At this point, it's basically a cliché that supply chains have been thoroughly disrupted.
I prefer to say "exposed" - adversity surfaced the weak links. Fixing them is another matter - that's why I turn to events like the Kinaxis Big Ideas in Supply Chain show.
I want to hear from customers who aren't just surviving. I want to hear from companies who are seizing this opportunity to rethink (for a big picture view of the show, see my colleague Phil Wainewright's Big Ideas in Supply Chain - Kinaxis CEO on supply chain planning myths and breaking down silos).
Better demand forecasting needs visibility
Of all the ways to change your supply chain planning, improving demand forecasting is high on the list. But moving to a truly consumer-driven supply chain is not an overnight project. Even giant consumer brands traditionally relied on bulk requests from distributors and intermediaries to drive those forecasts.
Call it what you want, but that's not really a consumer-driven forecast. When I saw Mars' Will Beery, Vice President, Global Transformations, demonstrate they've already achieved an improvement of 30 to 35% in their forecast accuracy, he had my attention. But how did they get there?
Beery's presentation at the Big Ideas in Supply Show, titled "How Mars flipped their supply chain script to make their customers ecstatic," gave Beery a chance to share Mars' demand-driven approach - and how Kinaxis' RapidResponse and concurrent planning platform supports these pursuits.
I won't spend much time on Mars the company here, as I think most readers are familiar with their presence in the consumer space. Mars is one of the world's top global confectionery, food, and pet care products companies. I'm not the only one who has chowed down a few extra Mars Snickers bars during the pandemic - though I'm trying to stick with nibbles of dark chocolate from here on out. Getting those supplies to the consumer is a monster supply chain endeavor. Consumers shift quickly, and so do lockdowns. As Beery explained to attendees, an old school approach was no longer viable:
Mars' supply chain was built at a time when demand was driven solely by scale, much like big post-World War era companies... We started with our factories, and we moved sequentially, and siloed to our customers, and it was hard-wired. It didn't even take into account a consumer.
The technology behind this was also old school. Beery:
From a technology perspective, it was, by and large, ERP-centric. Everything was built on the supply chain; ERP as the system of record. While this served its purpose very well over the past decades and gave us a lot of resiliency, the problem that we have is now is: consumer choices drive demand. This resiliency is being challenged because it's not fast enough. It's not agile enough to adjust to those consumer choices, and it's actually exposing us - and it's exposing that resiliency more now than ever before.
Supply chain transformation isn't a destination, but check out this next quote. You get a sense of how far Mars has already come:
When we head to the customer, we absolutely are doing the demand-sensing thing, using the best engines that we possibly can, to have the best Mars customer data access that we possibly can, combining that with consumer data to get a forecast accuracy improvement.
The goal? "A digitally interconnected, intelligent supply chain." Making that goal harder is the pandemic economy, which showed us how unpredictable consumer behavior can be. How is Mars getting the right demand signal? Beery says a big piece is diversifying your data sets, reaching beyond distributors - into the data of the end consumer.
No longer can our demand forecast not incorporate what the consumer might do, and just start with the customer. So with that in mind, we're augmenting our customer data sets with consumer data sets. Sometimes that's social listening; sometimes that's pandemic data sets, like which countries and geographies are open versus closed. Where are people returning to work versus not? This is going to constantly change and drive consumer and channel shifts in our demand. And we need to incorporate that into our demand planning capabilities.
Doubling down on AI, automation, and supply chain planning
Beery's team isn't stopping there. He sees AI playing a big role in their supply chain pursuits going forward:
By doing it more AI-driven, more automated, we're touching the forecast much less, and removing some bias in the equation.
Of all the supply chain investments Mars is making, Beery puts planning capabilities near the top:
One of the biggest areas of investment that we've got going on in 2021 and 2022 is in the entire planning space.
Beery looks to a diverse set of companies for fresh thinking and new supply chain models:
We've taken a lot of playbooks from technology companies that are focused on namely, external manufacturing, and trying to look at a heavy logistics and network pipeline for that. We've also looked at other sectors of CPG that are far more advanced in this space, like beauty. We're trying to put together the best strategy for Mars we possibly could in our areas of CPG, and it's very holistic. It covers demand and supply planning in a concurrent wa
To me, the meat of this talk is how Mars applies advanced technologies to the nitty-gritty of its supply chain. Mars uses AI to address inventory optimization. They are also starting to test AI algorithms at the point of order. This "order detection" can alert customers who might have made an error in their order, and prompt a conversation. Beery believes this will help to avoid "creating an inadvertent bullwhip back into our supply chain."
The process aspect is just as big: Beery shared their approach to supply chain segmentation. Instead of one supply chain, Mars is creating multiple supply chains, sometimes by product, other times by customer, or the dimension of that customer that matters to Mars. Each segment is "performance managed" and handled differently. Customer service operations are also transforming, with a heavy emphasis on RPA and automation for customer order handling.
Today's customers demand more supply chain visibility. Beery aims to address this via advanced analytics:
We're doing a lot in terms of route optimization, real time transit visibility to make sure the on-time dimension of the customer order is intact.
Then there is the use of digital twins - which prompted Beery to coin a new verb, "digital twinning." This stems from Mars' digital factory initiative, which they call "digital plants."
The areas that we're mostly focused on at the moment are mobile collaboration, inclusive of using the augmented reality technology that I mentioned previously in the warehouse space. But it's also trying to digitize our production lines with IoT, sensory and other types of equipment to get real-time data flows into the edge, and then being able to use that data to 3D model, and also drive optimization our lines through digital.
So what you're seeing here is actually real digital twinning. And it's using AI to better balance the packaging line, and it's driving great improvement for our plant operators. Our process engineers are gaining huge efficiencies out of it.
You certainly can't fault Mars for thinking small - this is a supply chain transformation worth tracking.
For more diginomica stories from Big Ideas in Supply Chain Summit 2021 visit our Big Ideas in Supply Chain event hub. The virtual event ran from June 1st-2nd and sessions are available to view on-demand until further notice. Click here to see all on-demand sessions.