Beyond Safety Stock: The Role of Probabilistic Forecasts in Supply Chain (with Sim Taylor) - Ep 138

00:00 Intro and Sim’s background.
01:56 How do you describe forecasting to non-specialist audiences?
06:10 How do you explain the difference between traditional and probabilistic forecasting to the average person?
14:10 How Lokad came to a probabilistic forecasting perspective.
17:20 Is safety stock excluded in probabilistic forecasting?
20:55 What led you to embrace the probabilistic forecasting approach?
28:18 How do you apply a non-traditional forecasting method in a company that is used to using traditional ones?
35:35 How does a business team address conflicting constraints in a traditional forecasting environment?
39:10 Discussion on how to properly frame the role of forecasting in a business context.
46:30 Is there a skills-based barrier to entry with implementing probabilistic forecasting?
54:57 How are data/supply chain scientists selected (and managed) for this type of forecasting?
01:03:46 The value generated with probabilistic forecasting.
01:11:49 Practical examples of applying non-traditional forecasting methods in a business setting.
01:21:34 Sim’s call to action for supply chain practitioners.


The meaning, role, and value of probabilistic forecasting are often viewed as mysterious and impossible to grasp without a degree in data science. On the contrary, probabilistic forecasting is rather intuitive, as we discussed with our guest, Sim Taylor. Sim is the Director of Analytics and Data Science at Petco, where he oversees a team of data scientists who leverage non-traditional forecasting methodologies.

During our conversation, several crucial issues surrounding probabilistic forecasting are put under the microscope, including the skills required to implement it, its function within a business setting, and how to convince people to adopt it. Recommended viewing for anyone looking for a crash course on the topic!