the gate becomes theater rather than governance.

Termination Decisions Are Good Project Management

The industry's tendency to advance weak assets is expensive. Estimates suggest roughly 60% of all R&D costs are attributable to attrition (link), and the overall likelihood of approval for a Phase I drug has fallen to 6.7% — down from 10% a decade ago — driven in part by failures that could have been caught earlier (link). A gate that produces a no-go is not a failure. It is the system working correctly.

Making those decisions well requires three disciplines. Criteria must be defined before the stage begins, not invented at the gate. Asset emotion must be separated from portfolio logic — scientific champions often conflate the importance of the underlying science with the viability of the specific program, and the PM's job is to hold the criteria accountable. And the decision must be documented; a gate decision with no written record is indistinguishable from a decision that was never made. Capturing the institutional learning from a termination is also what makes the next gate more effective — every terminated program should generate a lessons-learned review, not just a closeout.

Connecting Stage-Gate to Benefits Realization

Every gate decision is also a benefits realization checkpoint. The program's Benefits Realization Plan should be updated at each gate to reflect current probability of success, revised timelines, and commercial value projections — because a gate package that advances a program without updating the risk-adjusted value estimate may be greenlighting an asset whose benefit profile no longer justifies the resource commitment. Gate criteria should trace directly to the benefits defined in the original business case, and a hold decision should explicitly trigger a review of whether the program's strategic rationale still stands.

This is the connection that separates stage-gate as a process tool from stage-gate as a governance discipline. The process runs the gates. The discipline ensures those gates are connected to something that matters.

Conclusion

Managing stage-gate in drug development is one of the highest-leverage activities a project manager can perform. Done well, it protects capital and ensures pipeline decisions are made on evidence — not momentum. Done poorly, it produces expensive late-stage surprises and portfolios full of assets that should have been rationalized years earlier.

At Ganvion Biotech Solutions, our approach to stage-gate is built into the structural DNA of how we manage development programs — from target identification through clinical proof of concept. Whether you are building your first formal gate process or refining an existing one, we can help you design decision frameworks that are rigorous, actionable, and scaled to your organization.