In the early days of the pandemic, manufacturers struggled to keep up as consumers changed their spending habits. Similarly, while Procter & Gamble has decades of data to inform how its production lines work, predictions went out the door in 2020 as toilet paper and hand sanitizer flew off the shelves, said Guy Peri, chief data and analytics officer at the consumer goods giant.
“It’s nothing like we ever saw,” Peri said in a conversation with Jaime Fitzgibbon, founder and CEO of Ren.ai.ssance Insights, during a conference recently.
Initially, Procter & Gamble planned to use artificial intelligence (AI) to disrupt how its supply chain operated in 2020, but that plan did not include the coronavirus crisis. The company’s AI was supposed to rely on past data to drive decision-making, forecasting demand and creating new action plans.
But then the pandemic rendered P&G’s historical data nearly useless. Overnight, the company had to pivot to new solutions for maintaining its supply chain. It all began with its raw material inventory, government forecasts on consumer demand, and COVID-19 response data.
Data science has grown considerably at the company. Peri’s team is crucial to helping Procter & Gamble manufacture its wide-ranging line of consumer goods while minimizing waste and delivering satisfactory products to its clientele.
“For AI to be successful,” he said, “we need strong data management practitioners to ensure we have quality, reliable data going into the algorithms so the insights and actions recommended by the algorithms deliver the right outcomes for the business.”
This means starting small. At Procter & Gamble, where more than 5 billion consumers rely on its products every day, Peri’s team works on program pilots in small settings to study its algorithms and outcomes before scaling up.
COVID-19 may have thrown a wrench into the company’s plans, but Peri said the new features of “normal life” — fast delivery, remote work, and the growth of digital commerce — make this an interesting time for companies to test new business models.
In the new direct-to-consumer world, AI is a crucial tool to help companies predict how consumers want to shop. But it can’t be done with AI alone — 80% of the task is adopting a culture shift and 20% is rooted in the data, analytics, and technology needed to help the company solve real-world problems, Peri said.