The supply chain planning capabilities that many consumer goods companies needed prior to the pandemic became even more pressing during the health crisis — and many of those needs remain the same today. It’s not just the need for a speedier, more agile, more accurate demand planning process; it’s the need for full business planning capabilities.
As we enter year two of the pandemic, brands have already implemented — or are actively seeking to adopt — modern technology solutions that optimize their supply chain and full business planning capabilities. We talked with Shastri Mahadeo, co-founder and CEO of Unioncrate, to learn how AI can help solve these issues and even unlock new value.
CGT: You’ve talked about the industry needing a new benchmark for accuracy. What does that look like in today’s CG landscape, and what more needs to be done?
Mahadeo: That really depends on the size of the company. Most CG brands have around a 40% to 50% error rate when it comes to forecasting demand. They spend millions of dollars on shuttling products between warehouses and lose out on millions in sales when they are out of stock or aren’t able to manufacture in time.
The No. 1 cause of forecast errors are massive data silos between cross-functional departments, and a ton of manual data analysis. The supply chain planning solutions that have been around for more than a decade still require brands to manually collect data and build their own forecasts.
Some automate a statistical forecast using historical shipments but the analysis of store-level data, marketing investments, etc., are all done manually. And sales, supply, demand, and marketing teams all plan in silos. Can you imagine how tedious and inefficient that is?
What brands need to adopt is fully integrated business planning. They need to bring all departments and data together into one platform that can create a forecast for them based on the actual cause of demand.
CG brands need to realize that the old way and old systems won’t work in today’s CG landscape. They need to embrace the consumer’s needs, embrace data and embrace AI so they can make sense of it all without experiencing the headaches. With proper technology, brands can consistently have accuracy rates upwards of 90%.
CGT: How would you debunk some of the biggest misconceptions about integrating AI into demand planning?
Mahadeo: A lot of brands feel that if they implement AI then they would lose control. Or, that it would take them 1-2 years to even be able to integrate it into their existing processes and systems. When you’re talking about the AI applications that legacy providers are selling to brands, it’s not really a misconception, it’s a fact.
With today's technology, AI doesn’t have to force you to lose control and it doesn't have to take you 1-2 years to implement. There is a world where AI helps you to move faster, more efficiently, and can be unified with human intelligence, not replace it. It can even tell you “why” it’s predicting something and which errors cause any inaccuracies.
There is a world where you can get AI for demand planning up and running in 2-3 months. The major misconnection is that a solution like that doesn't exist or it’s not possible. But it does! And it is possible.
CGT: Likewise, what role can the human touch play in best-in-class demand planning?
Mahadeo: I firmly believe in the power of AI to transform the way CPG companies plan and execute agile supply chain strategies. I also believe that uniting the power of AI with human intelligence is the best solution a business can have to stay ahead of demand — and the competition.
AI won’t know everything. It’s not in the field with boots on the ground like sales, supply and marketing teams are (not yet at least). Those teams have unique insights into demand and supply that an AI is not initially aware of. Things like buyer communication that results in distribution gains/losses, supply constraints or last minute marketing investments. When CG companies have access to technology that enables them to easily blend human insights with artificial intelligence — and apply those learnings across the business during the planning process — the sky's the limit.