Leaders Opinion

Demand Planning is in a State of Flux

April 30, 2025 4 min read

Rohit Anand

Valvoline Cummins Pvt Ltd India, Head of Planning

Demand Planning

What if the forecast guiding your multimillion-dollar supply chain decisions was already wrong before you even acted on it? According to a recent survey, more than 70% of supply chain leaders admit that their forecasts are frequently inaccurate, resulting in a cascade of inefficiencies, ranging from excess inventory to stockouts that are painful. 64% of organizations expect unexpected customer demand variations to increase over the next 5 years.

In today’s highly dynamic environment, demand planning is not just challenged—it’s being fundamentally redefined. The traditional models and methodologies that once provided a sense of certainty are rapidly losing relevance. In their place, a new generation of tools and thinking is emerging, designed to cope with disruption, complexity, and rapid change.

Shape Picture 1, Picture

Conventional forecasting methods—based on deterministic logic and normal distribution assumptions—have long been the backbone of demand planning. But these models are ill-suited for the volatility now shaping global markets.

Underlying many traditional approaches is the assumption that historical demand patterns can reliably predict future needs. This presumes that the system generating those patterns remains stable—a presumption that no longer holds. From geopolitical shocks and pandemics to rapid digitization and climate impacts, the supply-demand equation is now influenced by far more than just seasonality or trend lines.

What we’re witnessing is not merely uncertainty, but persistent disruption.

A Landscape in Transition

To navigate this disruption, businesses must recognize a series of fundamental shifts in the demand planning landscape:

Picture 1, Picture

From uncertainty to disruption
Traditional planning accounted for variability. Today, organizations must grapple with structural changes that render historical baselines obsolete.

From deterministic to probabilistic models
Probabilistic forecasting allows planners to simulate a range of possible outcomes, rather than betting on a single-point estimate.

From short-term fixes to long-term strategies
Long-term structural change, such as the rise of e-commerce, sustainability mandates, and decentralized supply chains demands forecasting systems that look beyond the next fiscal quarter.

From linear to circular economies
With more businesses embracing circular models, planners must now forecast for reuse, refurbishment, and end-of-life product flows.

From static expectations to fluid consumer behavior
Demand is increasingly influenced by non-traditional signals: social media sentiment, influencer impact, and real-time market shifts.

Shape

Embracing the Digital Toolbox

To stay ahead, leading organizations are turning to a new generation of forecasting tools. These technologies are not just iterative improvements, they represent a fundamental rethinking of how demand can be predicted, shaped, and responded to.

Machine Learning (ML)
ML algorithms, both supervised and unsupervised, adapt over time. They learn from past errors and adjust forecasting models dynamically ideal for fast-changing markets.

Neural Networks
These advanced AI models are capable of identifying nonlinear, complex patterns in demand data—far beyond what traditional models can detect.

Generative AI (GenAI)
When historical data is limited (e.g., for new product introductions), GenAI can simulate synthetic but realistic data to train models and support early-stage forecasting.

Digital Twins and Scenario Planning
Digital twins of supply chains allow for real-time simulation of demand scenarios. Scenario planning, once theoretical, is now operationalized—helping businesses plan for multiple futures at once.

Big Data & Real-Time Demand Sensing
External data from social platforms, online reviews, search behaviour, and macroeconomic indicators is being harnessed to sense demand shifts as they happen.

Together, these technologies enable a more resilient, responsive, and intelligent demand planning function—one that’s built for uncertainty rather than hindered by it.

Shape

Beyond the Buzz: A Balanced Approach

Despite the excitement surrounding AI and analytics, it’s critical to acknowledge the limitations. Every model is only as good as its assumptions. Over-reliance on algorithmic forecasts, without human oversight or contextual understanding, can be just as risky as using no data at all.

Successful demand planning today requires balance. It blends the speed and scale of machines with the intuition and strategic thinking of experienced professionals. It’s a team effort—powered by data but guided by judgment.

Shape

Conclusion: Planning for the Unplannable

In this era of constant change, demand planning must evolve from a reactive forecasting function to a forward-looking strategic capability. The organizations that succeed will be those that embrace this transformation—investing in new tools, fostering cross-functional collaboration, and making planning a core business competency.

“The state of flux is not temporary. It’s the new operating environment. The sooner companies acknowledge and adapt to this reality, the better positioned they’ll be. Not just to survive disruption, but to turn it into opportunity.”

Shape

Explore the latest edition of Journal of Supply Chain Magazine and be part of the JOSC News Bulletin.