Adidas is using machine learning technology to spot consumer trends in a way that allows it to build much greater efficiency and responsiveness into its vast global supply chain. That allows the iconic sports brand to enable its customers to design their own products and importantly, take quick delivery.
The importance of digital technologies in this process was outlined on stage today at the SAP Sapphire event in Orlando by Adidas CIO Michael Voegele.
In order to drive it towards its goal of being the best sports brand in the world, Voegele said it was critical to not only understand what consumers were saying about the brand but to enable customers to impact product design.
He told attendees, “It’s about how we use open source to create products, and importantly it’s about how we invite people into that process, and the speed about which we can do that.”
That meant changing the supply chain paradigm in an industry that has behaved the same way with the same model for 50 years, he said.
Digital sits at the core of the transformation but Voegele was quick to stress that there is no separate digital strategy.
“It is an enabler, it helps us achieve what we do. It helps us to build direct relationships with consumers. When our founder Adolf Dessler started the company he was sitting next to the athletes when he was designing the product. We can’t do that as a global company these days, but we can use technology to help us build those relationships.”
“But how do we get all these creators out there connected to our brand, how do you engage them – and then link them to our capabilities. We need to marry their creativity to our manufacturing capabilities.”
He described how SAP and Adidas worked on a project to apply machine learning to product design.
“We thought ‘What if we could bring manufacturing of product into the store.’”
This lead to the creation of what Adidas calls the ‘store factory’. The logistical problem, of course, is making sure the raw materials are available to meet the consumers need at the moment they are required, and then knowing where they are in the supply chain.
“With machine learning, we can look at hundreds of millions of pictures to determine the trends in consumer desire and then translate that into a guided a design of the individualized product.
However simply possessing the machine learning capabilities was not enough, he said. “We need to be able to disrupt the supply chain to take advantage of these trends.”
Consumers don’t want to wait weeks or months for the product to be delivered, he said. “They expect it to ship within 24 hours.”