On the technical front, the challenge of applying AI across markets lies in the data itself. Mark is blunt about what’ s needed:“ To get the true value from AI’ s forecasting capabilities, the business should have access to high-quality, accurate data.” If the raw information isn’ t clean, no algorithm can produce meaningful predictions. Testing that data by checking format compatibility, removing duplicates and balancing values is essential. Only then can businesses trust what the model returns.
Southern Glazer’ s has already seen the impact of AI on its inventory processes. While it doesn’ t manufacture products, it still needs to avoid overstocking or understocking – a tough ask when customer behaviour shifts week to week.
Diego points to AI-driven forecasting as a way to reduce bias in ordering decisions.“ With the use of AI, we saw improvement in our forecasting BIAS metrics as human sentiment is taken out of the equation. This is helping us tremendously with significant inventory
70 July 2025