P&G's AI Factory Scaling Data Science Innovations
Consumer goods company Procter & Gamble (P&G) is placing big bets on artificial intelligence, looking to scale the technology across enterprise-wide efforts like product and package innovation, media planning and buying, distribution and retail activities, manufacturing, and back-office operations.
Vittorio Cretella, CIO at P&G, said the company is focused on becoming an “AI-first” business, sharing strategies at the recent Gartner for IT Symposium and highlighting key takeaways on LinkedIn.
The company has already begun innovating within this space, with Cretella pointing to success with a proprietary machine learning platform that is being leveraged across 80% of its global business.
“With unprecedented speed, the AI Factory is already proving to reduce complexity, making our data scientists 10X faster and more efficient,” said Cretella.
One specific use case includes Pampers MyPerfectFit, which is tapping the resources of the company’s AI Factory to develop an AI-based diaper recommendation for the right diaper size to prevent leaks.
“Parents can access the tool in our Pampers Club Mobile application, choose to opt in and provide the date or birth, the baby weight and height, diaper fit details, and the algorithm will recommend the right diaper size with a 90% size accuracy rate,” he said. “Thanks to AI Factory, we were able to rapidly and efficiently train and deploy the algorithm for several markets including U.S., Canada, Germany, France, UK, Spain and Japan.”
Earlier this month, P&G also shared that it is using AI-based smart algorithms within its fragrance development initiatives, giving the company better control over digital scent creation, increasing speed to market, and elevating processes across product development and design.
The company is looking to continue investing in AI-powered efforts, focused on three key areas to scale the tech: clearly articulating business purpose, building organization AI fluency and skills, and standardizing and automating developments in AI to increase speed and efficiency.
Specifically, this means moving away from “one-off initiatives and towards scaling algorithmic solutions across multiple categories and markets.”
Step one, said Cretella, is to not digitize for the sake of digitization. There needs to be business value, which then needs to be aligned to AI results.
He said companies should ask themselves, “How much is your consumer reach going to increase thanks to an AI-driven marketing campaign or execution?” for example, or “How much hundreds of thousands of dollars of product scrap are going to be avoided by AI-driven preventative maintenance of a production line?”
Second is a need to build in AI fluency across the organization, as AI is a tool to augment efforts, not replace employees.
“It is critical not only to insource key technology capabilities, such as data science and machine learning engineering, but above all, to ensure all employees become familiar with working with AI,” said Cretella.
Building trust within this space is critical, particularly as gaps in AI can lead to data inconsistencies and a “lack of explainability.” To combat this, the company offers an internal training program, in partnership with Harvard Business School, with two levels of certification focused on the fundamentals of data science and use cases that demystify AI and “invite replication of success.”
Last is standardization of AI. This is where the company’s AI Factory comes into play, allowing data scientists to access automation and machine learning tools that can build the infrastructure to support AI projects. This includes libraries of reusable and sharable source does and software development kits.
“AI has become a critical factor in brands’ ability to develop superior solutions that can deliver on consumer and retailer preference, reduce cost, and enable rapid and efficient decision-making,” said Cretella.
“It helps us reach consumers with greater precision – with the right content, on the right channel and at the right time, all while respecting privacy,” he added. “It helps to jointly create with our retail customers great shopping experiences across all channels. It enables superior and fast product innovation. It powers our operations including maximizing quality, resiliency, and the consumption of energy and water in our manufacturing plants.”
How Is AI and Machine Learning Changing the Way We Manage the Supply Chain?
Shifting consumer demands, geopolitical uncertainty, labor constraints, and residual pandemic-era behaviors can all have a negative impact on supply chain management. As a result, more companies are leveraging AI in the supply chain to improve their supply chain resilience. Learn how artificial intelligence and/or machine learning are being currently leveraged in CPG organizations
