Nestlé’s CIO says the value of the food giant’s AI investments goes well beyond efficiency

When Chris Wright is asked about the efficiency gains he hopes to extract from artificial intelligence, the chief information officer at food giant Nestlé is quick to discourage an extensive focus on bottom-line savings.

“One of the big learnings we’ve had is don’t focus too much just on the efficiency” says Wright.  “We’re finding in the way we track our business cases that the biggest value is in buying better, or reducing energy in a factory, not only being more humanly efficient.”

One pilot project that’s putting this viewpoint into practice is a generative AI assistant that can create fulfillment plans for distributors and retail customers. For years, Nestlé has standardized that process as much as possible, and for big customers—think Walmart and Target—order fulfillment is highly automated. But using AI in this manner can not just create the fulfillment plans faster than what a human can do, it can also produce better plans.

“That’s improving our customer order fulfillment, which can help us grow sales,” explains Wright.

It is also helpful for smaller retail customers that don’t have the digital might of a big-box retailer and still tend to place orders manually. AI is now being used to measure the impact of Nestlé’s sales recommendations that are offered to those smaller clients. Human representatives can spend some of the time they save on generating those fulfillment plans to now explain to those customers why these AI-generated plans best suit their needs. Customers that get that extra intel are proving to be more inclined to place the orders that Nestlé recommends.

Wright has worked at Nestlé for nearly three decades and has served as global CIO since 2021, steering all IT and digitalization efforts for the world’s largest food and beverage company, which ranks No. 97 on the Fortune Global 500. Nestlé’s business is vast and complex: it operates 337 factories and sells more than 2,000 brands including Nespresso coffee, KitKat snack bars, and Perrier water in 185 countries.

But Wright says 90% of the company’s core processes run on the same global template by SAP. He says this makes it easier for Nestlé to test various technology solutions in markets like Brazil or the Philippines, and then ramp up elsewhere if a pilot has proven successful. The company’s data foundation has been centralized across the entire business, so when leaders ask themselves difficult questions—how can factories be more energy efficient or in what ways can media spend be more efficient—it can pull insights from the same connected well.

“What we’ve increasingly tried to do is piece all of that together in a data foundation that connects across the value chain,” says Wright. “And we’re starting to see a lot of the value that’s coming out is our ability to rethink how you operate end-to-end, and rethink how decisions can get made end-to-end as well.”

A connected ecosystem makes it easier for Nestlé to turn on generative AI capabilities from vendors like ServiceNow and Microsoft. The latter’s Copilot is now used by around 100,000 workers on a monthly basis.

Another way it manifests is through pricing analytics. Nestlé no longer looks at pricing through silos: how much it wants to offer discounts to distributors to buy goods, what’s the company’s market share at any given point versus competitors and how it should adjust pricing to respond to those dynamics, etc. The company is using AI to look at the fuller end-to-end picture to produce more accurate pricing models, a pilot that’s been tested in India and the U.S. and can now be scaled to more markets globally.

“Because we’ve got common systems, common data models, and common processes, if I can get this right in one country, there’s a very high likelihood we can get it right in most of our countries,” says Wright. “That scale is what we are always going after.”

Nestlé is also using AI to help the company conduct auctions for transportation and commodity orders, more efficiently handle machinery shutdowns on the factory floor, and a virtual sales assistant. The latter was built after Nestlé took a look at the most time-consuming tasks—pulling sales information, details about the latest product launches, and the best practices to speak to a specific customer—and then built an AI tool that can quickly generate all those materials. 

“My belief on this—I couldn’t tell you that we can measure this yet—is that actually it’s going to make them more effective in those meetings,” says Wright. “So there will be a revenue-generating benefit.”

In October, Wright cleared a big IT milestone by completing Nestlé’s first major upgrade to SAP S/4HANA, which covered 112 countries across Africa and Asia. The upgrade covers core processes like manufacturing, supply chain, and finance, but Wright says he’s also bullish on the shift because it will allow the company to explore more potential use cases for SAP’s generative AI copilot, Joule. 

The SAP upgrade news was disclosed the same month that Nestlé announced it would lay off more than 16,000 employees under its new CEO Philipp Navratil, cost-cutting that would be more tilted toward white-collar jobs as the company targets “operational efficiency.” 

Wright concedes that “AI is not the only reason for that, but it is one of the factors.” He shares the example of a product manager that works in IT, typically supported by two project managers and a couple of business analysts. Nestlé could opt to develop AI agents to support the project manager, or alternatively, design those tools to support the top of the funnel. 

“We’re more heading towards the latter,” says Wright. “Now, you still need expertise in these areas. It doesn’t drop to zero, but it fundamentally reshapes it.”

He stresses that workers remain critical to Nestlé’s future. More than half of the company’s employees are in the physical factories, and there’s plenty of work in the field conducted by the sales team and agronomists. Because there’s more money to pump into data and analytics to make both factories and humans more effective, Wright does need to carve out efficiency gains. 

But, like in the case of the virtual sales assistant, that may not result in job losses. “Can we use that in order to visit more customers or spend more time with the important customers?” asks Wright. “It’s not necessarily a straight elimination. The question for us, is always, can we reuse that time for something useful?”

John Kell

Send thoughts or suggestions to CIO Intelligence here.

This story was originally featured on Fortune.com