Because you're a subscriber to this Bloomberg podcast, we thought you'd be interested in a sponsored podcast called The Ceo Radar, produced by BCG and Bloomberg Media Studios. It analyzes more than forty eight hundred Q three earnings calls worldwide to assess what topics merit a CEO's time and attention. Here's a recent episode when the company is investing in these AI pilots. Are they spending too long on them? Are they giving up too quickly on them? Or is it
just right? I wish it was just right. It's both.
Actually they're closing some of them too soon and some of them they're just stinking on for way too long.
But I think that focus on value creation and particularly where you can build distinctive value creations, that is the critical way to be thinking about where and how to apply AI right now.
The Ceo Radar reviewed earnings calls at forty eight hundred companies worldwide to see what's on the agendas of both CEOs and the market as a whole. I'm Edward Adams of Bloomberg Media Studios on this episode. I'm joined by BCG Global Chair Rich Lesser and Lad Lukic BCG Global Leader Tech and Digital Advantage. Rich and Blad Welcome to the podcast.
Great to be here, ed Nice to be here.
In Q three, tariffs remain the top one or two topics that were most mentioned during earning's calls, both by CEOs and by analysts, but the number of mentions went down from Q two.
There is a shift that I thought was really interesting.
And when we did this a quarter ago, all of the discussions are about policy related topics that will happen. What will happen. Now it's about, Okay, what are we going to do? There is still policy uncertainty. It's not that things won't shift, but now I think everybody realizes substantial tariffs versus anything we've dealt with historically in decades are now most likely here to stay, certainly as it relates to the US, and now we have to navigate it.
And I thought the interesting shift in the types of questions and conversations that CEOs and ANAUS were having was really quite reflective of that shift.
And along those lines, they were talking far more about growth topics we saw this year, particularly in AI, a topic we didn't think necessarily could be mentioned any more often. Than it had been.
What we observe in our research is a bell curve of behaviors where you've got about five percent of CEOs that are really at the cutting edge about how to embed AI into their core business, reshape how they operate and invent new business models drive the business. You've got another thirty or forty percent of CEOs that are well on that journey, and you've got about sixty percent of CEOs that are really struggling to get on that curve,
often either six zeros six percent. Right, No, it's a hard it's a hard transition to make because on the one hand, you have to change your mindset about how you're going to translate AI into impact at scale, not just do interesting pilots, and then you have to change your mindset to raise Yes, there is an enormous amount of interesting and important tech in here, but the hardest part of that change is often the human part of that change, how you embedded in and processes, leadership skills,
and companies just find that journey to be a bit hard, and so every year webs are more and more moving toward the more advanced and their ability to do it, but it's still a substantial percentage of CEOs that are struggling on that journey. But for the ones at the top of the list, I think they're thinking about it in very sophisticated ways. How do you pick fewer things rather than more and really deliver them?
LAD, you're oftentimes down in the trenches with these companies, not necessarily always at the CEO level. Are you seeing any disconnect between what rich is reporting at the top and what you're seeing when you're talking to people two or three levels down.
Many times you could have the direction lined up by the CEO and fully aligned.
By the executives.
But if you're running HR and now, you need to hire igentic engineers that know how to wire these workflows together. You need to change job descriptions, you need to introduce new titles, and many times they they don't know how to do that. They haven't done it before.
Richard told us that they are About sixty percent of companies are still struggling to get out of the incubation phase. Right with AI, If we're on the outside of a company, how do we recognize a company has moved beyond and it's begun to really get some ROI from from its implementation of that.
I mean, I would say focused on the bottom line results, like the really good ones immediately. However, we cut the data and when we look at it, the ones that have really leaned into AI are growing faster the numbers are there.
Are you finding that that CEOs are willing to talk about that on earnings calls to sort of be public about yes, we're improving in AI is one of the reasons why.
I think some are starting. One of the leading insurers talked at length about underwriting and how they've really thought about underwriting differently to be able to go faster and more effectively. And in commercial underwritings it's often a speed game. The ability to get to a proposal very soon for a client is critical to whether you're going to win the win the bid or not. And so a leading
mining company has talked about how it's automated. A huge part of Western Australia runs it out of a control center with TED people uses predictive and enerve AI to be able to make much smarter decisions about what time of data fill different kinds of ships, how to run the whole process. Loreal has talked about beauty Genius. It's desired to create a more agentic beauty platform that people can use online and get recommendations, but also fulfilled orders
and do things. And how they just did a major rollout here in the US and some other places this year that it's a learning mode, but those are real businesses, not just a small pilots.
It's a test to build off of that. There is need to communicate for several reasons. Son, you need to get new talent. The talent is now looking at am I joining a company that will be a winner in the space, and will I get all the tools in my hands to be successful. The buyers are looking at their suppliers and saying are you going to be cutting edge?
And are you going to be pushing on the innovation curve and on the cost curve fast enough to remain so the CEOs need to signal both to their employees and partners in the broader ecosystem that they're on top of this.
I want to find out whether or not you're seeing AI as a job creator or a job killer, at least at this stage of the process. Lad, you're on the ground, what are you seeing?
Yeah?
Counter to like what we might hear in the media, right. I'm actively engaged with probably twenty engagements right now. Not a single one of them has a thesis on reducing the number of labor. It is about growth opportunities and growing the business. Now that is in the short term. Does that mean it doesn't have longer term implications for sure?
Right?
But where I'm seeing the winning convoys when you get the efficiency and then translate it into new value propositions to customers, right and grow the business. So that's the conbo that seems to be winning so far.
I agree with that, but I would add the two things. One is, we have seen examples I mentioned earlier, the Australian example where a a mining company can control an enormous amount of land area autonomously have a control center with a very small number of people in it. I mean obviously that has labor displacement from the way they would have used to.
Formal we would have operated.
And then the second thing I'd say is there is so much capital going into AI and data centers now it's hard to see how that pays out in the long term if labor displacement isn't a part of that.
I'm optim is that it's going to generally result in net positive number of jobs, but they're not going to be the same jobs. People will have to be reskilled and moving to new into new fields, et cetera. And the ones that embrace this are going to be more relevant than ever. If you embrace this and do it, you'd get a superpower and you can be more relevant.
But at the beginning of the year, there's a lot of talk about agentic AI AI being able to accomplish tasks to a degree on its own right, to be able to book my flight to La get me a hotel, get me a dinner reservation without me having to get involved with it, as just one example. Are you still seeing that emphasis on agentic at the as we begin to close out the year, or has it has it shifted somewhat?
For sure, it's it's still there, and it's in two specific flavors. One is articulating, are of the possible? Oh boy, wouldn't it be cool if we could combine all of these things together? Right, the reality is we're just not there to pull it off. So where we're seeing agentic being deployed is in a specific workflow end to end, but then with a human to act on the decision.
What we're not seeing is and what a lot of folks are hoping for, is agents talking to agents and like stringing together a bunch of processes because you still have accuracies in the eighty to ninety percent range. So that means if I string together ten steps and each one of them is eighty percent accurate, at the end of it, I'm at less than fifty percent accuracy. So you do need a human in between that is interpreting and fine tuning the agents within.
Those specific workflows.
That's where we are, and for us to enable the next level, we'll need better data, we'll need better infrastructure. It was a fantastic catalyst to actually surface a lot of the basic needs that the companies need to rewire and rethink on their textag. So it's there as a catalyst and the promise of what it was can happen, and companies are leaning in and are committed to that motion for the years to come.
Which I don't need to tell you.
The valuations of the Magnificent seven and other AI related stocks are scott high. If those valuations were to tank, much as we saw the valuations of dot com companies tank during the dot com bubble. Do you think that companies will continue to invest in their own AI initiatives or do you think that may have a knock on effect in terms of how the worldwide corporate environment looks at AI.
I think it will have very little effect on the real economy. It'll obviously have a massive effect on the stock market, as they represent a high share of total stock market valuations.
If that's an aarrea where will occur.
But I actually think the more companies see the kinds of impacts they can drive, and the more they can point to examples even if it's of competitors or in other industries that they can rise, they need to take action, that will be what drives momentum. I actually think that companies are realizing you can fundamentally operate differently with these technologies. Even if the market is overvaluated, there's no question AI's capabilities will continue to grow.
Rich Are you finding that companies are investing in AI in the right kind of way?
Obviously a subset are, but I'd say too many more than half are still at the stage where they're either investing may need to do pilots and not thinking about scaling or Equally importantly, they're doing the kinds of things that are important to stay competitive, to be more productive all of the many AI tools that come along from tech players and others, but insufficiently investing to what will really lead to competitive advantage, which is where you're either
reshaping entire workflows or functions, where you're building new business models.
So lad when a company is investing in these AI pilots, are they spending too long on them? Are they giving up too quickly on them? Or is it just right? I wish it was just right. It's both.
Actually they're closing some of them too soon, and it's some of them they're just sticking on for way too long. We had a client that for three months was running a pilot and in an area that had so much value right and they were evaluating a tool that was available three months ago, and at the end of it was lukewarm results and they said, Okay, we're done. Well, there's so many new tools that got released in the last three months that can get the performance to the
next level, but they gave up. So if you know that there is a lot of value in that workflow, sweat it out, stick with it, don't abandon it too soon because the pace at which the new tools are coming is so high. Right on the flip side, there are a number of them that are seeing good, interesting results from the pilots, but it's in a process that's not a bottleneck, and even if you solve it from ten days to minutes doesn't change the overall outcome. You're
just increasing your cost. But they stick with it because the technologies, they can show cool demos and it's moving, but it ends up being a distraction. Stop stop those much sooner than kind of instead of creating zombies in the organization.
So apply AI to the core of your business operations.
Where there's business value in a very distinct and measurable way, as opposed to where you can do really interesting stuff that looks really cool. I mean not that we don't all like to have a little bit of cool. You want a little bit of pizaz in whatever you're doing. But I think that focus on value creation and particularly where you can build distinctive value creation that is the critical way to be thinking about where and how to apply AI right now.
Since we're almost at the end of twenty twenty five. One of the questions we want to ask you was what were the kinds of attributes that CEOs needed this year to be.
Effective rich resilience? Like this is another year that could be characteristic over this entire decade. But I just think this ability to take unexpected things and figure out how to navigate them and do it in a way that both anticipates as best you can, but then can respond quickly and adapt and then reimagine that set of attributes is turning out to be over and over again really critical in this world.
Are you finding that most CEOs have that skill set or is it something that any of them are still reaching for.
I think they've had to develop it a lot more. In fact, when I meet new CEOs that I do a lot of sessions with them, I always say, you have no idea how much this generation of CEOs is able to handle resilience. More than six years ago, pre COVID, it had been so steady for so long, I think people sort.
Of lost that muscle.
But now one thing after another, so I do think CEOs are generally more capable, but each shock requires something different, and the tear of challenges that we discussed on earlier podcasts this year I think required a new set of capabilities to be built. But people are more comfortable building new capabilities in this world, Blaed.
What do you think really good ones not?
As they were resilient, they were not shying away from asking some of the basic questions. I see too many CEOs that sometimes don't want to ask a basic question so they don't look silly in front of their technical staff or in front of the juniors. Right, and the really good ones went back to the basics. Okay, what problem are we solving for each customer? Where what will be the value created? So those the combination of resilience and going back to the basics was the winning combo this year.
What was the best thing to come out of this whole tariff chaos that we saw this year?
It forced people to step back and think, Okay, how do we make money in this business?
Where are we irrelevant? What is the value we're delivering?
And it really had them look at the business and I would argue for a lot of them to fall back in love with the business because there was a momentum they were just building off of.
Interesting are you seeing companies focus their attention?
I think this year pushed people to build geopolitical muscle that they probably needed to build in the long term anyway, but it really accelerated it. I mean often the people understood how to deal with tariffs and trade possees were three and four and five levels down understanding supply chain risks, not just in your own manufacturing, but deep into supply chains.
We're really not getting the attention that they deserved. And I think the sharp impact of tariffs, the increases that are so substantial versus the last eighty ninety years of history, I think that's caused people to build muscle that will served them well in the long term, but had not received as much attention in years prior to this one.
Rich and lad thanks for your insights today.
Great to be here, Nice to be here.
Those of you who would like to learn more about the CEO Radar can read the full report at Bloomberg dot com slash CEO Radar, and if you liked what you heard, we encourage you to subscribe on YouTube or your favorite podcast platform. I'm Edward Adams of Bloomberg Media Studios. Thanks for listening.