How pharmaceutical companies build forecast reliability

‍

About Frank Binder

Global supply chain leader with ~20 years of experience across Roche, Accenture, Novartis, Celgene, and currently Santen. With a PhD in physical chemistry and hands-on manufacturing experience (API production), Frank combines shopfloor understanding with global supply chain transformation and IBP leadership.

‍

The problem pharma planners face that others don't

Pharmaceutical planning isn't like planning for consumer goods or automotive. You can't aggregate demand across markets by default or substitute SKUs freely. Every decision carries regulatory weight.

‍

Frank learned this early. "We have to be sure which SKUs we are allowed to bring on the market in Malaysia or in France. It could be that they're made at the same manufacturing site, but maybe according to different recipes. We have to be 100% sure that we get it right."

‍

The stakes are real. Product information, packaging, and labeling are all approved by authorities. If you get the SKU wrong, you're not just wasting inventory. You're violating regulations. Getting the right SKU to the right market is non-negotiable before you even get to the business trade-offs.

‍

"You cannot plan demand planning in isolation and just look at numbers," Frank explains. "You need to know how your whole supply chain actually looks and how the product production process even looks." In pharma, data tells you one story. But the regulatory blueprint tells you what's actually possible.

‍

Most planners optimize cost or service. Pharmaceutical planners optimize legal constraints first, then try to make good decisions within those constraints. That changes everything about how you approach reliability.

‍

Build forecast reliability through disciplined collaboration.

Here's what Frank discovered: better forecasts rarely come from algorithms alone. They come from structured collaboration and systems that make the process repeatable.

‍

At every company, Novartis, Celgene, and Santen, Frank set up integrated business planning processes where commercial teams and supply chains sit together and align assumptions. "It is crucial to collaborate with the commercial teams," he says. "We have discussions on the demand plan where we really sit together and understand each other's assumptions."

‍

This isn't a quarterly meeting where sales hands over numbers. It repeats every month with the same people around the same table. "If we do it in a more informal way, it gets chaotic quite quickly," Frank warns. The structured process matters as much as the conversation itself.

‍

The collaboration extends beyond the commercial. Finance joins to understand the financial implications of different plans. "When we have to make trade-off decisions, I think it is very difficult to talk about replacing X million bottles of product A with Y million bottles of product B if you don't understand what is actually the contribution of that product or what is the gross margin," Frank explains.

‍

Money has become a common language. Instead of debating volume, you discuss value. That changes which forecasts people push back on and which ones they accept. A high-volume, low-margin product might need less accuracy than a niche medication with premium pricing.

‍

The forecast doesn't improve because your system got smarter. It improves because the people who know what's actually happening said it out loud in a structured setting.

‍

Standardize processes globally, but let markets make local decisions.

Frank manages supply chains across dozens of countries. Each has different customers, regulations, and distribution models. Yet every region uses the same planning methodology.

‍

"In the demand planning process with the different markets, we have to use the same approach, the same granularity, the same frequency of data collection, and the same frequency of discussion of the demand plan across all regions." This isn't a rigid bureaucracy. It's the only way to have meaningful global conversation.

‍

If Malaysia's forecast is built using one method and France uses another, planners comparing their books can't have a coherent conversation. "If we do not have the planning processes standardized, it will be nearly impossible to even have a meaningful conversation," Frank says.

‍

But standardization stops at distribution. "There is no point in making sure that the distribution in Malaysia works the same way as the distribution in France." Malaysia might have one warehouse. France might have satellites. Customers look different. Local teams have decided.

‍

The system, whatever it is, isn't enough on its own. Methodology and definitions must match across regions first.

‍

The lock-in points are data definitions (Malaysia's demand equals France's demand), planning frequency (monthly reviews everywhere), and planning methodology (same forecast approach). Everything else adapts to reality.

‍

That's how you avoid both chaos and over-centralization. You pick what matters globally and protect it fiercely. You let everything else be regional.

‍

Use your control tower for execution, not strategy.

Many companies have a "control tower" that's just a buzzword. Frank has one that actually works, and it's separate from planning.

‍

Integrated business planning handles six months to 24 months out. The control tower watches the next three months, asking: Where do we have inventory shortages? Can we accelerate production? What can our forwarders move faster?

‍

"This is really driving a lot by data that we have out of our systems, and also we look at data from our forwarders where appropriate to see what can be accelerated," Frank explains. Power BI reports surface the risks. People decide on the response.

‍

Distribution issues hit the control tower, not planning. Plant execution topics (like short-term production accelerations) typically land in the control tower. Planning sets the direction. Execution handles surprises.

‍

This separation prevents planning meetings from devolving into crisis management. You have different conversations at different horizons with different urgency levels. Strategic planning doesn't get hijacked by today's shortage.

‍

Your next steps

If you're trying to improve forecast reliability in a regulated or complex supply chain:

‍

• Schedule monthly collaborative planning meetings with commercial and finance. Not quarterly, not ad-hoc. Monthly, at the same time, same people. Align assumptions before you lock the forecast.
• Define your non-negotiable data standards globally, then document them. What does "demand" mean in every market? How do you count on it? Write it down and enforce it everywhere. Everything else can be flexible.
• Set up a separate 3-month control tower for execution issues. Don't let production shortages and distribution problems hijack your 12-month planning process. Different time horizons need different meetings.
• Audit your collaboration for structure, not just attendance. Are you actually aligning assumptions or just sharing numbers? Is everyone's department represented? Does it happen on the same schedule every month?
• Build your forecast accuracy measures around actual business impact. A 30% error on a high-margin specialty product costs more than a 10% error on a low-margin commodity line."

Forecast reliability in complex environments comes from disciplined collaboration, not better data or smarter systems. When everyone sits down with aligned assumptions using consistent definitions within regulatory constraints, the forecast gets better, not because the math improved, but because people stopped guessing.

‍

Start with one structured monthly meeting. You'll be surprised by how much clarity that creates.