Many companies are investing heavily in artificial intelligence right now, hoping to improve both efficiency and innovation. But, as with any technology that sees widespread adoption, AI itself won’t be enough to build a long-term advantage over competitors, says Jay Barney, professor at the University of Utah’s Eccles School of Business. Yes, leaders need to deploy these new tools, especially those that use GenAI, to stay relevant. But they also need to think about how AI can be applied to their business’ differentiating competencies and offerings to truly add value. Barney is the coauthor, along with Martin Reeves of Boston Consulting Group, of the HBR article “AI Won’t Give You a New Sustainable Advantage.”
ALISON BEARD: Welcome to the HBR IdeaCast from Harvard Business Review. I’m Alison Beard.
It’s pretty basic business strategy. Find something that your company can do better than any others out there. Focus your investments and reap the rewards of competitive advantage. For anyone leading an organization right now, the big question is how generative artificial intelligence might change the game. Those who are quick to incorporate it into their business operations are certainly finding gains in terms of efficiency and even creativity, while those not using it yet are probably falling behind.
But today’s guest notes that early adoption of widely available new technology rarely yield sustained momentum. Instead, long-term success with gen AI will depend on applying it to existing advantages. He’s here to talk more about that.
Jay Barney is a professor at the University of Utah’s Eccles School of Business with Martin Reeves of Boston Consulting Group. He coauthored the HBR article AI Won’t Give You a New Sustainable Advantage. Hi, Jay. Welcome to the show.
JAY BARNEY: Hi, thanks for having me.
ALISON BEARD: I want to start with some of the ways that companies are currently using AI to gain a bit of advantage. What are you seeing as some of the most successful early tactics?
JAY BARNEY: Well, actually, it’s hard to keep up. Almost every day, some new wonderful thing that gen AI can do is described in the business press. We know everything from writing reports, creating new products, improving online customer service, creating new drugs, even composing new music.
The most obvious ones are efficiency-oriented. That is, we’re taking costs out and making processes that used to rely heavily on human beings. Now, less so. And there are some sectors where the expectation is at least that there will be substantial impact of gen AI on, for example, in the legal profession, a lot of the sort of boilerplate. Contracting work and other kinds of work that the lawyers historically have done is just perfect for gen AI.
Another area where it seems to be taking off very, very quickly is in the area of computer coding. Firms are using gen AI to right now produce the first draft of code, and then a human being will go through and then update and make it more elegant or whatever. But there’s every reason to believe that a lot of basic coding will be done by gen AI going forward. Right now, a lot of value being created most by cost reduction and increasing efficiency. The broader societal implications of that, of course, are pretty profound and deserve a lot of discussion.
ALISON BEARD: But why is that not enough to give competitive advantage over the long term for these first movers who are doing it really well, creating their own internal systems, even using it for more creative tasks?
JAY BARNEY: The problem of course is that gen AI is available to any company. Even though it creates value, even though it reduces costs, it reduces costs for anyone who applies it in sort of a generic kind of way. And thus, we don’t expect to see a sustained competitive advantage being there.
That doesn’t mean it’s impossible for gen AI when linked with other aspects of organization, that it will be able to generate competitive advantage in those settings.
ALISON BEARD: Many leaders would think that, “Okay. Well, the obvious next step, I can’t create any sort of advantage using the publicly available models, but can I build a proprietary model or layer my own information on top of an existing LLM? Why isn’t that enough?”
JAY BARNEY: First of all, building your own proprietary system, it’s unlikely if it’s just general-purpose artificial intelligence that a particular user will have the scale and expertise that some of the other firms that are out there that have been doing this for a long time will have.
It’s kind of deciding, “Gee, we want to get a competitive advantage from Word processing, so we’ll build our own Word processing system.” I’m sorry, you should probably shouldn’t be trying to be producing a general AI system for yourself.
But proprietary gen AI might be possible. It might be necessary, for example, if you have unusual data requirements or if the patterns you’re trying to recognize are slightly different than other patterns. The thing about proprietary software is that if one firm can develop proprietary software, then its competitors can also develop that proprietary software. And so, even that is unlikely to be a source of sustained advantage.
The bigger one that people actually talk about a lot is proprietary data. That is, we have data that no one else has. And so, our data is unique and thus, therefore, it can be a source of competitive advantage as we apply our data to analyze it with gen AI.
And again, that’s possible. We’re not denying that it’s possible, but it turns out that in order for proprietary data to be a source of advantage in this space, it also has to be unique in some sense.
So, for example, suppose that your company has developed a very long, very complex database that summarizes the information you collected about your customers, your suppliers and your employees for the last 15 years. That’s unique to you, no doubt about that.
But if one of your competitors has collected its own database about its own customers, its own suppliers and its own employees, and it’s done so in a way that is what we’re saying is functionally equivalent to your dataset, then while your dataset may be unique, the patterns that will be revealed in that data are not likely to be sufficiently different to be a source of competitive advantage.
Another thing people talk about, I have a good friend who’s a three-time entrepreneur and I was explaining to him the thesis of this paper. And he says, “But Jay, some of these companies have 10 billion data points. Isn’t that the source of competitive advantage compared to firm that says data has only 50 million or whatever the small number is?” And my response was, not so obvious actually. if the patterns that the software is looking for are revealed with 50 million just as well as they’re worth 10 billion, then 10 billion sounds like not that helpful, not that necessary.” In fact, as someone who used to teach statistics for a living, sampling theory suggests that we should actually be trying to figure out what’s the smallest dataset we need to have in order to generalize and to understand what patterns exist in a dataset.
So, in fact, there’s a lot of research going on right now in gen AI, trying to figure out what is the smallest dataset we can use in order to generate the patterns that we need to see. So, 10 billion sounds like a winner, but actually it may not be all that special.
ALISON BEARD: So, it sounds like you’re saying that gen AI is going to be the same asset for everyone, like Word processing-
JAY BARNEY: Pretty much.
ALISON BEARD: … or the internet. But isn’t there an advantage to being the company that’s experimenting with it first, finding out what it can do, or is it such just that the pace of development is so fast that the gap between the first mover and the fast follower is just minuscule?
JAY BARNEY: I mean, I’m not going to pretend that there aren’t possible examples of some first-mover advantages that last a little bit of time. One of the ironies, of course, of gen AI is that when you use gen AI to do something, the results of the analysis you’ve done then feed back into the dataset. And so, the gen AI becomes smarter after you’ve done your work.
So, in some sense, there’s a second-mover advantage because the second mover gets to see not only the generic data but also whatever data you added to the dataset. I’m not suggesting that firms need to or can or should delay and delay and not use gen AI. Obviously, you should be using gen AI pretty much everywhere you can as quickly as you can, especially since access to gen AI is actually not that expensive right now, at least.
So, that makes a lot of sense, but it’s unlikely that those are going to be sources of sustained advantage that just based on the gen AI. Now, we’ve talked about linking gen AI with other aspects of your organization. Now, that’s more likely to be a source of competitive sustained advantage rather than gen AI, per se.
ALISON BEARD: Okay. So, let’s talk about that. How do you ensure that the way you’re using gen AI is focused enough that it will propel you further ahead than your competitors because you’re leveraging it in a better way?
JAY BARNEY: You’re leveraging in a way that is unique to you, So, the general story here is to gain a sustained competitive advantage. You have to have resources or capabilities that are valuable, that are rare and costly to imitate.
So, what you have to do to really get competitive advantage out of gen AI is link it with some attribute of your firm, of a firm’s resources and capabilities that are in fact rare and costly to imitate.
In order for Amazon to run its business model, it has to have relationships with its suppliers and customers. And there’s millions of them and it’s deeply complicated. This is exactly the world where gen AI can be of great value. As you put your system through gen AI, it’s going to suggest all sorts of ways that you might be able to reduce costs and increase effectiveness.
But because you have a virtually unique business model, those insights are only valuable to you. They’re not valuable to anyone who doesn’t have that same business model that you have. Then your use of gen AI to improve the execution on your business model could be a source of sustained competitive advantage. But that’s not because of gen AI.
What gen AI does is it can tee up, suggest to you new exciting ways to improve your business processes and gain some competitive advantages, at least in the short term. But firms differ in their ability to take advantage of those. In the paper, we call it agility.
Some firms are able to take that new technology and apply it in an agile and a clever and creative way, in an agile way that their competitors can’t. Now, in the long run, will the competitors catch up? Probably. But in the short to medium term, that kind of agility can also be a source of competitive advantage. But again, it’s not the gen AI per se, it’s the gen AI plus the ability to change and manage change effectively.
It turns out that when you’re trying to discover these resources and capabilities that are rare and costly to imitate, the things you want to look for are resources and capabilities you control that are socially complex. They’re things like friendship, trust, relationships with suppliers, brand names which are a promise to consumers about what your brand stands for.
Another thing you want to look for is resource and capabilities that really take years to develop. You can’t go from where you are to where you want to be just by working hard. You have to go through a series of steps and process to get there. Those are also likely be sources of sustained competitive advantage.
And then finally, things where it’s unclear what steps you would take to create the advantage. So, I can know that my competitor’s advantage is fully dependent on their unique organizational culture and that I don’t have that organizational culture. But it’s not obvious what I should do to change my culture. I mean, that’s not an easy thing to do. It’s not a simple thing to say, “Hey, get rid of my old culture, build a new culture and move forward.”
So, those are the things you really want to look for are resources and capabilities that you control that are socially complex, that are path-dependent and that where the steps to create these new resources, capabilities are not obvious.
ALISON BEARD: It sounds a little bit like what you’re saying is that gen AI is going to help winners keep winning when I think a lot of us think of it as this great democratizing technology that allows the young startup to do something more quickly and more efficiently or more innovatively to compete with the big guys. How should I reconcile that? Are there still opportunities for the non-Amazons of the world to use this new tech to their advantage?
JAY BARNEY: Remember my example with agility. So, agility is a socially complex resource or capability. One firm is able to manage change more effectively than another. Large firms are not known for agility, they may have other advantages but agility may not be among them. And in that world, a smaller firm who is maybe more agile, may be able to take and apply AI in a way that our larger firm cannot do.
It is the case, however, all that said, on average, we expect that firms that already have sources of sustained competitive advantage embedded in them somehow and are able to then leverage those more effectively used with gen AI, those firms should do better than firms who don’t have anything special and then try to use gen AI to make them special.
Now, there are a lot of small firms out there, so it’s not just large firms are going to benefit from this. A lot of small firms, very innovative, many unique business models that gen AI could apply in that world to generate insights that aren’t available to other organizations. So, agility, unique business model, all those can link with gen AI and be potential sources of competitive advantage. But you got to start by knowing what your organization’s core competencies actually are and then figure out how to leverage those core competencies with gen AI. And that’s likely be a source of competitive edge.
ALISON BEARD: So, all the other work that you might do, the sort of cost-cutting, maybe brainstorming for marketing or new product development, all of that is periphery. What really needs to happen for any business is you need to figure out how to apply gen AI to the one thing that makes you special.
JAY BARNEY: I think all that other work that you described, efficiency, enhancing, making, we use the example in the paper of the toothbrush example, asking gen AI to generate a list of seven new kinds of toothbrushes. All that work is fine and you should use gen AI for all that, but that’s a source of competitive parity.
If you fail to do gen AI and to use gen AI to do that work, you can put yourself at competitive disadvantage. So, we’re not arguing against doing those things. We’re only arguing that they’re not likely be sources of sustained competitive advantage.
Now, it is the case that sometimes firms get so excited about all the cool things gen AI can do, that they fail to do the work of trying to figure out how to really leverage gen AI to generate a sustained competitive advantage.
I know in my world, in the academic world, I go to these conferences on gen AI. And firms are just talking about all the cool ways they pull out costs in various processes in their organizations. And I’m just sitting there just mystified on two dimensions.
First of all, they’re telling all their competitors how to imitate them. I just don’t understand that. So, sooner or later, the competitors will figure this out on their own. But why don’t we just delay that as much as possible and keep quiet about what we’re doing? So, that’s the first thing, it mystifies me.
The second is there’s no conversation about how we can leverage gen AI to be a source of sustained competitive advantage. Now, by the way, ironically, you could actually talk about that source of sustained competitive advantage because the kinds of resources and capabilities that you leverage with gen AI are not easy for your competitors to imitate. But just pulling costs out, that’s not hard and everyone will do that quickly.
There’s another point that needs to be made here. Managers need to temper their expectations about what it is they’re going to get out of gen AI. Suppose you do an analysis and discover from your analysis that the kinds of sources of sustained competitive advantage you might have are not easily or not obviously leveraged by the application of gen AI. That means gen AI is not likely to be a source of sustained competitive advantage for you.
That’s not a disaster because there are other ways to get sustained competitive advantage. It does say, “I probably don’t want to spend millions and millions and millions of dollars trying to build a proprietary system that’s not going to be a source of sustained competitive advantage for me. So, I got to temper my expectations which temper my investment.
ALISON BEARD: So, to leaders out there, what advice do you have in terms of how they need to be thinking about gen AI in the next six months and then over the long term, say five, 10 years?
JAY BARNEY: My first advice and I’m talking to CEOs here, because functional managers, people who are running large complex processes and need to simplify them and pull costs out, those people need to focus on creating value with AI even if it’s not a source of competitive advantage. CEOs, on the other hand, should be focused on understanding, nurturing, protecting, growing these sources of sustained competitive advantage And once you understand that, with a clear view, then you can then start leveraging your gen AI to help take advantage of whatever that is, whatever those the resources and capabilities are.
And in terms of six months, over the long term, I think in this technology world, six months is a hard prediction. Five to 10 years, a really hard prediction. I will say this. I am quite confident that three or four years from now, another new technology will come along. This happened many times in my life and career. I remember when PCs were first introduced, this will change everything. And in fact, it did change everything. But the great irony is because it changes everything, it is unlikely by itself to be a source of competitive advantage. So, I think it’s important to not get caught up in the enthusiasm of these new technologies to recognize, “Yeah, they can be very valuable. I had no doubt about that. And we should use them to create value.” But that sources of succinct competitive advantage usually found elsewhere.
ALISON BEARD: But given how fast this new technology is developing, how do you advise CEOs that you work with to stay current, to make sure that they’re on top of it, using it in the best way they can to sustain that competitive advantage that we’re talking about?
JAY BARNEY: If I had the resources, I would hire someone whose job it was to make sure that gen AI is being used efficiently as possible in my organization, from a cost-reduction efficiency point of view. I would go to those conferences. I listen to what those people are saying. I duplicate it whenever possible. And I would continue to focus in that area.
And then, don’t lose track of the core. Keep that most in your mind and then make sure that the technology people, gen AI people don’t lose track of that as well and actually continue to force that conversation.
The CEO and the CIO, it seems to me have to be in a constant conversation about are we engaging in technology investments in gen AI or other technology investments that put our core at risk, that sort of reduce the value that could be created by our core competencies? Or are we actually using gen AI in a way that actually allows us to leverage those more effectively? That’s the conversation that the CEO and the CIO need to be having constantly.
ALISON BEARD: And is there anything that lower-level managers should do, either on the frontlines or in the middle to ensure that the organization is focusing on how gen AI can be applied to the core competitive advantage?
JAY BARNEY: Yeah, super question, because you were talking earlier about the democratization of business associated with gen AI. And part of that democratization, to the extent that it exists, is that people at low levels of organization can apply gen AI just as well as people at the high levels of the organization.
This really goes back to this question of agility that I talked about earlier. One way to think about gen AI is that all gen AI should be approved by the IT department in a very systematic, bureaucratic way.
Another way to think about gen AI is we have 10,000 employees. That means there are probably 20,000 gen AI experiments a day, some of which had no value at all, but some of which might be quite valuable that no one at IT has thought of before. How do you marshal that enthusiasm and power and experimental knowledge? That’s really interesting problem. That’s agility, you see.
ALISON BEARD: So, how do you harness it though? All those experiments?
JAY BARNEY: you can do lots of simple things, like you can have gen AI competitions where you Say, “Hey, if you come up with a really cool application, gen AI, submit it to this group and we’ll evaluate it along with everyone else does it. And we’ll give $1,000 prize to everything that works.”
The challenge with gen AI and the opportunity with gen AI is that everyone becomes their own source of new information and new creativity. And so, we have to find ways culturally and through systems and other processes that you can put in place to try take advantage of that.
But I said earlier, agility is a potential source of sustained competitive advantage because not all firms can do this. Some firms are just locked in an old mindset and in a command-and-control mindset that is pretty inconsistent with gen AI.
ALISON BEARD: Well I appreciate you being with us today. Thanks Jay.
JAY BARNEY: Thanks for having me.
ALISON BEARD: That’s Jay Barney, professor at the University of Utah’s Eccles School of Business and co-author with Martin Reeves of the HBR article AI Won’t Give You a New Sustainable Advantage.
And we have more episodes and more podcasts to help you manage your team, your organization and your career. Find them at hbr.org/podcasts or search HBR in Apple Podcasts, Spotify or wherever you listen.
Thanks to our team, Senior Producer Mary Dooe, Associate Producer Hannah Bates, Audio Product Manager Ian Fox, and Senior Production Specialist Rob Eckhardt. And thanks to you for listening to the HBR IdeaCast. We’ll be back with a new episode on Tuesday. I’m Alison Beard.
