240. Improving strategy with AI

First of all, we understand the limitations of the tools, but we understand what they bring. And what they bring is the ability to formulate hypotheses a lot more quickly. So it changes number one, uh the time that you have to actually s discuss, develop those hypotheses. And some of the roles related to the thought partner or the simulator also help you validate, test, prove, disprove the hypotheses that you have. And the team has a lot more time to be creative, to think out of the box.

And the box now being the way AI is actually framing some of the strategy problems, right? Which is which is very important. Number three, it comes with responsibilities to understand what the tool is doing.

From McKinsey Company, I'm Sean Brown, and welcome to Inside the Strategy Room. You just heard Bruce Delteil, one of our guests today, talking about how generative A I and A I more broadly can help corporate strategists in their role. And today, we'll focus our discussion on the emerging role of A I in the strategy room. While artificial intelligence can't, and our guests believe won't fully replace human logic and interpretation in a domain as complex as strategy, the technology is already being used to significantly accelerate the strategy planning process, allowing teams to focus on bringing more creativity and innovation to the process.

And now, to talk us through some of the opportunities, risks, and approaches, we're joined by three guests who co-authored a recent article on this topic that we're also linking to in our show notes. Bruce Delteil is the managing partner of our Vietnam office and leads our strategy practice in Asia. He serves clients on all aspects of strategy development and, more recently, the impact of gen A I. Bruce, welcome to the session.

Thanks very much, and very happy to be here.

And Antoine Montard is the director of client capabilities for our strategy practice and based in Lisbon. He leads our strategy research and insight center, helping develop novel approaches to strategy, building the capabilities of teams and clients, and also serving as a strategy expert on client projects. Antoine, welcome to the session.

Thanks for having me.

Andrea Tricoli is an associate partner in our strategy and corporate finance practice and based in London. He leads our initiative to apply generative A I to our strategy method. Andrea also serves organizations on how to use analytics and A I to drive better decision making, specifically across the finance and strategy functions. Andrea, welcome to the podcast.

Thanks, Sean. Great to be here.

Antoine, let's start with you. Before we get into a more detailed discussion about A I in strategy, perhaps we can start with really defining what is strategy. What do we really mean by that?

There are many definition of strategy out there. The definition you use is that strategy is a coherent set of hard reverse choices made ahead of time in the face of uncertainty to create and capture economic surplus. So three different elements. So the first element is all about choices and and choices that are coherent, which. That's much about what you decide to do than what you decide not to do. That are hard to reverse, meaning that they involve a level of commitment.

Second element is those choices are made in the face of uncertainty. If uncertainty wasn't there, you wouldn't need a strategy, you would probably need a plan. And lastly, the goal of strategy is to create and capture economic value. And the way we think about it is, uh, through what we call the power curve. The power curve basically is, uh, we we stack rank, uh, economic profit, uh, for the largest companies from kind of the bottom to the top.

And that forms our curve, where what we see is that the if we divide these companies into five quintiles, the middle three quintiles almost create no value. It means that markets are efficient. It is really hard to beat the market. And you have a handful of companies at the top that creates most of the value in the world. And the way we think about strategy is all about moving right on that curve. And we know that about 10% of the companies that start in the middle three quintile end up in the top quintile over a ten-year period.

So the real question, strategy is how do you move right? How do you how are you part of this one in ten companies that makes it to the top?

Indeed, and Antoine, you've been working on something called the strategy method. Can you tell us a little bit more about that?

The the strategy method is the way that we think about how we develop strategy. So the strategy method has kind of three broad stages. I'll go through it quickly. The first stage is about designing the strategy. So it's all about uh agreeing on what's the core challenge or the the strategic imperative, uh, to start with in in aligning, then building a granular fact base to get insights into the situation and the potential. Then there's an exploration stage, which is where most of the creativity part of strategy comes into play.

But also when we'll see some of the A I tools and and lastly, in the design, we need to commit, commit about what we're gonna do and not do, and build conviction around the management team. So that's kind of the first stage. Often kind of we we we talk about strategy and execution and we we miss the middle stage, which is, uh, mobilized for execution. It's a stage that shouldn't be, uh, under invested in, because that really sets the, uh, the condition for success.

And mobilizing, uh, for execution has kind of a few components that look and feel different depending on look on your, uh, on your strategy and your bold moves you have in place. But there's a common, uh, common elements. Those common elements are about, you know, empowerment, engagement, so you need kind of clear accountability and empowerment of your teams to execute. You need to do the hard work of translating kind of bold strategic ideas into real initiative, right?

What are the steps? And, you know, it can be kind of building a business case in some instances. And so that's real work that takes a lot of time. Then it's about reallocating resources. We talk about capital allocation, but it's also about talent and how the management team spends its time. And lastly, it's about execution, right? Forcing the first step and forcing execution momentum is a lot easier if you have kind of real initiative. You're clear about the first steps and you have people empowered.

Then strategies are hard and so you're going to encounter issues and challenges along the way. And it's all about supporting the performance and overcoming these these challenges. And then the last two two steps, testing assumptions. You've made a lot of assumptions in the strategy and you have to, you don't know what the future holds. But you need to be able to to test them over time. And if they turn out to be wrong, just to be able to course correct.

You know, every company has kind of a number of S curves. It's always kind of being on the lookout for the next S curve and restarting the process again.

Thanks, Antoine. So now that we're all aligned on what strategy is and the strategy method, Andrea, could you take us through the roles that you see gen A I playing in that strategy method? And how can A I help strategy leaders make some of those hard to reverse choices?

We decided to invest some time in really bringing to life what it means to do strategy without A I, because this allows, uh, I I to to really bring to life the different roles.

The first five that we are seeing of how generative A I and A I in general can, uh, be, uh, a of assistance to to the strategy leaders out there in their tasks of going through the strategy method that we have seen. We see the researcher, uh, the interpreter, the thought partner, and the simulator, and then communicator as these five emerging roles. As, yeah, as we as Antoine was highlighting, there's different moments in the in the life of a strategy where you just need the information.

And getting that information is not just about data, it's about, uh, getting it right, getting it often, frequently, for example, when during the execution, you need to confirm your hypothesis over time and that that the underpinnings are there. You need to aggregate them. There's a there's a job for for systems to to bring that late data, uh, to you, uh, and to us as as strategists in general. But the interpreter instead comes the the the step after and says, Okay, I I do have data.

How do I think about how use, how useful that data can be in understanding markets, in understanding competitors, uh, or growth of spaces or other other aspects of the um of the both market and the company? The thought partner we'll see is, uh, is a different role. It's not about researching or data or putting it together, but it's really about providing a source of the biasing, a second opinion, but even a source of ideas to the strategies through, uh, what we have, uh, most of us have experimented with the chatbot, but applied in context.

We should consider also the ability of A I of connecting dots, uh, across scenarios, and, uh, almost play games. I mean, over the last, uh, um, decade, we have, of course, seen the emergence of AIs that can play chess or go, pro popular games, but increasingly more, especially with the addition of generative A I, which is able to understand and and deal with concepts that are not procedural like, um, like the more traditional games. Uh, we see with the emergence of the of the uh role of a simulator, so the ability to playing those games across, uh, uh, different market reactions.

And finally, last but not least, the more understood and simpler role, but still valuable of communicator, uh, drafting, uh, drafting narrative, uh, from, uh, from a few, uh uh, elements into longer narratives that can be helpful during the mobilization, or even cascading the strategy at different organizational levels, uh, with the with the right, uh, with the right medium. For sure, the researcher and interpreter are crucial in the phase where strategy is being designed, uh, as well as when it's monitored, because knowing about the market, getting the data and interpreting is crucial.

Uh, but, uh, but the the thought partner can be very helpful as, uh, the decision making, the the thinking of the strategy is the focus, uh, of the leaders of the organization. But also when that strategy is converted into initiatives, are there blind spots? Are there risks not being considered? Uh, I would not leave the last word to the A I. We'll talk about it, but that can be an extremely interesting source of, uh, of insights.

Similarly, simulator and and communicator have a role in different moments of the strategy, uh, notably the communicator, uh, for sure, when in that conjunction between the strategy having been decided and the mobilization, so that we, you know, the organization can find multiple ways to engage different stakeholders within and without.

So do you recommend using or developing different types of A I or different assistance or agents for each of these different roles and types of tasks, or is there sort of one all-encompassing platform that one can use that has all these capabilities?

In general terms, I would expect that at the moment it's, uh, it's more efficient and more effective to to look at different use cases or families of use cases, uh, individually. There are, uh, some of these tools are are being available or being researched or are easier to sp to to spin up with more specificity and and focus.

Thank you. So you might use a different agent or capability for each one of these different roles, then. And so Bruce, can you take us through some of the practical examples of the use cases we just discussed? Maybe what using A I to help developing corporate strategy looks like in the real world.

So we it's a bit of a scenario, uh, which is primarily related to the strategy design portion. And, uh, the scenario is the following. Um, we we consider European player financial services, which is, um, eager to expand in Asia or in emerging markets, and in this case, in in Vietnam. And um, the strategy team or the organization obviously will need to go through the typical strategy questions, which are understanding the market, what are the market dynamics, what is the competitive landscape, uh, is there any opportunity to acquire in order to enter the market faster and to grow faster, uh, but also getting a view of how to mitigate risk.

And um, genai at today actually gives us some um some help. So the first tool is going to be on to what extent genai or a i today help us identify faster potential targets, what kind of outcome we get from the tool or from the model, what can we do with it? Second is to what extent, um, a I and generative a I can give us some keys to better understand or actually understand much faster, uh, market dynamics, which will be the source of the initial strategy or potentially, uh, competitive advantage moving forward.

And then, uh, how can we leverage A I as a thought partner as we're going through all this in understanding the risk and what we should bear top of mind? So this first part, uh, as I mentioned, those three um steps are actually more related to the design, the strategy portion of what, uh, Antoine presented earlier. Looking for a target, identifying, uncovering opportunities of companies operating in a market that potentially could become, um, a acquisition target for us is related to assessing the market and re and and related to the explore phase.

And this is the initial research, uh, step that we need to go through.

Thank you, Bruce. So, Andrea, what kinds of tools are available to actually execute this?

The ability of these type of tools is to go through a lot of material related to companies, be it their own annual reports when those are available either publicly or in in databases, uh, uh, news about the company, uh, if they have patents, if they have, uh, their own marketing publications. All reading is, of course, a way to anthropomorph- anthropomorphize it, but, of course, the AI is ingesting and thinking about this in a different way. By by going through all these data points, the AI is able to create a graph of all these companies.

And and then with the prompt, is able to convert that large graph that has in that looks at companies globally into a filtered version of that. And understand, uh, which ones of these companies could be a good, uh, target.

So, in this case, Bruce, how comprehensive and accurate would that target list of companies be that's been identified by the tool?

It's interesting that that list is, overall, a good list for who knows the market. So I was putting my hat on knowing the consumer finance and the banking market in Vietnam. But then it gets into some sorts of complexity. Sometimes the the companies are actually, um, referring to a different parent company. Um, the tool does not necessarily tell you to what extent there is option to buy a stake in those companies. Um, and then, um, the tool, obviously, in this case, would require to do a little bit more work on each of them to understand what's the the financial health of those companies.

So, uh, great improvement in terms of speed, quite reliable reliable view in terms of list, still a lot of work to do to make sure that you come up or that we come up with the top three or four that really makes sense from, uh, all angles, as we would do in target identification and and screening.

So M and A targets can be a very sensitive subject. How do you protect your strategic intentions if you're using one of these tools? Because my understanding is they can also ingest your searches to get smarter. So do you recommend building your own proprietary gen AI tool, or are there ways to use generative A I to maintain that confidentiality?

I'll I'll I'll start. I mean, this is we did not go into the, let's say, infrastructure and security side of things. Uh looking to to start from, let's say, the use case, but that is a very valid concern. And, uh, every organization when using, uh, um, when using tools from from anything that is on the cloud that leaves their own secure premises, of course, as a number of, uh, uh, policies and and focus on security that, uh, that is there for a reason.

So, um, uh, we would encourage everybody that wants to use these tools to consider the trade-offs between using, you know, uh, public tools and that confidentiality. And most likely, most organizations already have in place, uh, uh, strict policies on what is allowed and what is not. One extra thing I'll say is that it is possible to build guardrails and secure environments for these type of tools to be used so that that security is and confidentiality is enforced.

Thank you. So let's discuss another example of AI in action in formulating corporate strategy.

So we we looked at the companies that potentially we could consider as targets. And as we're doing this, as part of the typical strategy, uh, formulation or design, we try to understand better the market. And understanding the market means understanding the players, understanding whether they are clusters of players, understanding, uh, all sorts of, um, uh, criteria related to which one is the most innovative or a little bit less innovative, whether they are experimentation in different business models and so on.

And then, um, a number of tools, um, um, AI agent, uh, driven, provide, uh, an initial view of what those clusters or what those dynamics are.

And Andrea, can you tell us a little bit more about the AI agents themselves that are driving those tools?

Yeah, thanks. This is this is a very exciting, uh, opportunity. What I'll, uh, I'll try to explain a a bit more what we mean by using AI agents in these type of things. Uh, understanding and interpreting data is not easy. It's, uh, something that, of course, we have seen to a certain extent language models can do. Uh, but what we have found, what we have seen more successful in the in the applications that that we have we have tested and seen is when we combine our way of approaching problems, uh, with the capabilities of language models.

What do I mean by that? Is when we apply frameworks, processes, workflows that we know work when they are in the hands of of, uh, expert and talented individuals with, uh, with the language models. And we have found that chatbot do not do not are not sufficient for that. What really brings this approach in the right place is splitting the work, and when we can split the work into tasks that can be assigned to, uh, to different agents.

Now there's been a lot of conversation about agents, and there will be more, uh, in in the market on on many other general AI applications, uh, in customer service or in other type of more operational settings. But in strategy as well, uh, we, uh, we have seen that agents can be, uh, uh, brought into into the picture. There can be the, uh, researcher agent, the the expert, the critic that checks that the analysis was appropriate and and checks against sources, avoiding, you know, when there are hallucinations that they get come come through.

So there's a number of these of these, um, agents that can be brought in the same place and and the tasks can be assigned to. And these tasks can be are the ones that allow for this competitive, for example, landscape or market review to to to be done from understanding which the competitors are, what are the product segments, what are the sentiment of the public on a specific or consumers on a specific, on specific products and and offering and and many more.

And the result is is very helpful. It's not, uh, in itself, uh sufficient, but it's definitely a great, uh, great start.

And how hard is it to use generative AI in the way you describe? Is this something that you can take an off-the-shelf option to do, or does it take a bit more custom programming work, for example?

I I I it's, uh, I think the straight answer is that, uh, in these first two examples, we are sharing, uh, um, ways to apply GENI to strategy that is rather complex. It's not really off the shelf. It does not mean it's not, it's not achievable for organizations that are committed and and and motivated.

So do you have to be a really big company with a massive data set for this to actually work, or can all companies do it if they're committed and motivated?

Look, I think the key thing to to consider, I think, is that it's actually not, uh, for only for large companies. Actually, it starts from the strategist, the individual person. There are a number of applications that, uh, are available for every strategic thinker out there. And of course, as organization size and complexity grows, uh, the opportunity of institutionalizing certain capabilities, uh, investing in systems, uh, grows, and the return expected from those, uh, as well grows.

Understood. And will gen ai allow companies to be more agile in terms of continuously evolving their strategy, or is this something that's more used at the early stages of strategy formulation?

Um, the answer is, uh, we believe yes. Uh, as these capabilities become more widespread and available, and we'll see some of them are already available, we expect that the speed and the ability to apply, uh, a so a a layer of automation on top of the typical analysis and investigations and reflections the strategists have to make will allow for more agility, um, uh, quicker reaction times, and also deeper insights, uh, in certain cases.

So are there any risks that we haven't talked about involved in using gen AI in the strategy process? Like what can go wrong in the example that we've been discussing of this European financial services player looking to expand into Vietnam?

So we have experts who understand banking, who understand Vietnam, who understand the markets, and then we looked at what was generated. Now, along the way, there are a couple of things that go wrong. Um, in this particular model, we are trying to understand, uh, who potentially would have developed an innovative business model in a market which is in the making and where the key differentiating factor from a strategy perspective will be for the first player that managed to import or develop a business model that's going to be fundamentally different from the rest of the pack and take advantage of the very large Vietnamese market.

And when we ask this question, and when we get the tool to answer this question, that's where we we see certain confusion. That probably something to do. And Andrea can add to this with, um, what is in the statements or in the press, or how it's worded, the word innovation. It's not necessarily related to a new business model, it's not necessarily related to a new target segment, which is targeted in this case, right? So seventy eighty percent good view, but same thing as what we said for the first example.

Absolutely important to get local expertise, understand, put things in perspective, and make sense of, in this case, the the metrics that, um, that the the model come up with.

Yeah, and I think Bruce, you were mentioning, this is just one example. How do you look at banking? One of the way to look at players and rank them is through their, uh, non-performing loan ratio. This is one of the industry metrics. Of course, there are others from from the assets or in the balance sheet and so on. There are many others. What we wanted to say is that these type of workflows can be engineered so that they can provide the strategy for with a starting point.

Also across a number of metrics, as ensuring that the thinking about strategy is complemented from a a good fact base. There are publicly available commercial, uh, general AI chatbots, of course, and very many organizations are either considering or have already introduced their own version internally that meets all those security and confidentiality requirements we were mentioning earlier. Uh, but but even to the, you know, one shot, one, uh, one person strategist out there, these tools are, um, are available and can be a great, uh, source of what we have called the thought partner role.

Uh, in terms of brainstorming, uh, idea generation, challenging hypothesis, and so on. Of course, and what we this is what we are seeing organization do, uh, when they build those, uh, um, uh, those capabilities internally. The more proprietary data, the more these tools are integrated with the the ability to search the web or specific sources internally or externally, the better, the more specific ideas and suggestion come. So that is that is important. And and I think we mentioned the point on, you know, check with your organization what is appropriate and not that is that is, uh, um, it's it's better to be secure in these cases.

What about hallucination? That's something that, you know, we frequently hear about. And although gen AI is maturing quickly, is this still a concern?

Yeah, no, that is a very important point, and that's why we we we mentioned earlier. Uh, it's important to have the strategies in the loop, right? The the the person, um, but including these commercial chatbots, but but but also the ones that we are seeing organizations bring to life in their in their domains. Um, increasingly more, there's, uh, an automated source verification, so many claims, uh, are linked back to the source so that the person can go and check.

But even when that does not happen, especially for s precise numbers, uh, uh, stories, uh, uh, and, uh, that that are specific, it is important to cross-reference.

Understood. So it's important to have a human strategist in the loop. Also, how well can the AI itself spot potential risks in the results that it gives you?

So, uh, so same same analysis as what we did before, right? You know, at first, it's not necessarily impressive, because if you ask what what are the risk, um, in entering any markets, you will probably get something related to, uh, regulation, something related to political and economic risk, and so on and so forth, until financial risk, right? Now, in this case, um, the tool provides us interesting input, but the problem is you don't know that they're interesting.

I I I take an example, um, regulatory risk, the foreign, uh, ownership limit is the most important one. I mean, of of course, they are regulation changing, but that's that's typically what could make, uh, or break an entry strategy, uh, in the context of, in this case, Vietnam and banking. So the tool does not necessarily tell you that, but someone who is closer to the market will say, yes, actually, that's true. On that particular regulation point, uh, let's let's have a closer look at the the ownership limit.

Um, and then the same thing applies, for example, on acquiring local talent, uh, and retaining local talent. So Vietnam is known to have a lot of great digital talent, attracting and retaining them becomes very important. Any, uh, any financial institutions, and in this case, any new, um, um, institution entering the market, we need to be very, very close to this and find the keys to make that happen very quickly. There again, that requires a lot more analysis to understand exactly what the tool is giving us.

So let's change gears a bit and turn to what strategists should do differently starting today, after listening to our podcast. Antoine?

So there are really kind of three main takeaways and and and implications today. The first one is start using gen AI today, even yesterday, right? Start using the available tools, start understanding what's available out there, kind of third-party tools, understand the mean limitations, and because you're kind of strategists, you you actually know kind of and be and are able to assess how good those tools are. But, uh, that's really kind of, even though some of them may not be kind of ready and mature yet.

It's the first imperative is is to really kind of start using. The second one is, you know, we see kind of gen AI and AI tools as kind of automating a lot of task and analysis that, you know, we used to spend a lot of time on. On what to do with that free time, so to speak. The imperative here is is that creativity and innovation will come at a premium. And so you should, uh, think about how you can become more innovative and and creative by kind of linking to more kind of stakeholders around you by exploring different analogies with other industries and and so forth.

So really kind of trying to think through how you're going to differentiate yourself and how you really kind of boost this creativity and innovation in in your role. And the last one is is strategists will need to actually become gen AI architects to a certain extent. For the handful of of use cases that are most relevant for you and most valuable, you may want to build your own tools. And to do that, you really need to understand how it works under the hood, try to understand the limitation, the current state of the of the technology.

Which data set do you have that will become really important and how you can invest behind it? What's the technology stack that you'll need to put in place? Where's the the cyber risk and then how to manage it and so on, right? It's kind of that's kind of the third imperative, and that's a role that we see is gonna become increasingly important for strategists in a way to to be able to to really leverage the power of of those tools, uh, to come.

Super, super helpful, thank you. So what are the implications for strategy teams? Do you need to hire a data scientist now for your strategy group? And what does it mean for how teams will be organized when these tools are available, especially with relation to the roles you described earlier and where gen AI can be involved?

Right, it's a little bit connected, Sean, to the initial question you had on on how, um, organizations become more agile, uh, from a strategy perspective. Uh, so look, um, let let's try to go through two or three implications. So, first of all, we understand the limitations of the tools, but we understand, you know, what they bring. And what they bring is the ability to formulate hypotheses a lot more quickly. So it changes number one, the um, the time that you have to actually s uh, discuss, develop those hypotheses.

And, you know, some of the roles related to the thought partner or the simulator also help you validate, test, prove, disprove the hypotheses that you have. And, uh, that's number one. Two connected to number one is the team has a lot more time to be to be creative, to think out of the box. And the box now being the way AI is actually framing some of the strategy problems, right? Which is which is very important. Number three, and that's what, um, Antoine was alluding to.

It comes with responsibilities to understand what the tool is doing. From my own experience, what was very much missing for me at the beginning, when I when we started leveraging and experimenting, is I didn't know what the model was trained for. And therefore, uh, when you get for the first time, um, uh, an outcome or an initial set of analysis, you turn to the data scientist and you say, well, what was behind what I'm getting? And then the data scientist can tell you, well, that's what it was.

Now we could go faster by doing this or using another tool or having a different technique, which, as a strategy, I didn't know. So that's that's very much that capability development approach that needs to be taken in strategy, division and and organizations. And maybe the last point is related to that part on agility and and simulation. I mean, we used to talk about scenario envisioning and all these tools that were trying for us to master uncertainty a little bit more.

Well, AI helped us do that even more, um, by also taking advantage of the data that we have internally. And and just one last quick illustration. I don't want to be too long, but you could think on our M and A example of a case where you start asking the model to what extent you can be successful at integrating a particular target based on the historical data that you have and based on what you've done. We're not there yet.

I mean, at least we we haven't presented that yet, but that's, um, these are big changes for a strategy team.

Fascinating, thank you, Bruce. And before we close, I'd like to just see if there was anything that really surprised or excited you as you've worked on using AI for strategy, and as you developed your recent article. Maybe Antoine, let's start with you.

Sure, I'll I'll I'll share a simple example, but, uh, that that works surprisingly well. One of our colleagues, uh, created a small agent, and, uh, we have something that we call the ten time list test of strategy, and it's a way to to assess a strategy. And we have kind of a vertical document that that goes through kind of the logic of it. It's about forty pages. He uploaded that, or actually the link to that document, and then uploaded an actual strategy of a company.

And asked the the agent to actually assess it against the ten test. And, you know, what are the strength and what are the areas where it could be strengthened? It works surprisingly well. And, you know, it it took him probably no more than ten minutes to do that, and it's repeatable, right? And so I think there's, um, we present it today, especially the first two kind of more complex use cases, but there are some simpler use cases that can add a lot of value, uh, in the short term.

And that's kind of a a a good surprise that that I've discovered recently.

Thanks, Antoine. We always like good surprises. Andrea?

My kind of my last kind of vocal thought that I would leave would would want to bring to the table. It's a bit less about the surprise, but it's really about the shift that I'm seeing of, uh, uh, really spending more time in the architecting and thinking and creating also, um, the opportunity for these, uh, uh, systems, for example, the one on agents that we described, to tap into proprietary and, you know, internal data or sources of data that only is one organization has.

There is a big shift versus doing the analysis itself. And I think strategy teams out there will see that if they use the same tools, you know, that that are being, they're being in a way democratized compared to others. Uh, they may not get that, you know, special insights that they want. But if they start linking, uh, uh, data and, uh, or talent that they only have, then all of a sudden they can, they can establish, uh, you know, their their own version of AI that is, uh, that is, you know, architected for success.

And I think the last the the as the time, uh, moves from doing actually to architecting, it actually also moves, and I think we were mentioning that, into the opportunity to move to action. And, uh, and I think that is one of the big, uh, dividends that will come from this, uh, revolution, uh, of moving, uh, instead of spend spending more time converting this, uh, the insights into conviction.

Thank you, Andrea. And Bruce, you get the last word.

So, you know, we hear often that AI will provide the same level of data and analysis to everyone, and therefore where the differentiation will be. And is there even a chance to differentiate? Uh, I think, you know, we we try, uh, you know, in a humble way today to to show that, yes, there is even more than before an opportunity to be differentiated in what you do with what you're getting from AI, uh, and how far you wanna go in terms of using the insights, the data to develop, uh, recommendations or action, um, you know, actionable or initiatives that will be fundamentally different from competition.

And that that is great and very exciting for for a strategist.

Awesome. Really appreciate the time, Bruce, Antoine, Andrea. It's been a great conversation. Thanks again.

Thank you very much.

Thank you.

And thank you to all of our listeners for joining us today. We really hope you enjoyed the conversation. And if you'd like to read more of our latest thinking on the emerging role AI is playing in the strategy room, we've included a link in the show notes to our recent McKinsey dot com article titled How AI is Transforming Strategy Development. As always, we welcome your feedback and ideas for future podcasts. You can email us at itsr at mckinsey dot com, which stands for Inside the Strategy Room.

You can also share your ratings and reviews on any podcast player with many thanks to all who have already done so. We really appreciate the comments and feedback we've received and encourage you to please keep them coming. And if you enjoyed this episode and you'd like to subscribe, you can easily follow our weekly series on any podcast player, where you can also access our entire library of more than 230 episodes. We also offer an Inside the Strategy Room podcast collection page available at mckinsey dot com slash itsr, where you can easily browse our prior podcasts across nine major themes and also access written transcripts of all those conversations.

And finally, if you'd like to automatically receive our latest publications and insights, we encourage you to sign up for email alerts on our insights page at mckinsey dot com slash scf, or connect with us on LinkedIn at the McKinsey Strategy and Corporate Finance Practice page. Thanks again for listening. We look forward to having you join us again next week inside the strategy room.