Why you probably should stop chasing improving Forecast Accuracy by Dr Alan Barnard (15min version)

Why You Should Stop Chasing Forecast Accuracy?

Many companies are investing heavily in AI, hoping it will be the holy grail for improving forecast accuracy to reducing avoidable shortages and surpluses at the retail or consumption point level.

However, in my keynote at SCTECH24, titled "Why You Should Stop Chasing Forecast Accuracy," I made three key points that challenge this belief:

1️⃣ Forecast Accuracy at the Point of Consumption will always be poor:
At the retail store level, where consumption happens, forecast accuracy is typically in the range of 20-30%. Despite advancements in AI, achieving the level of improvement needed at this level to significantly reduce shortages or surpluses is simply not feasible. While AI might improve accuracy at higher levels like distribution centers, this is not where the major issues of shortages and surpluses occur.

2️⃣ Focus on Switching to Demand-Driven Rules
Rather than focusing on forecast accuracy, companies should shift from forecast-driven rule sets like Min-Max and EOQ to more dynamic, demand-driven rules like Theory of Constraints (TOC) or DDMRP. As seen in the attached results from our Inventory Digital Twin, TOC outperforms Min-Max Perfect World by a factor of 3x even in real-world conditions. What’s fascinating is that the gap between the perfect world (where forecasts are always correct, and suppliers deliver on time) and the real world is not as wide as one might expect when using Min-Max. This underscores that switching to demand-driven rules offers much greater leverage for reducing both shortages and surpluses.

3️⃣ Prioritize Agility and Anti-Fragility
Rather than fixating on forecast accuracy, we should focus on building supply chains that are both agile and anti-fragile. This means creating systems that can minimize the impact of negative stresses while maximizing the opportunities from positive ones. By doing so, companies can navigate uncertainty more effectively, capturing upside opportunities while mitigating downside risks.

In short, the real competitive advantage lies in creating supply chains that don’t just rely on better forecasts but are built to thrive amid uncertainty.

PS: For those interested to know if forecast accuracy is a major leverage point or not, and how much your company will benefit from switching to demand-driven rules like TOC or DDMRP, you are welcome to reach out to me via DM to explore doing a pilot with our Inventory Digital Twin - for more information visit www.supplychaindigitaltwin.com