Case Study: How We Uncovered a Hidden Capacity Crisis at a Top LATAM Pharma

In late 2025, a leading pharmaceutical manufacturer in Latin America came to us with a familiar problem: they knew something was wrong with their capacity, but they couldn't see exactly what or how bad.

Their production teams were working overtime. Orders were being expedited. The ops team felt overwhelmed. But when asked to quantify the problem — which resources, by how much, over what timeline — they couldn't answer with confidence.

Six weeks later, we delivered a report that changed how they think about their operations. Here's the story.

The Client

Our client is a top-5 pharmaceutical manufacturer in their country with three manufacturing sites, over 250 active products, and decades of market leadership.

They operate across multiple therapeutic areas: cardiovascular, respiratory, pain management, dermatology, and more. Their product portfolio spans tablets, capsules, liquids, and topical formulations.

Like many pharmaceutical manufacturers in emerging markets, they had grown organically over decades. Equipment had been added as needed. Production had been allocated across sites based on historical decisions. The result was a complex, interconnected operation that no single person fully understood.

The Challenge

Leadership had three questions:

1. Where are we today? Which resources are bottlenecked? How bad is it?

2. What can we do without capital? Can operational changes resolve the constraints?

3. Where should we invest? If capital is needed, where does it have the highest impact?

They had tried answering these questions internally with spreadsheets. But every time they thought they understood the situation, reality surprised them.

Building the Digital Twin

We collected data on 35 active resources across three sites, 250+ products with defined production routes, demand forecasts for six years, and OEE parameters by product-resource combination.

By week two, we had a complete digital twin of their three-site operation: every resource, every product, every route, every constraint.

The Baseline Shock

When we ran the baseline simulation, the results were worse than anyone expected.

The worst bottleneck — granulation at one site — was running at 226% utilization. That means the plant needed that resource for 2.26x more hours than physically available. Across both main sites, seven resources exceeded 100% capacity: granulation (226%), dry mixing (153%), blistering (125%), coating (169%), granulation at the second site (163%), cartoning (139%), and blistering (133%).

What does 226% mean in practice? It means that even if the granulator ran 24/7 with zero downtime, they'd still only make 44% of what they needed. The result: constant expediting, missed targets, and an ops team that felt like they were always behind — because mathematically, they could never catch up.

The ops team knew things were bad. They didn't know they were this bad.

Why the Spreadsheets Missed It

The planning team had calculated utilizations before. Why did their numbers look so different?

1. They calculated averages, not peaks. Annual averages looked manageable; monthly peaks didn't.

2. They didn't account for OEE variability. Theoretical capacity vs. actual capacity differed by 30-40%.

3. They couldn't see the cascade. Upstream bottlenecks starved downstream equipment, but downstream utilization still looked fine on paper.

Finding Operational Levers

We started with operational changes that don't require capital investment.

Lever 1: Adding a Third Shift

Four critical resources were running two shifts. Adding a third shift increased available hours by 50%.

Lever 2: Operational Excellence Program

A 20% improvement in OEE through maintenance programs, changeover optimization, and scheduling improvements.

Lever 3: Production Rebalancing

Transferring products between sites to balance utilization across the network.

Combined result: The three operational levers reduced the worst bottleneck from 226% to 111% at one site, and brought the other site from 169% to 94%. Seven over-capacity resources became four — still challenging, but manageable without capital investment.

The Long-Term Challenge

Here's what the simulation revealed that spreadsheets never could: even with full operational intervention, demand growth of ~5% annually would erode the gains within two years. By 2028, resources would start exceeding 100% again. By 2032, the network would be back to crisis levels.

Operational levers buy time. They don't solve the structural problem.

The Investment Roadmap

We modeled two capital investment scenarios:

Scenario A: Three Plants — Maintain all three sites, expand one plant with modern equipment, renovate critical equipment at the main site. Investment: $4-8M.

Scenario B: Mega-Plant — Consolidate into two sites, close the main site, build out one facility as a modern mega-plant. Investment: $4-7M.

Counterintuitively, the consolidation scenario was slightly cheaper because it eliminated the need to renovate aging equipment at the main site.

Both scenarios achieved the same goal: zero resources over 85% utilization through 2032, with room for additional growth.

The Outcome

The client now has:

• Clarity on exactly which resources are constraining production
• A playbook of operational levers to implement immediately
• A roadmap for capital investment tied to demand triggers
• Confidence intervals on all projections, not just point estimates

Most importantly, they stopped guessing. Decisions that used to be based on intuition are now based on simulation.

This is what capacity planning should look like: data-driven, scenario-tested, and actionable. If your plant feels like it's running at the edge, it probably is — but you won't know how close until you simulate it. Let's talk.