AI Isn't the Answer. It's the Start of the Question.

Speaker A: But what we've seen is this sort of popular and public explosion of, uh, very simple to use, very applicable AI tools. And I think that sort of pace of how that's kind of gone up the curve and into people's minds is what makes it particularly interesting. It's going to be unknown. It's going to require us to gird our loins to cope with different things that we haven't seen before.

Speaker B: Hello and welcome to the no Bull ideacast. I'm Becky Holland. Let's talk about AI again because while the headlines are full of hype, most organizations are quietly stuck. Only 7% of business leaders feel genuinely ready to harness AI and the rest, not sure quite where to start. Bumbling around, nervous about the risks, struggling to turn potential into progress, and maybe focusing on the tech at the expense of the impact. My guest today is James Easterbrook, partner at Morehouse Consulting, and we've been working together on a major piece of research exploring exactly that, the data and AI readiness gap and what to do about it. The strategy and the research behind this project was something that the team at BHP helped shape from the ground up, digging into not just what leaders say about AI, but how they really feel about it and what's holding them back from meaningful action. I'm so excited to share what we found out. We talked about the fear factor, the invisible blockers, and why AI needs to be seen not just as a tool, but as a strategic lens. Along the way we get into behavioral science, loss aversion and the cultural shifts that need to happen if AI is going to drive real and positive transformation. This episode is for anyone who's tired of surface level AI talk and wants to get under the skin of what's really, really going on. Good morning James, and welcome to the show.

Speaker A: Morning Becky. Thanks very much. Very happy to be talking about this incredibly exciting topic.

Speaker B: Exciting, but a bit complicated and um, a bit of a rabbit hole I think for a lot. But before we get into that, what's your journey?

Speaker A: So I started out actually in advertising. So my very first job was in France. I worked as part of my degree, uh, working for a publishing agency. And uh, it was very much about the creation of content for marketing purposes. That was really the crux of what they were doing. And then through a couple of hop skips and jumps, ended up back in the uk and then I ended up working in, um, in more traditional ad agencies and I, ah, really enjoyed that. Again, super exciting, interesting world, you know, doing crazy things. But I found that, um, a lot of the clients I was working with would say things like, I know it's not what we pay you for, but we're having a customer strategy day or we're having a session on our customer experience and we'd love you to join and give your thoughts. And I started to realize I was enjoying that stuff way more than I was cranking out the ads. So I ended up moving to PwC, uh, was there for two and a half years and then I moved eight and a half years ago to Morehouse.

Speaker B: Let's jump into the research and obviously this is, um, something we work together with you on. So, um, it's quite exciting to see how it's shaped out. But I want to start with the big picture. When we launched this research, we were really curious about why AI wasn't delivering on its promise inside large organizations. What do you think surprised you the most once we started to get the data in?

Speaker A: I think it's easy to assume agency when you look at large organizations or even, you know, small to medium sized organizations. It can be easy to think everybody else has got it nailed, everybody's got it sorted, they all understand these things, they're all got it figured out and then we're just, you know, waiting to see what they come up with and it'll be launched and then it'll all be good. And I think, you know, one of the biggest surprises was just the degree of fear, inertia, uh, lack of sort of grip that people had. Everybody knows it's going to be a huge opportunity, everybody knows that it's a challenge they've got to address. But, uh, people, I think, are really struggling to know where to start. And the scale of that sense, I think it was 7% said that they felt well prepared to tackle it. And I mean, that's really a tiny proportion. The scale of that gap was probably what surprised me the most. And I think that in one way, when you've got through the initial shock of it, when you think about what's actually involved in the scale of change and what people need to create kind of get their heads round in order to really get value, it's maybe less of a shock because it is incredibly complex. And of course it's not one thing. There's not this sort of homogenous one thing that AI is. And I think when people sort of start to unpack, um, it a bit and figure out what it means for their roles, their teams, their departments, that's definitely the realization that people are having.

Speaker B: No, I completely agree. And I think when you hold the mirror up to organizations, it's very easy for people to talk about things that they're doing with AI, a system that they've implemented, how they're using Copilot, all these different bits and pieces. But using AI to help with various tasks you do in an organization is not the same as embracing what AI can potentially do and really understanding, I guess, how it works in the, in the real world. Did the results from the research match what you're seeing on the ground in the work that you do with clients?

Speaker A: Yes, uh, in one sense, in that it reinforced some of the things that people are asking us for. So, you know, we get a lot of people saying, where do I begin? You know, what are the use cases that, you know, how could this help me? I know it's out there. And that obviously came out in the research that people felt that they weren't quite sure where to start. But I think we're so early. I mean, I know it feels like it's been around for a million years and it's the top topic of conversation on everybody's lips and every LinkedIn post, and it's all there. But in reality, we are still very early in this journey. If you think about organizations in certain fields, like finance, for example, actuarial science, or what goes on in insurance, they've been using machine learning and AI for years and years and years. It's been part of the mix in terms of how they assess risk and what they do for product development for a long time. But what we've seen is this sort of popular and public explosion, uh, of very simple to use very applicable AI tools. And I think that sort of pace of how that's kind of gone up the curve and into people's minds is what makes it particularly interesting and is why you're getting that awareness now. So it's not just very specific IT teams or technology teams or data teams that are talking about this. It's much broader in the organization. And I suppose that reflects the research reflects. What we're seeing is now the questions are coming from a much broader set of people within an organization.

Speaker B: Was there a moment in the interviews or the survey results when they were being played back to you, that made you just sit up and go, oh, yeah, that's the thing I was looking for, or that I was expecting to hear.

Speaker A: I mean, there was something that came out in one of the discussions which was around a really good question around, is AI a tool or is it, uh, fundamental to the strategy of the organization? And for me, that was quite a good distillation of the challenge of is this something that should sit in a technology team and be a product that they push out to the wider organization, or is it something that people need to think about underpinning everything they do? And I think that was a moment of clarity because the answer, of course, is not the same for everyone. It's different for different organizations. But. But if you think about AI as a tool, it's a very limiting way to come at it. And I suspect organizations that do that, uh, are going to struggle to take full advantage when they're not truly embedding it in all of their thinking.

Speaker B: You mentioned m that only 7% feel truly ready for AI, which obviously means 93% don't think they're ready. Why do you think that is? Why do you think only 7% feel truly ready?

Speaker A: So I think it's a number of reasons. The first one that, and I think this came out in terms of, particularly our interviews that we did where you got a bit of a feel for how people were tackling this is fear. It's impenetrable. There's so much to know, there's so much to get my head round. It's so sort of difficult, et cetera. You know, how are we going to take advantage? That definitely is a driver. I think also investment. Right. We are in incredibly volatile and challenging times economically. Any big steps, anything where you need to be thinking about new tools, about making large integrations, large transformations, it costs money. And so I think that's another factor of people are thinking, well, you know, how am I going to free up the time and the headspace and the money to invest in this? And I think for me, those are probably two of the critical drivers. It's sort of getting over that human factor of being almost the paralysis of inertia about, I just don't know where to begin. And I'm scared, genuinely. People are actually thinking, what might this mean for my job? What might this mean for my team? I think that that's part of it. But yeah, again, being able to produce a meaningful business case because there's so many unknowns, to be able to turn around and say, well, I can do this and it's going to generate X return. That's incredibly difficult right now. But still, it's still a very difficult task because the technology is moving on at pace and, you know, if you miss a week or two in the life cycle of some of these tools, it's making you, uh, know it's a huge difference in what's available and what you can achieve and what you can do.

Speaker B: What kind of skills do you think are missing? What's needed to be able to fully embrace what m. AI can can offer?

Speaker A: Clearly there are some core hard skills that are required. So the, the ability to understand, you know, to have your data in good order, to be able to kind of use your data, manipulate the data in the right ways, the ability to use the tools that are coming on stream, you know, uh, all of that is critical. And outside of that, probably where we're seeing some of the biggest skills gaps is on the softer side. It's the resilience to understand that this is driving change at a pace that people haven't seen before. It's the ability to assess and interpret the potential value and how you might fit these tools into your workflow or into the product development that you're doing for clients. Clients. That's definitely a skill which uh, I think is lacking. You know, a very practical level. For example, a lot of people's first experience of a basic LLM, probably something like copilot will be well I put in a question and yeah, the response wasn't really what I wanted. Right. And then they go away with a little bit disappointed and ah, uh, well it's overhyped and it's not there. But realistically the ability to engineer your prompts, to think about what you're asking this tool to, to do is actually it's a critical skill in itself. And you can see a massive difference between where that is done well, using best practice and then sort of self optimizing as you go versus uh, not being kind of sharply defined or clear enough. The other thing from a skills perspective is I think there's something around the ability to innovate and I think that is probably a shift that we need to see. More people need to be thinking about that as a core part of their work rather than it being someone else's job. And they'll tell me when there's something new that I need to implement.

Speaker B: So this, this project is a very good example of that where um, you know we have a um, an LLN that's been built with all of the data that's the qual, the quant, the customer surveys and everything. All, all of the analysis, all the feedback that came back from your consultants, all in one central repository. And I think that was a big factor in allowing us to create such a robust piece of research and to be able to analyze it really quickly and Easily. Um, and there is definitely an element of, um, you know, we need to practice what we, what we preach on projects like this. If we're saying we need to understand the power of prompt engineering to allow us to I guess take insights from data, for want of a better way of putting it, then if you don't put that into practice, then um, you're a fool. And I guess we also obviously used it in terms of the execution because uh, the final execution, in terms of the creative route that we've come up with again embraces AI. Ah. Um, and my goodness, that needed a lot of prompt engineering to make it work. But it's interesting to see how that's applicable across a whole different range of, of industries. And I think that's really key.

Speaker A: Yeah. And, but I think that point alone reinforces the fact that tech plus human beats tech or human any day. Right. The, the human in the loop point here is absolutely critical. If you're taking a blanket view that while this is just going to do all of those jobs and they'll never need to be a human and we won't need to do any of this, like, of course there are some areas where that's going to be, you know, that's going to be the case, vast majority. It's about enhancing the way people do their jobs of allowing you to come at them in different ways, allowing you to create new and additional value. I mean we see it with our own consultants just the time that they can spend doing more critical thinking, more value adding work goes up. Um, because the tasks like gathering research or distilling it or making sense of it are done, you know, done so much quicker.

Speaker B: Um, I wanted to touch as well on the split of opinion between the different leaders that we spoke to because it certainly into the interviews that I was involved in and we've seen it throughout the research, it was quite polarized. There were some that were very, I guess, embracing of AI as an adventure, if you like, something that could help them achieve more. But others, it really felt like they were not charging ahead, they were dragging their feet. There were not so many people in the middle. And I just wonder what your thoughts are in terms of why there is that polarization.

Speaker A: I think it comes down to a degree of skepticism about, or perhaps through perception or experience of their own organizations. So this all looks brilliant. These tools could be amazing for us. But God, we've got so many basic problems that have got nothing to do with AI that we need to fix. That's definitely something. So people's perception of uh, how much advantage is out there is colored by the fact that there's stuff that they just don't think works right now and they'd like to fix that before they worry about the shiny new thing. I think in, in more effective organizations, um, there's probably a sense of a glass floor. They think, okay, well this is, this is working and this is sort of sorted and this is okay. Therefore we can now build on and think about what might come and what might, what might we do that's new. And then I think you also have to put play through in each of those different archetypes of organization. Where are they at from a financial perspective? If an organization is struggling or it's been buffeted by this sort of economic headwinds, then the sense of having the time or the money to invest in new things, to innovate and to push things forward is more limited. Whereas in organizations that are doing better, there might be a little bit more freedom to say, yeah, I've been given a bit of headspace or a bit of time to try this out and to push something on. And I think that probably where we saw the distinction was probably reflective of, you know, where those organizations are in terms of their own sort of economic life cycle, but also where they are in terms of their maturity of ability to kind of innovate and take new things on board and take advantage of them.

Speaker B: Is there a danger, do you think, that too many organizations are just waiting for somebody else to go first? Uh, they don't take the risk.

Speaker A: That debate about everyone waiting for anyone else to go first does depend on where you are strategically. And that would always have been the case with anything. I think with AI particularly, it becomes a less viable option to just sort of wait for everything. I think you have to have some degree of experimentation, you have to have some degree of view of how you can use AI to optimize what you do. Otherwise you genuinely are going to get

Speaker B: left behind, I think. And I guess that leads, leads quite nicely into what I wanted to talk about next, which is really what behavioral science tells us about transformation failure. It reminds me of something that we talk about a lot uh, at ah, bhp, James, the idea of loss aversion. People fear losing what they've already got a lot more than they value the possible gain. And I think that came out quite strongly actually in the research. What do you think are some of those invisible blockers that hold transformation back?

Speaker A: You know, if it's seen as something which can do us out of a living, then nobody really wants to embrace it and address it. So I think that's one of the first points is do I feel like this is going to take away my job rather than enhance my job? That's clearly a barrier. And that's a human instinct, right? If it's, if it's fear, loss, aversion, as you say, that will stop people from embracing it. I think the other factors are people not feeling equipped, so people not feeling that they've got what they need personally, in terms of, uh, skills, abilities, knowledge, that sort of thing is definitely a barrier. There's also a fear, I think, of getting it wrong. So do I have the remit to do this? And I know some of the organizations we talk to, large organizations, they, you know, for example, where they had a central head office and then different business units or different regions where, you know, you might be talking about 50, 100,000 employees, maybe more, um, across the globe. Things were done at, uh, a central level and they didn't have the remit to do that locally. They were sort of banned or, you know, they didn't have the mandate to kind of experiment or push things forward. And that is another barrier. It's where the central control or grip, maybe from a regulatory perspective or maybe from a strategy perspective has sort of said, well, we're going to do it and this is the way it's going to be done and you're going to need to wait for this to happen, which doesn't allow, you know, the thousand flowers blooming of different teams around different parts of the business sort of cracking on and using tools and doing stuff. And, you know, we've seen lots of organizations where the use of, uh, these tools is banned. You know, it's not considered safe from a data perspective. It's not considered to align with, you know, cybersecurity rules and client, um, confidentiality and everything like that. And I think that that's clearly a barrier of people feeling, am I allowed to do this? Have I got the. Have I got the remit to be. To be able to explore these tools and how they can add value? I think weirdly, the fear thing is there's a flip side to that coin, which is fear can be a big driver. There's a lot of people, I think, and definitely some of the people we spoke to were, uh, embracing AI and using it and trialing it and playing with it and doing things with it because they felt scared that they needed to make sure they. I mean, I certainly feel that myself. Right? They felt scared. I must understand this, I must be on top of this because this is coming right? Whether we like it or not, this is going to be there and we're going to have to respond to it. So I think if you can get the, if you can use fear in a healthy way, then it can actually be a driver. Uh, it can actually encourage and motivate people to take that leap and to start experimenting and start playing. Because there is no question that it's going to be, it's going to be present and we're going to be required to be using this. I have seen mandates, for example, from organizations who've said for any business case, if you're asking for headcount, you need to justify why this couldn't be done by AI first. So straight away you're now seeing that embedded in governance, you're seeing that embedded in the way. And you know, these are potentially, uh, slightly more advanced organizations, but where they lead, others will follow in a matter of weeks, months.

Speaker B: This is where this piece of research that Morehouse has done has come in. There was a huge amount of insight that was gathered which obviously can, can be used when working on transformations, but also you've put together uh, an in depth report into what this looks like and something that people can use immediately. From your perspective, did it feel like this was being developed, I guess with a bit more of a behavioral lens in it and that was that it was action focused. And what role do you think this kind of creative storytelling element has in bringing it to life? Because you can just present the facts, everybody presents the facts. That's on AI presented every day. But what do you think was, was different about this project?

Speaker A: So for me that is, that is at the crux of it. So anyone can go on LinkedIn, they can go on any kind of news service and like a huge volume of the content that they'll see will be about AI, this fact, that fact, this thing, this thing you need to buy this tool, etc. What we wanted to do with this research and with this report was to create something genuinely practical. We sort of joked and talked about we wanted to create a lonely planet for getting value from AI. Something that you could genuinely use to sort of break down the problem in very functional, practical, pragmatic ways to start addressing it. Because so much of the content I read, you know, there's a catchy headline and it sort of says, oh, you know, here's the answer, or this is a story, or this is something you should know. But then when you dig under the Surface. It's like uh, ah, okay. You haven't actually really told me anything. This isn't very useful. That's, that was really the goal for us. We wanted to create something super practical. So we, we included things that broke down this, it sort of eat the elephant. Right? So broke down the problem into. Right. Think about four different types of AI, not just AI, and where might this apply? Think about how you build your use cases. Think about the process you go through in terms of ideation, in terms of um, adoption and in terms of scaling. So very much sort of frameworks that people could follow, ideas that they could take on board, uh, examples of where other people have done things that they might be able to pinch and try for themselves. That was very much the goal. And I think that is at its core behavioral. It recognizes that this is a massively complex, challenging, in depth, uh, problem. And actually the way to tackle it is to break it down. And that is behavioral. It's being able to embrace and think about, well, the little steps I might need to take or how I might find my way in. That's right. For uh, my team, in my organization, in my industry, that was very much the goal. And I agree, I think there's no shortage of detail and information about the tools and what's out there. The behavioral bit is where you convert those tools into value for the organization. One of the things we pulled out in the report I think was, um, don't let the perfect be the enemy of the good. You won't figure out everything and you won't know all the things. And the example, I think we, I think some of the research indicated that one of the big barriers was around data fragmentation. And people said, well, I can't really take advantage of AI because my current data is in such a poor state. And the message we were trying to say here is that it's better to get going with something and accept that problem, uh, because you will find that you find your way to value. And equally some of the problems you're having around data fragmentation can themselves be solved by AI. So it's actually momentum which is critical. And I think that that came out very strongly actually in the report and the feedback we had, particularly in the qualitative interviews as well, that sense of, um, what do I need to have in place? It didn't need to be as significant or as polished or as perfect for you to start getting value. And then that became a snowball. You could then start to see the advantages in its sort of uh, virtuous

Speaker B: Circle, I guess if somebody is a chief data officer, where would they start in terms of defining success based on everything that we've learned?

Speaker A: For me it goes back to that point around is this a data problem or is it an enterprise problem? Is it about the organization as a whole or is it a data thing? And I think if you're for any chief data officer, anyone working in that field, they are always at their best when they are in lockstep with the business and they understand what the business and its customers are trying to achieve and they're reacting and then they're also being proactive about what's possible and, and what could change. And I think that two way relationship is where it's best. So for me, success is that we're able to add AI into that mix and maintain that two way street. So here's what I'm trying to achieve for my customers. Can you help me? And equally, here's a new thing that we know is possible. Can this help you? Um, that I think is a critical sort of flow. I think at a different level. For me, it's impossible to build a picture that is going to stand up to the change that we'll see week by week, day by day, hour by hour in this space. So it's about identifying areas of advantage that you can drive. Can I make that process quicker? Can I provide a little bit more value for the customer? Can I take a little bit of cost out of that over here? Can I take that team and enhance its activities so that it can do something different or that it can add different value? That's the way I think you can determine, determine value best. Rather than saying here's a picture of the organization as it will be in 10 years time, have we got there or not? Because I just think it's so varied and variable and you're so unlikely to get it right.

Speaker B: It comes back to actually something where we kind of started this conversation, but just the idea that AI should be there to help you achieve your vision, achieve your mission, achieve your strategy. And so actually measuring success in terms of AI isn't about, well, have we got copilot enabled and what percentage of our work is done by AI toolkits or all the rest of it. Actually the question is if this is our goal, if we want to be the most sustainable beer company on the planet, or we want to transform the way people travel around Europe or whatever. The thing is, the question is not how are we using AI in our systems, but are we uh, are we achieving what We've said we're going to do in our roadmap and how have we used AI to help us achieve that? And if it's not helping you to achieve it, then don't do it. You somehow have to remove that noise from the system, don't you?

Speaker A: M. Yeah, definitely. I totally agree. I think the build on it is being able to look back and say, where did the use of AI not just allow us to enhance what we do or make what we do cheaper and easier, but where did it help us to understand that something different could be done? Where did it give us a new piece of insight about what people might want? I saw an interesting stat the other day which talked about the fact that, um, for a lot of, um, online businesses, particularly in the E commerce space, the attention they're getting through search, you know, what they're seeing in terms of their sort of marketing, uh, and outreach performance has gone, uh, significantly down. They're seeing much less referral, they're seeing much less kind of traffic, but their sales are up. So that's really interesting. Right. So it suggests that the morass of inquiries that they were getting before through search were unqualified and people weren't then buying. The way people are using, uh, LLMs like ChatGPT now is creating a more qualified buyer. So when they do come through, they are more likely to be the right customer, they're more likely to be more informed, they're more likely to be able to navigate more quickly to what they want. And I think we've got to read that as a good thing. We've got to read that as a good thing that people are getting more information and more insight that allows them to make better decisions, whether that's the supplier or whether that's the customer.

Speaker B: So we did a lot of research and we've talked about this for, uh, quite a lot. But what's something that you're still chewing on from the research? Uh, is there something in this that challenged your assumptions or that you would like to explore more in the future?

Speaker A: I mean, that's a really good question. As ever, when you do any piece of research, you go in with a question and you want to get an answer to that question, which we got right. We got a sense of what the barriers were, we got a sense of where people were excited, we got a sense of how far through their journey people felt that they were. But again, as you would hope, it actually ended up raising more questions that we then wanted to go and tackle. So I don't think there's one kind of question that it's left us with. I think for me one of the areas that is probably now coming to the fore as something I would like to spend more time on is understanding genuinely not just on the ground, how do we optimize, you know, how do we make customer journey better or how do we make, you know, how do we push more value add more time to value adding rather than non value adding for our agents or our care workers, whoever it is on the front line, it's the impact on strategy and strategic decision making that is uh, I think uh, a Pandora's box to open to open up now. But I think that definitely came out for me as a sort of stimulus from the report.

Speaker B: Okay, so final word. What, what word or phrase would you use, um, to describe where do we go next with AI? And I just want one word.

Speaker A: Excitement. Excitement and any excitement is tinged with a bit of fear and a bit of optimism. And I think that's a fair summary certainly for me and I think reflects the client conversations we're having. It's going to be unknown. It's going to require us to gird our loins to cope with different things that we haven't seen before. It's going to require us to behave in new ways. And you know, any change like that is difficult and requires us, you know, sort of um, to think differently. But it's also there's so much opportunity, it's going to throw up so many different ways to do really, really positive things. So I think, yeah, that would be my word of choice. But the unpacking to make sure that we're clear that it's not just all positive, it comes with a little bit of trepidation is definitely my thought.

Speaker B: Thank you so much James. I've absolutely loved having you on the show. We will talk again soon. Thank you so much for joining me.

Speaker A: Thanks very much Becky. It's been fascinating.

Speaker B: Thank you again James for joining me and uh, for being so open and generous with all of the thinking behind this research. Super, super interesting. And we have got plenty more coming this season. From nature based finance to fast moving consumer brands, from B2B behavior change to public sector innovation. And yes, there's is beer coming. Every episode is an exploration of inside out impact. Don't forget to subscribe. Wherever you get your podcasts, Google BHMP, message me on LinkedIn. You can get transcripts, you can get downloads, you can get more behind the scenes insights. Lots and lots of fantastic content. So please, please do subscribe and if you're working inside a complex organization and wondering how to move from insight to action, that's what we do at BH hmp. Our Impact Thinking framework helps organization turn strategy into creative systems that genuinely work. Thank you so much for listening and I will see you next time.