Forecasting becomes more accessible with AI, but data still makes the difference

Argon & Co argues that forecasting is becoming more accessible for two reasons: supply chains now have more usable historical data, and AI-assisted development reduces the effort required to select, build, test and improve forecasting models.

COVID distortion is gradually receding

The article notes that the pandemic disrupted demand patterns, supply behavior and relationships between variables. As more recent years return to more consistent seasonal and long-term patterns, forecasting methods regain comparable history that can help distinguish recurring behavior from anomalies.

AI lowers the development barrier, not the data requirement

AI can help choose suitable error metrics and model approaches, while AI-assisted coding can reduce the resources needed to develop tailored forecasting capabilities. But those gains still depend on a continuous and reliable historical record.

Missing history can erase the benefit

Argon & Co highlights data gaps caused by system defects or migrations as especially damaging because forecasting depends on time continuity. Reliable data infrastructure, retention and ownership therefore remain prerequisites for sustainable forecasting improvement.