This year marks a pivotal opportunity for organizations to optimize their operations. Leading companies are already setting a course to redefine their operations as a source for competitive advantage and growth.
In this episode of McKinsey Talks Operations, host Daphne Luchtenberg is joined by Dan Swan and Ruth Heuss, global co-conveners of McKinsey’s Operations Practice. Together, they explore the defining operations themes for 2026 and look at what we’re seeing the bold movers do as they set themselves up to win the race to rewire.
The following conversation has been edited for length and clarity.
Daphne Luchtenberg: Looking back for a moment, we know clients were focused on remaining resilient, looking for ways to increase productivity, and charting the course for growth. What bold moves did we see some of these clients embark on in 2025?
Ruth Heuss: Let’s start with productivity, which was obviously a huge theme in the last year for many regions and many clients. It’s growing without growing the amount of resources needed. And I think some of the fabulous players have actually done that really well with all their digital and AI business models. And many other companies are looking to learn from these examples, to try to do their manufacturing with limited resources, at least on the people side, and just working much more toward a leaner and more efficient operating model. I think that’s one big theme we saw last year.
Daphne Luchtenberg: And Dan, what about from your side?
Dan Swan: Anybody reading the newspapers over the last year will have recognized that there’s been pretty unprecedented volatility, whether that’s inflation, the implications of tariffs and trade, or geopolitical unrest. And in the midst of all that, companies have had to figure out how to make sure they can source their products, deliver them the right way, and, quite frankly, minimize the disruption to their customers and their operations. So I think the implication has been a lot of people figuring out creative ways to get products and services to their customers, and a lot of people spending time on being more resilient.
And the one thing that cuts across both productivity and resilience is the leveraging of technology, AI, generative AI, et cetera. I think 2025 marked a bit of a turning point for clients going from testing and learning and making small bits and pieces of progress to real separation from the winners using technology and AI to drive the value that they’re seeing, both in productivity and to help them be more resilient. So it sets the stage for an exciting 2026.
Subscribe to the McKinsey Talks Operations podcast
Daphne Luchtenberg: As we look toward 2026, our insights point to three big imperatives: rewiring operations, accelerating tech-enabled decision-making, and building resilience to drive growth. Where do you see the biggest economic opportunities emerging in manufacturing and supply chain, particularly as automation and robotics move from pilots to scaled deployment?
Dan Swan: I think productivity is going to be as, or more, important in 2026 than it was in 2025. I think we see our clients in so many companies with aspirations for performance improvement this year that far exceed what they’ve done historically and on a regular basis. So the need for new technology and new capabilities to deliver on those objectives has never been higher.
So, we look at if the actual use of physical automation—whether that’s robotics, humanoids, et cetera—has really continued to grow in the manufacturing or the distribution centers. And the expectation is that the automation will get more and more affordable as time goes on and therefore continue to scale.
I think we also see that, outside of the physical supply chain, whether in planning, procurement, or call centers, the leveraging of AI is a “once in a generation” opportunity to rethink how end-to-end processes get done. And importantly, choosing to use those words “end-to-end processes” very carefully, because when we’ve seen people make pockets of investments in one or two small areas, it really hasn’t had very good ROI. Where clients have tackled end-to-end processes that can transform their business and leverage AI, generative AI, and automation to build the capabilities of their people altogether, companies have started to see really transformative results.
Ruth Heuss: To add to that, we’ve seen in the last year—especially in the services sector, banking, insurance, and similar industries, and also in telco—is that all those end-to-end-process-related industries have embarked on leveraging not only generative AI and large language models but also agents and even groups of agents—so really building out an agentic workforce.
Many of our industrial clients are a bit behind on that because it’s much more difficult for them. If you imagine a discrete manufacturing operation, it’s always more difficult to have the complete stream digitized. And there we see some leaders experimenting with ways to teach the workforce about all the different steps in the process, and the manuals, and so on, if you think about shifting one employee to another station. But there it’s a bit more difficult and, as I said, the service industries are really storming ahead here.
Daphne Luchtenberg: Do you feel that those priorities are going to hold steady? Have companies done enough yet to build true resilience?
Ruth Heuss: Well, what is true resilience in our current times? I think the name of the game is also agility. How fast can you actually make changes in your supply chain if something happens? And that’s not only because of geopolitical things. We saw this already in the semiconductor crisis. We see it when disasters and climate catastrophes strike us. So I think the faster you are in and actually changing your setup, the better you’re prepared.
Daphne Luchtenberg: Dan, let me bring you in here. The acceleration and agility that Ruth was talking about—is that where AI and large language models can help and facilitate?
Dan Swan: The ability to have more information at your fingertips and the ability for AI, large language models, and agents—as Ruth talked about earlier—to take some decisions and help you take action faster positions companies to be far more decisive and responsive than they have been historically.
I think one of the things that we’ve seen is that AI is necessary, but not sufficient. And one of the things that requires you to get to that level of sufficiency is actually building the capability of your employees to understand how they use the agents, how they use the LLMs [large language models] to actually take decisions and move quickly.
Again, it comes back to one of the things that we talked about earlier, which is that ability to really understand an end-to-end process and make that change. Whether that’s how you do supply planning, or how you do your new-product development, or thinking about the places where your employees need to have different capabilities and where you need to have and leverage technology differently is really the magic sauce.
So again, where we’ve seen people just try and point technology at a small issue, we’ve not necessarily seen companies get the ROI. But where they’ve looked at tackling that end to end, it’s been incredibly valuable.
Daphne Luchtenberg: Let’s stay with the idea of rewiring. It’s hard to generalize across entire regions, but are leaders grappling with fundamentally different challenges and opportunities? Ruth, I wanted to come to you first with this question.
Ruth Heuss: The opportunities are probably the same everywhere, but the way that people are tackling them is a bit different. It also has a little to do with the freedom you have to make choices. I work a lot in the automotive industry, so here are two examples of where we see those differences.
In China, there is an ecosystem of electric-vehicle players that have no legacy to start from. For example, they have a complete digital tool chain. So that starts from having your requirements in the tool chain, then following the requirements through, up to the testing. While, for example, in Europe, we have a lot of companies that have different tools to do one thing and then do the other.
Obviously, it’s much more efficient if you can follow the requirements through the whole process instead of having something in this tool and then in that tool and then, all of a sudden, you don’t really know if there’s any influence from that piece to this piece. And I think there it has to do with both the legacy systems that I spoke about and the tech affinity, if you will. And I think that is very strong in China and also in the US.
Daphne Luchtenberg: How are you seeing these universal principles being applied in the local setting?
Dan Swan: One thing Ruth said that is certainly true is that a lot of the same technologies and opportunities exist across industries and across regions. But how they’re applied and where they’re worth more or worth less is obviously quite different. So some of those things we see that are incredibly consistent include the notion of what are the data requirements and the data availability. How you think about really redesigning an end-to-end process. How you think about building the capabilities of your employees. How you think about the mix of automation, robotics, agentic AI, generative AI, and the application of all those different technologies into one kind of coherent technology strategy.
There are a lot of these truisms that exist across companies and across regions. There are some things that are very much driven by the specific regional context or the specific industry context. Obviously, where the cost sits in your organization can be quite different, and therefore what you pursue can be differently prioritized. You think this notion of: “What are the capabilities around your company?” And you think about the broader ecosystem of suppliers and vendors and technology providers, because you’re working in a broader ecosystem to do this stuff now versus doing everything yourself.
And then the last thing I would say is that it comes down to, “What’s the reality of any given specific business?” Whether that’s the performance of the business or the constraints of people, or time, or dollar investment, et cetera. Pretending there’s an off-the-shelf solution is probably not right, but making sure that people are learning from one another, or where the value is actually being created, is the super important balance to get right.
Daphne Luchtenberg: Let’s dig a little deeper into technology. We’ve moved from experimenting with generative AI to deploying agentic systems that automate planning, scheduling, and optimization across an operation’s functions. What do you see as staying consistent, and what do you see as changing?
Dan Swan: I would say a couple things. On the technology front, 12 to 18 months ago, everything was about LLMs, and now people are talking about and pursuing agentic at a pace that hasn’t been seen before. My hunch is that this will continue to evolve, and we’ll continue to find the next best thing, and the pace of innovation is really unprecedented for what we’ve seen.
One thing I think will continue to be true is that the way to drive value is not by asking the questions “How do I leverage agentic?” or “How do I leverage AI?” Instead, ask what are your most important business problems, and how do you leverage a full suite of technology and potential solutions to solve those business problems—being impact focused in how you leverage the technology.
I think the second thing that will continue to be true is this notion of the ecosystem. How companies are working with multiple vendors and multiple providers to help deliver these end-to-end technology implementations is going to be super critical. There are places where it makes sense to leverage more industrial solutions. There are places where it makes sense to build your own custom solutions. There are places where it makes sense to use a best-of-breed provider that might have a robotics solution, or an LLM solution. So really, companies getting very smart on how they want that to fit together and being very strategic about it is super important.
Maybe the third thing that I expect to be very consistent is the fundamental need to rewire the capabilities of our employees. If you think about someone working on the shop floor, the skills it takes to work with a robot and work in MES [manufacturing execution systems] are very different than the skills to physically use your hands, your back, or your legs to lift and be dexterous. If you think about it in a planning environment, what it takes to understand where there are variables that need to change or where there are disruptions to the plan is very different from what it takes to actually build the plan the first time. So teaching things like how you interact with the systems, how you drive problem-solving, and really get our humans focused on the things that only humans can do is a really powerful truism that I think will continue into the future.
Daphne Luchtenberg: Ruth, I want to come back to you, as you were alluding to the adoption of some of these things happening much slower in Europe, for example. What do you think the real calls to action are in Europe from leaders in manufacturing and process industries to keep pace with the rest of the world?
Ruth Heuss: I think we had probably one or two decades of enormously prosperous growth in Europe. And as sometimes happens in those cases, you are not always on your toes.
We see that if we look at our global lighthouse network initiative—which we have had for almost ten years—it’s very clear that there is a trend for many years to work much more in digital with generative AI and also with robots. And to give you just a few facts, in 2018, when we started the measurement, roughly 10 to 15 percent of the plants were experimenting with that. But today, 80-plus percent of the best plants have generative AI, and roughly 5 percent also already have robots in their plants—so fully autonomous. And I think that is something that we in Europe have to catch up with.
I also believe that industry leaders have now recognized that there is something to catch up with. And the other thing, which I think is also very promising, is that there’s still a lot of things that you can do even in an existing footprint. And that makes me quite positive that, looking at demographics and a potential future shortage of skilled labor, we will still be able to compete.
Daphne Luchtenberg: When I listen to leaders, and I hear what they’re saying and what they’re hearing from clients, I’m hearing a sense of measured optimism. There’s a notion that, as we enter 2026, while geopolitical uncertainty continues to prevail, there will also be a new vector of opportunity to create value, especially for organizations that seem bold enough to embrace rewiring and tech-enabled operations models.
So let me come back to you both and ask what are the one or two key things you hope our listeners will take away from today’s discussion that will help them achieve that impact in 2026.
Ruth Heuss: The one thing I’m really convinced about is that the speed of change will not be any slower. We see innovations coming into all kinds of operations systems at an unbelievable speed, which on the one side is a bit threatening, but on the other side it also brings us a lot of opportunities. While yesterday it was only about large language models, you can almost skip that now and concentrate on the agents. So that makes me quite positive that we can also do quite a lot, even if you start a bit later.
And we also see in surveys that 80 percent of companies are experimenting with all kinds of things, but only 20 percent are already reaping the benefits. And that means even if you start now, you can catch up. You just need to be very serious about it.
Daphne Luchtenberg: And Dan, how about you?
Dan Swan: I think the combination of this unprecedented need for productivity, the challenging growth situation for many industries, and just the reality that there’s likely to be continued supply chain disruptions means two things.
One is that it means no COO or head of supply chain is going to be bored in 2026. The second thing is that it means there’s a “once in a generation” opportunity to leverage those things, to create the focus and the clarity within your organization to drive transformational impact.
I think the encouragement would be to really think big about how you rewire that organization. What are the fundamental processes that need to change? How do you think about it as a transformation at scale rather than a pilot in one or two small areas? And how do you think about this as a fundamentally different way of doing business—from the technology adoption to the way the process works, to the capabilities of the individuals, in order to get very different results than if you pick small parts and try to make them marginally better.
Daphne Luchtenberg: Love that. Let me finish with one last question for you both: What is the one thing that you are most optimistic about in 2026 when it comes to the operations agenda?
Dan Swan: I think there are examples of companies that have started to get the value that everyone’s been promised over the past decade. It’s an exciting situation that hopefully gives people the chance to believe and the excitement to go after it. And that makes me really motivated and energized for a lot of companies to achieve pretty unique results.
Ruth Heuss: As Dan also said a bit earlier, it will not be boring. There’s tons to do. And I think it’s a great time to start now. Why not start now with all these new challenges and opportunities and just make 2026 the year of rewiring?
Daphne Luchtenberg: It’s great to end on such a positive note. Thank you both, Dan and Ruth, for being with us today, and we’ll call on you throughout the year to hear your latest updates.
