PRICING AND PROMOTIONS
Fresh Flowers Group needed better ways to forecast supply and demand. This was particularly challenging given the perishable nature of its products and the extreme volatility in demand (driven by various calendar events such as Valentine’s Day).
We fed in historical point-of-sales data and augmented it with significant calendar events (e.g. Chinese holidays). This model was tested with 3 months of hold-out data, comparing losses from overstock/understock scenarios.
The insights generated by daisee Harmonee allowed business rules to be created in order to make predictions about future stock purchases. AI prediction gave significantly better results than a traditional Excel regression method
There were peaks and troughs in our business that I did not realise were there. We had totally missed a couple of peak periods.