Proven Strategies to Successfully Increase Forecasting Accuracy

Forecasting in healthcare, especially within the Pharmaceutical and Medtech sectors, is a cornerstone of strategic decision-making. It remains one of the most challenging areas due to evolving market dynamics, regulatory pressures, and technological advancements. In this context, the ability to accurately predict future trends and outcomes has never been more critical.

In this article, we have the privilege of speaking with Luis Enrique Caballero, Partner at LECA Solutions. He shares his insights on the most pressing forecasting challenges, explores effective methodologies, and highlights best practices to navigate the uncertainty and complexity of this high-stakes industry.

Karsten: Thank you for taking the time for this interview, Luis. Jumping into our topic right away, what would you say are the biggest forecasting challenges currently faced by the healthcare industry, particularly in Pharmaceuticals and Medtech?

Luis: Thank you for having me. Currently, healthcare companies face these challenges: 

a) Lack of reliable data, such as analogues to estimate product launches,

b) Difficulties in foreseeing market events and assessing their impact, 

c) A competitive landscape with greater complexity, fewer me-too products, and more innovative therapies, 

d) Uncertainty about National Healthcare Systems’ pricing and reimbursement decisions, and 

e) Market access restrictions (e.g., reimbursement limited to a specific patient group), which can significantly impact product performance and is typically unknown until the launch date approaches.

Karsten: How do external factors like regulatory changes, market competition, and economic fluctuations impact forecasting in healthcare?

Luis: Regulatory changes have a significant impact, both on prices and volume. For instance, changes in reimbursement status or the imposition of discounts or price cuts can directly affect forecasts and P&Ls. A VISA on a product prescription can severely reduce volumes, as only some specific specialties may prescribe it. While economic fluctuations may have a limited short-term impact, their effects can ripple over the long term. For example, an economic downturn could prompt regulatory bodies to cut healthcare expenditures, creating lasting challenges for forecasting. In the short term, these fluctuations tend to impact non-reimbursed products more significantly. It is important to recognize that each product is different, and all these external factors need to be evaluated separately, usually in a customized way for each market segment.

Karsten: What forecasting methodologies or techniques have proven most effective for managing the complexities of healthcare?

Luis: Advanced techniques like AI, ARIMA, and Exponential Smoothing are commonly used, but their added accuracy often comes at the cost of complexity and opacity. In contrast, traditional methods such as Monte Carlo simulations and regression models remain popular for their balance between accuracy and practicality. Ultimately, business expertise often proves more effective than relying solely on advanced tools.

Karsten: How can organizations strike a balance between long-term strategic forecasts and the agility needed for short-term changes?

Luis: Long-term and short-term forecasts serve distinct purposes, which is why most organizations handle them separately. Long-range planning (LRP) typically addresses strategic goals, such as market positioning or resource allocation, while short-term forecasts focus on operational needs like demand planning or sales force incentives. However, these forecasts are interdependent, and collaboration between the teams responsible for each is essential. By aligning methodologies and assumptions, organizations can achieve greater consistency and reduce inefficiencies. That said, fully integrating long- and short-term forecasts into a unified process is often impractical. The complexity of calculating monthly sales across multiple years and the diminishing utility of such granular data beyond the second or third year makes this approach less efficient in most cases.

Karsten: How are technologies like AI and machine learning reshaping forecasting and decision-making in healthcare?

Luis: AI is transforming forecasting in four fundamental ways: 

1) Much faster and more efficient turnaround of forecasting results, scenarios, and simulations,

2) Better accuracy of forecasted sales, 

3) More solid decision-making, and 

4) Management of higher volumes of data to uncover patterns, trends, and correlations that traditional methods might miss.

Karsten: What are the unique forecasting challenges associated with launching new therapies or medical devices?

Luis: The key challenge in launching new therapies is predicting market share, particularly the peak-year percentage and timing, which drive critical strategic decisions. While defining the overall market size is easier, market share forecasting is far more complex. Tools like industry benchmarks, Zipf functions, and analogues can help, but no single method suffices. Combining approaches weighted by data quality and product similarity often yields the best results. Further complicating this, factors such as competitive dynamics, reimbursement policies, and government pricing decisions remain uncertain until late in the product’s lifecycle, making accurate forecasts even more difficult.

Karsten: How can healthcare organizations ensure alignment between sales, marketing, and supply chain in their forecasting processes?

Luis: Short-term forecasting is normally the responsibility of the supply chain but must be aligned with annual operating plans (Finance, GM, HQ, etc.). Long-term forecasting is a rare gem: most departments depend on it, but it is rarely built collaboratively. In some companies, it is the responsibility of Business Insights or Commercial Excellence; in others, Marketing is in charge, and normally Finance, Sales, and international HQ have the final say. Our recommendation is to create a multidisciplinary team to align on the process, methodology, algorithms, events, and their impact as soon as possible to avoid endless back-and-forth between departments.

Karsten: What strategies can companies use to improve resilience and adaptability in forecasting during times of high uncertainty, such as pandemics?

Luis: Adaptability starts with investing in skilled teams and leveraging AI or machine learning models that can deliver fast, automated results, avoiding delays caused by manual processes. For resilience, scenario planning is essential. Companies should use historical data as a base but expand with optimistic, realistic, and conservative scenarios. Monte Carlo simulations and sensitivity analyses can identify high-impact variables and reduce uncertainty around key factors like prevalence or pricing. Lastly, strategies like maintaining higher inventories, local manufacturing, and shorter lead times can help mitigate risks such as out-of-stock situations during demand fluctuations.

Karsten: How can healthcare companies enhance data collection and integration to improve forecast reliability?

Luis: The precision of forecasting is expensive. Market data from audits and primary or secondary market research can be costly, and the expense skyrockets when conducting a multi-country conjoint analysis with a robust sample size. If you want accurate forecasts, you will need to invest heavily in the inputs of your model.

lllustration: The Healthcare Forecasting Ecosystem: Key Components and Interconnections

Karsten: What practical advice would you give to healthcare leaders looking to improve their forecasting and decision-making?

Luis: Sales forecasting is critical and must be a company priority. Treat each product uniquely—what works for one therapy area or country may not work elsewhere. AI can assist, but it cannot replace sound business judgment. Lastly, establish strong internal processes with cross-functional collaboration to achieve your goals.

Karsten: Thank you, Luis, for sharing your valuable insights and expertise on this important topic! As we move into an era of unprecedented change, staying ahead requires not only understanding the tools at hand but also rethinking how we approach forecasting. I am confident this interview has provided our readers with actionable insights and sparked new ideas to tackle their forecasting challenges.

In summary, effective forecasting is a blend of science, technology, and business acumen. While advancements in AI and data analytics provide powerful tools, the human element—judgment, collaboration, and adaptability—remains irreplaceable. By addressing challenges head-on, leveraging cross-functional collaboration, and embracing innovative solutions, healthcare organizations can turn forecasting into a strategic advantage.

#HealthcareForecasting #ForecastingAccuracy #CommercialExcellence