At one of Danone’s high-speed dairy plants, operators now monitor production lines from tablets instead of control rooms. Machines predict when the oil in an industrial packaging line’s motor is about to run hot, which can lead to equipment failure, and alert a factory employee before the meltdown. Tech upgrades like these are making Danone’s operations smarter, leaner, and faster.
Leading the operational transformation at Danone is its chief operations officer, Vikram Agarwal. Since joining the French multinational—whose portfolio spans dairy, plant-based products, waters, and specialized nutrition and includes brands such as Activia, YoPRO, Nutrilon, evian, and its namesake label, Danone—Agarwal has reimagined operations “from being a cost center to being at the center of growth.” A veteran of Unilever with more than three decades of experience in operations and supply chain, he’s focused on Industry 5.0, where “a computer can’t replace a wrench, but it can predict when one will be needed.”
Recently, McKinsey Senior Partners Pierre de la Boulaye and Søren Fritzen met with Agarwal at Danone’s new Paris headquarters. They discussed how Danone is building production capacity with precision, using data and AI to optimize performance, and preparing the company’s more than 47,000-person operations workforce (which is about half of Danone’s overall head count) for a digital future. Their conversation has been edited for length and clarity.
A transformation blueprint
McKinsey: Since you joined Danone as COO in 2022, you’ve led the company through an operations transformation. What were the main operational challenges you set out to address through this transformation?
Vikram Agarwal: One of the first challenges was to redefine the role of operations from being a cost center to being at the center of growth. Over the last three years, the goal has been to build a truly end-to-end, consumer-responsive supply chain, one that feeds consumer demand rather than simply fills factories and pushes products onto the market.
Overall, we needed to be more cost-competitive and customer-focused, and we needed to build capacity with precision—knowing where the business is going and how it’s growing. We also needed to invest in the future through digital systems and, most important, talent.
McKinsey: How did you structure Danone’s operations transformation? What were the key initiatives?
Vikram Agarwal: There were three parts of our transformation: first, digitalizing our planning processes. We upgraded our planning tools, which were running on legacy systems.
Second, we needed to focus on anticipating shifts in consumer tastes and creating capacity to meet those needs. Fortunately, our strategic bets now align with what’s growing, so we’re in the right categories, such as high-protein and plant-based yogurt options as well as specialized nutrition.
When we talk about building capacity, there are three ways to do it. First, you can release capacity—get more output from existing assets—by adopting manufacturing excellence systems. For us, that’s Integrated Work Systems1 and digital manufacturing. Second, you can transform: When a product or category isn’t growing, you can convert that production line into one that is. You change the equipment, but the people remain the same. If you change the equipment and hire new people, that’s new capacity creation—the third path.
This distinction shapes the financial metrics of investment. When you invest to create new capacity, you need a small multiple of that in net sales growth to pay it back. But if you transform existing capacity without adding fixed costs, that ratio drops fast. Therefore, how and where you plan affects not just the capital you spend but also your P&L. Once that investment is commissioned, you start to very quickly leverage the investment in your P&L.
Partnership, precision, and portfolio upgrades
McKinsey: You’ve also launched initiatives in areas like procurement and supplier partnerships. How are these programs helping Danone create value?
Vikram Agarwal: One example is our Partner for Growth program. Any company that manages product development entirely on its own will eventually hit a ceiling, limited by its knowledge and resources. So why not rely more on our key suppliers? One of our major partners specializes in dairy ingredients, for example, which is expertise we couldn’t replicate in 20 years. The idea behind Partner for Growth is to bring the best technology into Danone products, ideally giving us early access to new developments from our partners.
The same goes for sustainability and decarbonization. No company can make global agriculture less carbon intensive on its own; it must be a joint effort among partners with complementary capabilities.
McKinsey: What does Danone’s portfolio today reveal about the future of packaged foods?
Vikram Agarwal: When I look at the portfolio today, it’s very different from what it was even three years ago. What has taken over in the food industry, particularly in dairy but also in plant-based and other categories, is the focus on health benefits: high-protein, gut health, and immunity, for example. Food is much less about being something that fills your stomach; it’s about the potential impact it can have on health and well-being. That’s what consumers are increasingly looking for.
I also see personalized nutrition becoming more prominent. Every gut microbiome is different. It’s also where digital can potentially play a role. One example is our AI-powered Stool Tracker, which helps caregivers understand their child’s gut health through image recognition of the stools.
McKinsey: A few years ago, the World Economic Forum recognized Danone for its digital manufacturing work.2 How have you built on that momentum to take Danone’s digital manufacturing to the next level?
Vikram Agarwal: We started with three or four factories under our Digital Manufacturing Acceleration, or DMA, program. It now covers more than 80 factories worldwide, and another 40 will join over the next year. But what qualified for a Lighthouse Award a few years ago wouldn’t even be nominated today.
We know about Industry 4.0, but I prefer to talk about Industry 5.0, which focuses on the symbiotic relationship between machines, artificial intelligence, and humans. Industry 4.0 emphasized data connectivity, analytics, and AI but didn’t fully recognize the human interface that drives them.
We have a few pilots underway. One is a remote-control tower, combining an asset digital twin and agentic AI, allowing an operator to control, simulate, and predict the run of a machine from a distance via a tablet. The idea is to elevate operators into digital experts rather than purely physical attendants. Of course, a computer can’t replace a wrench—but it can predict when one will be needed, shifting from breakdown maintenance to predictive maintenance.
Such centralized control towers will create breakthroughs in line performance, leveraging super-experts in remote performance monitoring. Maybe they’ll get an alert that says, “Your oil temperature is running high; take these steps” and advise operators in real time.
My one nonnegotiable principle is that we won’t digitalize for the sake of digitalization. We do it only when it delivers measurable business value—whether through better capacity utilization, improved customer service, or lower costs.
McKinsey: Danone has taken steps to drive efficiency and innovation across the business in areas like portfolio optimization, one example being its Design to Superior Value (DtSV) program. What impact has it had so far?
Vikram Agarwal: We have been delivering record levels of added value over the past three or four years, but the program has been challenging because our portfolio is highly complex.
We’ve dialed up our efforts by connecting product design to the competitive benchmarks in the market. As part of DtSV, we are also simplifying the portfolio. If you look at one category in one zone, say, dairy in Europe: We put in the best high-speed production lines available in the market. The minimum order quantity of a single SKU run, on average, used to be a relatively small number of pallets. Because each SKU run was so small, the line had to stop frequently for a changeover to the next product, which reduced efficiency. You produce the required amount, you stop, you change over—even if it’s a simple changeover like a label change.
Simplification of the portfolio, which in turn leads to more efficient line management, is the key to getting this right, though it’s hard to quantify the improvements. When I ask people, “What are the savings coming out of simplification?,” I get a few pennies here or there. But the impact on the totality of the business is much, much larger. So, you have to look at simplification as an overall business initiative rather than just a factory program.
The human side of operations
McKinsey: Beyond digital manufacturing, where else do you see fundamental change needed in operations?
Vikram Agarwal: We are making great progress in procurement with gen AI, using it for cost forecasting and developing should-cost models for each ingredient.
On a related topic, our COGS [cost of goods sold] forecasting system has improved over the past few years. Unless you’re able to forecast the COGS accurately, you cannot forecast the P&L. Every year, if I have to land around €13.7 billion COGS plus-or-minus 2 percent, then we have to move with surgical precision. I cannot achieve that with manual systems, because manually analyzing the data takes months, so you’re always trailing behind. By the time you’re getting into your next forecast cycle, the markets have already moved.
That’s why we use AI and machine learning tools that can predict cost changes, using inputs related to the market or Tiers 2, 3, and 4 feedstocks, for example, to predict what our costs will be. Those tools consider that data, apply it to our bill of materials, and show the financial impact on a single country or category. So, we’re moving from managing a few thousand variables to a few million variables, enabling much greater accuracy.
Beyond digitalization, a major transformation lever has been our work in sustainability. I’m particularly proud of what our operations team has done on decarbonization. Danone was the first company to commit to a 30 percent methane reduction by 2030 from fresh milk, and we already achieved a 25 percent reduction in 2024, largely through regenerative agriculture. We’re helping to make milk farming both less carbon intensive and more economical for the farmer. Our playbook—which covers soil, manure, cattle genomics, animal welfare, and other factors—is now being deployed at farms worldwide.
McKinsey: All of this transformation clearly depends not just on technology but on people. How are you building the skills and mindset needed to make this shift?
Vikram Agarwal: We cannot run Industry 5.0 or AI top-down. You can create systems and software top-down, but the use cases for those things have to come bottom-up. I need all 47,000-plus people in operations contributing to that, rather than ten people sitting in this office trying to define what the use cases would be.
To get there, you need to train people. We established the Industry 5.0 Academy at one of our factories in Poland. It’s housed within the factory site. Through the academy, we have already trained about 20,000 people on, for now, the basics of digitalization.
We just started it in mid-2025, but we already have 100 user stories. Imagine the power of that, especially since it has the potential to be rolled out across more than 140 factories.
There are also social benefits. The newer generation is no longer interested in being an industrial worker. Going to a shift from 6:00 p.m. to 2:00 a.m. is not something that inspires people. If I transform the shop floor, then we can attract the next generation of workers.
McKinsey: Cultural change is at the heart of any transformation. How do you measure and track progress on that front?
Vikram Agarwal: We use what we call the 4C Performance Management Framework: customer-centricity, cost competitiveness, cash efficiency, and carbon reduction. How deeply are those ideas rooted in operations? And are they recognized by the P&L holders? What’s the speed at which we are making progress on those fronts?
The way we set our annual targets as a team is based on those four Cs, and it cascades down, starting with me. If you go to a frontline manager in one of our factories and ask about their goals, they will talk about how they’re aligned to the four Cs.
Reflections on a transformation
McKinsey: Looking back to when you began initiating the transformation of Danone’s supply chain in 2022, is there anything you would have done differently?
Vikram Agarwal: I wish I’d been able to read the consumer shift into the high-growth segments earlier.
We need to be able to do long-term capacity planning proactively from the signals available in the market, rather than following a classical method—where marketing creates a forecast and category strategy, then other people convert it into a sales forecast, then the operations team converts the strategy into a demand and supply forecast. That linearism has to go. So, I just wish I’d started much earlier.
McKinsey: What advice would you give to other COOs based on your experience leading this transformation?
Vikram Agarwal: The most important thing for any supply chain leader is people leadership. In our case, you have to lead more than 47,000 people in the same direction. To do that, you need leaders who are authentic and inspirational and who others will follow.
The second point I’d note is that in operations, you cannot take a sabbatical from launching new initiatives. You blink, and the world has moved past you. Perform in the present but plan for the future.
The third thing I would say is to have a can-do attitude. Very often, supply chain people are known for saying we can’t do something, whether for costs or other reasons. But that attitude reinforces the idea that supply chain leaders wait to be told what to do rather than participating in the growth story.
