Pharmaceutical Manufacturing’s Three Crises Are Really One Problem

James Glover
James Glover
Pharmaceutical Manufacturing’s Three Crises Are Really One Problem

The FDA issued 105 drug-quality warning letters in FY2024, the highest count in five years. The workforce is thinning at the same time: McKinsey reports that 80% of pharma-manufacturing companies face a skills mismatch, and industry projections have long warned that 60% of needed positions could go unfilled. And the financial exposure is real: when a single facility goes down for quality reasons, the losses run into the hundreds of millions, as Johnson & Johnson learned when a plant shutdown cost roughly $600 million in lost sales.

But these aren’t three problems. They’re three symptoms of one: getting people to consistently do what they already know they should do.

The Knowledge-Behavior Gap

Pharmaceutical companies have solved the knowledge-transfer problem. Operators complete exhaustive training on aseptic technique. They understand contamination risk. They know gowning procedures inside and out, and they pass the tests. Yet warning letters keep citing the same violations. The knowledge is there. The behavior isn’t.

So why does the industry keep retraining people who already know the rules?

Because when it comes to CAPA, the corrective action (retraining) is an easier problem to solve than preventative action. Corrective retraining helps prove a response, and it is important. But it treats compliance as an information problem rather than a behavior problem. Consider the data. On average, almost 30% of quality defects are attributed to human error, yet that same analysis finds retraining addresses only about 10% of those errors, because it’s focused on knowledge. Looking at that numbers from the other direction, retraining will fail to solve 90% of human errors, because the errors are behavior-based, not knowledge-based. The operator already knew the procedure and slipped anyway, under time pressure, distraction, etc. Preventative action needs to focus on behaviors and habits. If compliance hasn’t become habitual, it becomes optional the moment pressure mounts.

The warning signs show up across the industry:

  • Operators repeating the same mistakes despite multiple retrainings
  • Most companies naming human error as their leading root cause for quality incidents
  • A persistent “mistakes are inevitable” mindset on the floor

Proactively addressing the behavior gaps can help manufacturers demonstrate preventative action.

Why the Crisis Is Accelerating

Regulatory pressure is intensifying. The FDA is expanding its focus, with data integrity now a major enforcement priority and quality-system issues making up a large share of citations. The bar is getting higher.

The workforce shortage compounds it. Many new hires arrive with no pharmaceutical background at all, and getting a new operator productive on the floor takes months.

The financial stakes keep mounting. The J&J shutdown was not an outlier; it was a warning. These are not abstract quality metrics. They are balance-sheet impacts that executives have to explain to boards and shareholders.

Together they create an impossible triangle: you need operators who follow protocol automatically (the regulatory requirement), you need them productive faster (the competitive requirement), and you cannot afford the recalls that follow when either fails (the financial requirement). The traditional training model cannot deliver all three at once.

What Makes Behavior Change Different

Organizations that have solved this aren’t doing more classroom training. Since behavior develops through repetition in context, they’re embedding intentional practice into daily operations: a gowning verification built into the shift start, a contamination check folded into a real production run, systematic practice of the exact behaviors that tend to slip when schedules compress. None of it interrupts the work. It is the work, done deliberately enough to stick.

Research on habit formation shows it takes roughly 10 weeks of consistent practice to make a behavior automatic. That’s longer than most training programs run, but it’s still faster than the proficiency timelines the industry currently accepts, and it usually produces behavior that holds when pressure mounts.

This mirrors what manufacturers are learning about turning SOPs into habits: systematic behavior change through on-the-job practice addresses the root cause that retraining never touches.

Building Compliance Into Operations

The advantage goes to organizations that make compliance automatic rather than aspirational. That means making SOP adherence the path of least resistance, building quality verification into the normal workflow, and engineering the work so that doing it right is easier than cutting corners. It also means measuring different things:

  • Not training-completion rates, but behavioral consistency during actual production runs
  • Not the hope that protocols will be followed, but tracked adherence over time
  • Not human error as individual failure, but as a signal of organizational design

Organizations making this shift connect behavior change to the outcomes boards care about: deviation rates, audit findings, and capacity. They can draw a direct line from systematic behavior development to the KPIs the business already reports on. And the validation doesn’t rest on a single before-and-after snapshot. It comes from multiple points across a program: the operator self-assessing readiness, the supervisor confirming the behavior is actually showing up on the floor, and documented real-world examples. That matters most in a compliance environment, where you need evidence the behavior is happening, not just a record that someone sat through the training.

The Competitive Imperative

Regulatory pressure, workforce shortage, and quality culture share one solution: making compliance a habit. Organizations that solve it will gain advantages that compound, since fewer warning letters, faster proficiency, and durable behavior change reinforce one another. The work is no longer teaching operators what they already know. It is building the behavior that holds when production pressure arrives.

About the Author

James Glover is CEO of Flint Learning Solutions, where he helps organizations change behavior through AI-personalized, on-the-job activities that integrate into daily work. He has spent more than a decade using machine learning to predict and change behavior across industries including financial security, marketing technology, and manufacturing.