Email
sinha.rk.dr@gmail.com
Forecast accuracy continues to be a pain point.
Supply chains which depend on monthly forecasting often find that poor forecast accuracy is their major pain point. Should they go for more sophisticated AI/ML based forecasting models? Should they collect more granular data? Their experience in adopting these two approaches hasn’t been encouraging either.
Our analysis indicates that the main reason for poor forecast accuracy is not the sophistication of models or the quality of input data, although these two do have some impact.
The main reason is the process of forecasting… the time lag involved, the frequency of refreshing the forecast, and the need for aggregation across product and geography dimensions for a meaningful discussion at the S&OP meeting.
Unless we solve for these three underlying issues, I am afraid these companies will have to live with poor forecast accuracy.