AI is changing the world of insights, but just how much change are we likely to see, and what will the insights teams of tomorrow look like?
In this episode of The Persuasion Game podcast, two AI marketing experts join us to offer their views on what is an ever-evolving landscape.
What will AI mean for how insights are gathered and interpreted? What will it mean for where resources are directed? And, crucially, where do humans fit into all of this.
As Nick puts it: ‘The whole industry is at a crossroads.’
The question is – where will it end up?
Jonathan Williams is Founder of One Strategy Studio and Nick Graham, formerly of Mondelez and PepsiCo, is Founder of Vertemis.
Episodes are released bi-weekly. Follow us on LinkedIn for updates.
Subscribe to The Persuasion Game Newsletter on LinkedIn.
Want to know more about us? Visit our website here: thisistheforge.com
This is an 18Sixty production for The Forge.
CHAPTERS:
(00:00) Introduction
(02:43) The Shift from Data to Decision Making
(05:37) AI’s Transformative Potential
(07:42) Challenges and Opportunities with AI
(14:01) Adopting AI: Strategies and Pitfalls
The Persuasion Game is available on all your favourite podcast apps: https://link.chtbl.com/PersuasionGame
Episode transcript:
What next for AI and insights? With Nick Graham and Jonathan Williams
Nick: I think the whole industry is at a bit of a crossroads right now.
Jonathan: People generally say that the impact of technology is sort of overestimated in the short term and underestimated in the future, and I imagine this is going to be the same.
Simon: Welcome to The Persuasion Game, The Forge’s podcast about growing brands and persuading consumers in the modern age.
So I’ve spectacularly failed to persuade any of my colleagues, no Adam, no Laura, no anyone, else to come and join me today. So whilst I’m flying solo, I’m joined today by two brilliant guests and we’re going to be talking about the future of client side insights and analytics functions, particularly in a world of AI.
And my two guests today, Jonathan Williams, founder of One Strategy Studio and Nick Graham, who has held insight VP roles at Mondelez and at PepsiCo. So two people who are expertly placed to think about the challenge of what a future fit client side insight and analytics function looks like within this world of AI and the disruption that is coming.
And what you’ll hear from them today is a story that, yeah, faces into the realities that this insight industry will change and that there is disruption coming. But also you’ll hear a story of optimism, one to fuse with a belief that this is a great opportunity for the insight industry and real opportunities for all of us to take.
Welcome Nick. Welcome Jonathan. Lovely to see you both again.
Nick: Lovely to see you too. Great to see. It’s been a while.
Simon: It has indeed. It has indeed. Well, great. Thank you so much for joining us today. Talk to me a little bit about what do you see as the future of the client side insight function, and maybe Nick, we’ll kind of start with you on that one.
Nick: Sure. Being honest, I think, not just the client side insights teams, but I think the whole industry is at a bit of a crossroads right now. I think we’re stuck, or for too long we’ve sort of been stuck in output mode, thinking about being seen as methodological experts, subject-matter experts, very much focused on the what we do as opposed to what we deliver and what the outcome is of what we do.
You know, obviously we’ll talk a lot today about AI, but I don’t think it’s just AI. I think businesses are changing. I think technology more broadly is enabling and forcing us to operate in a different way. And I think fundamentally there’s so much data now that the sort of old model of working just doesn’t work anymore.
That’s why I say we’re at a crossroads, because I think we’re at a real opportunity, a moment of opportunity I think for the insights teams. I think now this is the moment, you know, again, having worked in client side insights teams for 10, 15 years- the teams really want to move from being stuck in the data and the doing to actually advising the business, driving decision making. And I think this is that moment, this is that moment when I think insights teams using technology as part of that can start to make that shift being much more of a decision driver as opposed to sort of like a research analytics output team. But I think that’s not easy. That’s not an easy shift to make.
Simon: So it’s really about a kind of transition, a shift from, from being in the doing.
Nick: Yes.
Simon: To being in the real kind of delivery and the value creation?
Nick: Yeah, I think it’s from being in the doing to being in the decision making or to shaping the decision making, I think that’s the big shift that I think insights teams and the whole industry therefore needs to facilitate.
Jonathan: So yeah, Simon, I mean, I really agree with all of that. I think the insights function has really always been about the consumer focus and helping organisations put the consumer at the heart of decision making and really helping these organisations with evidence-based decision making rather than gut. So really making sure those decisions go in the right direction. I think that isn’t changing, but I think there’s been a lot of assumptions about how you do that in the past.
Methodologies have been built up around that whole systems and organisations have been set up around how that has worked in the past. And you know, I’m looking at this very much from an agency point of view. So I’ve been agency-side all my working life. But the same things apply agency, I guess as client side, as in so much of the focus of these methodologies have been about collecting data, processing data. How to get the most out of those data to get them into the shape where they can influence decisions that typically, I don’t know about client side, but agency side maybe 5 or 10% of the effort and time and the value delivered is around what it actually means
Nick: Yeah.
Jonathan: And driving these decision making where that can change now in the future. That’s the transition point I think that we’re talking about.
Nick: And I’d say honestly, even in client side, insights teams, 80%, 90% of the time is in the middle. It’s in the middle of the funnel, right? It’s not enough time is being spent or is able to be spent at the beginning, which is sort of framing the brief. What’s the problem to solve? And then delivering into the business because just the data collection, data cleaning, data processing, interpreting the data takes up so much time and effort and just gets, honestly, despite technological progress.
Actually, that bit’s just gotten worse because there’s now more research, there’s more analytics, there’s more data, and so what teams are saying is despite the sort of promise of transformation, often they’re just stuck on faster, busier hamster wheels in the middle parts. Rather than really being able to free up to do the beginning and end better.
A lot of the time I feel like we’re just optimising or augmenting the existing process as opposed to saying, how do we use this and that. That’s why I think AI is such a critical technological and transformative shift, but I think it will fail if all we do is just throw it on top of our existing processes and methods and sort of make that a little bit better as opposed to saying, hang on, let’s use this as a moment to totally reimagine how we operate.
Simon: Let’s dive into that a little bit more, because if I’m playing devil’s advocate, 10, 15 years I’ve been hearing, go to conferences and people stand up and go, insights need to step up. We need to get to the table. We need to be driving decisions. We need to be the kind of the champion of actual decision making. We need to get out of the methodologies. So that ambition has been around for a long time. Why do you both have such optimism and belief that now is this moment that actually the long talked about and kind of hyped, sort of, well, hyped is the wrong word, but the often discussed kind of transformation that needs to happen?
Why now?
Nick: I think for me at least, AI and particularly generative AI, gives us radically different opportunities. You’re absolutely right and you’re right to question it and you’re right to push on what’s going to be different this time round. But I do think what AI enables us to do will help overcome a lot of those barriers that either insights teams have felt or they’ve sort of used as a crutch to carry on doing the thing that they wanted to do.
And it’s a bit of both, to be totally honest. So, you know, for example. If you run a, you know, digitised quant piece of work, you still have to write a report. You still have to tell a narrative about it. You still have to analyse the data and interpret the data. There’s only so much that previously technology was able to really help you solve for. But I think what AI enables us to do is to transform, particularly the sort of analytical, interpretive analysis, storytelling,opportunity spotting. It’s like the next bit of the process, and I think obviously that’s where the real transformation could come. But I will take your point, which is it only comes if we seize the opportunity and we seize the opportunity to jump into it.
You know, I work with client side teams, some of whom are embracing it, and some of whom are still resisting it and pointing out all the problems with AI and all of the hallucinations or biases or whatever it might be. And yes, of course that’s true, but if we just try and sit on the sidelines and go, well, this doesn’t work.
You know, I work with client side teams, some of whom are embracing it,and some of whom are still resisting it and pointing out all the problems with AI and all of the hallucinations or biases or whatever it might be. And yes, of course that’s true, but if we just try and sit on the sidelines and go, well, this doesn’t work, it’s not going to change, not going to create the transformation that we want.
So if we’ve got to jump in and embrace it and make it useful for us and make it what we want, as opposed to just sort of sniping from the sidelines.
Jonathan: I totally agree. A lot of technologies of the past have promised transformation and I think probably over promised in terms of actually what’s going to change and what’s possible, and you can see why there’s invested interest in that.
I see the AI technologies as truly transformational and can transform every aspect of what we do in the world of insight, whether it’s data collection, analytics, storytelling, ideas, strategy. It can impact everything. And so there’s the transformational potential of the technology, but I think there’s also a cultural shift.
Simon: Yes.
Jonathan: That I think is important. That is this technology is changing everything at the same time, so people are experiencing it in their everyday lives in a way, if you talk about other technologies, like online surveys or panels, I mean these technologies were touching their lives a little bit, but the sort of the transformational power that’s so inherent in AI that people can see in their everyday lives.
There is a hunger from organisations, I think, to embrace this and to change, and I think that strength of cultural potential is there, and there are people who are positive and there are people on the sidelines,sniping, as you say. But I think on balance there’s a large shift or focus on where can this go and how can we embrace this? And so I think that puts some impetus behind what is a technological potential? There’s a kind of, a cultural embracing, I think, you know, with legitimate concerns on top and on a focus on getting things right. But there’s a momentum behind it that I think that perhaps hasn’t been before.
People generally say that the impact of technology is sort of overestimated in the short term and underestimated in the future. And I imagine this is going to be the same. That it won’t happen as fast as it potentially could or people are hungry to happen and there’ll be these kind of organisational shifts that have to happen, but when they do the transformation will probably be more extreme than people expected further down the line,
Simon: Obviously you are both working with organisations all the time who are embracing this. Let’s talk a little bit about the spectrum of what you are seeing. The common phrase I hear in, and this is about all types of application of AI, not just within an inside industry, is what’s the use case?
Nick: Working with a lot of corporate side clients, they’re being pushed to create an AI strategy and to adopt AI. But to what purpose? To what end ultimately, which is your point about use cases. Sort of broadly speaking. I think there are three big buckets that I’m seeing clients work through.
One is streamlining things. Even simple things like, you know, coding open ends, coding open ends, but coding open ends and adding more texture and more interpretive analysis. Stuff that isn’t necessarily complicated, but just takes time, right? Takes time and effort and energy today.
There’s a second bucket, which is around surfacing things, so patterns in data, opportunity areas, innovation ideas. So using more of the generative capability, right? So actually using it to identify opportunity areas. I think that’s probably one of the most rich in reality spaces because I think there’s a ton of opportunity to use that again, for insights teams in particular, to move beyond the, here’s what the data says or here’s, you know, the analysis too, so here’s what you could do. Here’s the opportunity area, here’s the innovation. So it starts to give insights, much greater opportunity to step into the advisory role as opposed to the reporter or the analyst role.
The last bucket that some of the sort of clients are a bit more advanced in the journey is the sort of simulate bucket. So you’ve got sort of like, you know, streamlined surface and simulate. That simulate bucket, which is where digital twins, simulated research using AI, synthetic data, et cetera like that. But I think that’s where there is the most question about what’s the value of that and what’s it really doing for businesses.
Jonathan: We see clients moving at different paces for all sorts of reasons. I also think it’s interesting the benefits that client organisations are looking for. What we tend to find is a headline message. Speed is something that often connects very early on, but there’s a speed element and then there’s a cost saving element, and those are attractive and I think they can really drive people to think about doing things differently. But actually the two benefits that matter more after people start to embrace. One is different ways of working, iterative ways of working and actually transforming that human journey and making it a more positive experience. And the second is getting to the same or better outcomes.
Simon: Yes.
Jonathan: The ability to actually get to better ideas, better insights, better opportunities, and ultimately drive better decisions. And I think if people can achieve those two things, new ways of working, better outcomes. I don’t think they’re actually bothered whether if it was quicker or as much the price.
Of course, companies want price saving, but you know, if they’re saving money on this, they might just be doing more projects of this type in order to get to more, better decisions, if you know what I mean. It’s not, the ultimate end goal isn’t to move incredibly fast and save lots and lots of money. There’s what are the headline benefits that catch the eye and what are actually going to make a difference and move the needle within the organisation.
Nick: Yeah, I couldn’t agree more. I think you are right, the opening message is often about efficiency, but actually it’s the effectiveness that actually is much more compelling and much more sticky. I think for clients, particularly for insights teams, who again, I think a lot of them are coming with a, not always, but with a bit of a skeptical lens. Can this really do as good a, as a human consultant or a human researcher?
Simon: You’ve both talked at the beginning around some of those sort of realities of things that have been promised in the past, and without getting all Kevin Costner on it, we can build it, but will people, will they come? You know, will people actually use this stuff? Can you talk to me a little bit about the human dimension, the behavioral dimension, and what are the changes in the transitions that we, as insight and analytics people, that we really need to be doing to make the most of this, this opportunity for transformation?
Jonathan: There’s a lot of software out there that promises a lot, that have been adopted by client side organisations that have struggled to deliver against the promises against them. I think there’s a certain amount of elements of just the actual time, but sort of mental effort of engaging with software that is counter-cultural to how organisations want to operate and make decisions, but also to rely on partners to provide that journey such that software isn’t necessarily just the solution.
It will fail or not create the potential that the technology has. If we assume, we can throw some technology at people and assume that it’ll be adopted and change the world.
Nick: Yeah. Well, I mean, such as the history of technology, right? Or certainly in organisations, and again, we’ve seen this with previous technological shifts, right?
I mean I saw this through the digitisation of research when I was at PepsiCo. We invested in digitised platform, but it failed or didn’t succeed to put it positively at the beginning because we didn’t think about the full change management that has to go with it. Like there wasn’t full support for people in terms of making that transition. Suddenly you were just landing them with the, you’ve got to set up this survey, you’ve got to do this, you’ve got to do the analysis, and that’s a huge change in behavior. But it was the same people with the same amount of time that we’d thrown that onto.
Having a full change management process to go alongside this, again, I’m not talking about if you’re doing a small pilot, but if you want to really adopt it, you need a change management process to go with it. You need to think about what training people need, what skills they’re going to need to be able to be successful, what processes need to change. And that’s why you shouldn’t be doing 27 huge big AI transformation projects. You should be doing lots of pilots and then picking a few that are really successful, and then scaling those. But scaling them intentionally, thinking about how are they going to really land in the organisation.
The difference between those who are like dabbling at the edges and those who are really embracing it is those that are really embracing it, they’re picking things that they know are transformative and they’re truly thinking about how to land them into the organisation. How to drive change and how to overcome people’s hesitations and skepticism and all of that stuff like that has to come with it as well.
Simon: That, for me feels really, really key that that psychology part of this, right? Which is, we’ve all read the new stories and the fearmongering about what it could be, the extent of how many jobs can be replaced, et cetera, et cetera. That sense of identity is this stripping away my sense of identity. And how do you address those fears and those potential things that are holding people back from fully embracing and taking that opportunity that you both are so optimistic about.
Nick: Well, I think to your point, is any change like this and right it’s because of all the fearmongering, is going to create loss aversion and it’s totally human nature. There’s something’s coming along promising to do the thing that you do as well, if not better- of course, you’re going to poke holes in it. Of course you’re going to say no, no, it can’t possibly. No, no, no. Look, all the mistakes it’s made. No, no, no. And of course not.
So I think we need to help. We need to help that transformation journey. We need to be realistic, need to be realistic about what AI can do and what it can’t do, and what it should do and what it shouldn’t do. Corporations need, agencies need the guardrails and be explicit about what is and isn’t in scope. But I also think we need to help people realise that what’s the to here? We know what the from is because we know we’re living it today, but what does the to look like? What are we trying to free up people’s capacity to be able to do?
And that’s the bit that I don’t think many of us are focused enough on is like, what’s the so what for the people in our companies? Like what do they now get to do as a result? We sort of said, oh great, it’ll be, you know, be faster, it’ll be cheaper, it’ll enable you to do more. And then what? Then what does that really mean for me?
And actually I think, and this is where maybe I’m just the eternal optimist, actually, I think there is a good to story, which is the ability then to free people up, to spend more time on framing the problem, understanding what the business issue is, and actually potentially even using AI to help understand and help to frame up what the business problem is in the first place.
Jonathan: I see it as two areas of optimism for the future and the direction where it’s going. One very much reflects that and we think of it as kind of releasing the 10%. Traditionally, as we talked before, a lot of people’s time is spent on the sort of the doing and the steps on the process and managing companies and organisations to get to the great point of thinking and ideas. It often feels like you’ve left it to the 5, 10% of time of the actual the really strong thinking and the guiding and those things are sort of crammed into small amounts of time to release that 10% to be a much bigger focus of people’s time in terms of, yeah, asking the right questions, framing the challenges, going through initiative processes, landing the thinking within organisations influencing impacting people. I think that’s very exciting.
But I’m also excited about the craft. There’ll be times when models, data sources are a great way of getting to certain amounts of information, but there’ll be other times when you just need to connect with consumers, with people to have empathy, to understand situations, that’s not going to go away. We’re not saying that’s going to replace that. The balance between different types of research and when you go to consumers when you don’t might change. But my hope is that it actually frees up more time and budget to do the times when you do need to go to consumers when you do need that really strong consumer craft, or whether it’s semiotics craft or cultural craft. When you really do need that, then you’ve got more time in budget to actually invest in it properly and do it properly.
So I’m kind of optimistic for both, releasing the 10% and then really focusing on the craft where it’s going to be most powerful and most impactful. So actually the craft sort of thrives as well.
Simon: I think that’s really important. I remember being at a conference a few years back and just hearing someone kind of talk about, oh, we shouldn’t sort of prejudge different types of projects. You know, there’s no such thing as a sort of a simple kind of research brief, you never know what you don’t know, and you should always…and I remember at the time thinking I’m sort of pretty uncomfortable if we don’t recognise that. If we assume that all projects are…
Nick: Right. It’s terrible advice actually.
Simon: Exactly. If we don’t recognise that some projects are really complicated, right? Where we almost don’t even know what the question is yet, alone what the answer is, and some questions are, should it be red or blue? Then we are wasting our time. You know that’s what at The Forge we spend our time doing is, those big questions where frankly, you know, there are an infinite number of potential solutions and questions that you could ask. They are very, very different projects from ones where actually it’s a defined set of possible outcomes. It’s a much simpler question, and as you say, if AI can help us free up time, so these projects aren’t all equal and that actually we get the time to really invest in thinking about the big challenges that surely has gotta be a win for everyone.
Nick: Completely. Right? Because again, if you talk to insights folks today in client side companies, they’d say 80%, 90% of their time is taken up with repeatable, fairly standard questions. There’s only 20% of their time actually focused on the big strategic questions, the big knotty problems when they really need to be using their brains.
Jonathan: I think in the past, and for good reasons, people think we want to be consumer centric, so they have some big knotty questions. They think, well, the first thing we should do is talk to consumers. And then you’ve spent a lot of time and a lot of money going into consumer time when you don’t actually, you haven’t really framed the challenge, you haven’t pushed the thinking. And I think that’s one of the really powerful things now with the AI methodologies is you can do AI driven entire methodologies in just a few hours, takes a few days in terms of the human journey, and you can see how far you can get. You can see how much you can learn.
And you know, in these automated projects you’re referring to, you might be 95, 99, or a 100 % of the way there. In these big strategic projects it might be 90% of the way there, 80, 50, whatever it is you’ll get a long way. You’ll have strong hypotheses, you’ll have strong insights. You’ll have some early ideas perhaps, but then you can decide ,okay, now we want to go to consumers
Simon: Right.
Jonathan: Now we can really focus that precious time with other people where it really matters where we know it’s going to have an impact. Rather than just launching into large numbers of groups, with we need to learn stuff because we’ve got a new question. I think that’s what’s exciting.
Simon: What would your advice be to people who are thinking, you know, in our industry or wanting to come into the industry, what would be the skills, the things that you would be saying, these are the areas that you should really focus on because they are going to be what will make the difference in the next 2, 5, 10 years?
Nick: To be honest, I think they’re some of the same skills that we talked about for decades, right? We’ve talked about, I think framing. Strategic framing of problems, I think, is going to be a really important skill. It’s always been an important skill, but I think it’s going to become more so because in a world where we’re using LLMs more. Understanding how to frame a problem to then ultimately prompt an LLM I think is going to be really important. And I think it is a special skill that insights people bring to organisations is that strategic framing of, I’ve got all of this stuff, how do I turn this into what the real problem is that we’re trying to solve for the real questions that we have?
And then I think if we imagine that the bell curve of time that people spend is flipped, instead of spending all the time in the doing, we’re spending time on the framing and the landing into action. The other part is influencing skills. The organisational psychology. How do organisations work? How do they make decisions? Understanding those, storytelling. Those skills, I think are going to become even more important because again, even if the AI can get you to a good solution quicker and faster, you’ve still got to then sell it into the organisation and get the organisation to take action on it. So to me, those would be critical skills still to build.
Jonathan: And I think if we think about the skills, there’s a broader question as well about how and why and where does new talent come into the industry? And I think that’s a really important one. In a way I like to be optimistic about this because people say, you know, without all those junior jobs, how do people enter the industry? Well, what are, or were those junior jobs? A lot of them were, let’s get somebody in and for the first three to five years of their working life they’re doing a lot of the doing, a lot of the process stuff. Yes they’re learning, but not necessarily in the most enriching and positive way. They’re sustaining and, I’m leaning a bit on my agency experience here more than the client side, but they’re kind of there to sustain that pyramid model and that 90% doing -is that the best way for new talent to enter the industry? And is there a brighter future in this new model?
And I like to think of it as, rather than people coming in to sustain the pyramid model, they’re coming in more as apprentices. New talent is coming in, they’re working with experienced people on knotty challenges. You know, that shift to the 10% that they can start bringing, the knowledge, the expertise they have or at least the skills, the core skills of thinking and they can start applying that straight away on knotty challenges rather than serving time. And I think that that’s exciting. And I think the same goes for the craft as well for people who want to learn the craft consumer connection and qual or semiotics or cultural analysis, whatever it is, to have a much more rich way of engaging with that at the early stages of their career and learning.
Simon: Nick, all the soft skills that you talk about and Jonathan as you say, so as long as humans are involved somewhere in the process, right, it still has to deal with humans. And humans are messy as we know. It’s just so, so complex. What’s always struck me is that is going to be true for anyone in a commercial role, anyone at all, sales, marketing, supply chain, whatever it is, you are going to have to do those soft skills. Who’s better placed?
Nick: Correct. Exactly!
Simon: To have really strong soft skills about understanding people, about what’s not said, about reading body language, about reading interpersonal dynamics than insight people. This is literally what we do. Not only should it be a like, oh yeah, but actually should be a source of reassurance and confidence for the industry that kind of goes, if anyone can lead this change within a client organisation, if anyone can show the power of humans in an AI-enabled business, surely the insight industry is, is perfectly placed to do that
Nick: Ultimately, so we’re still dealing with humans in organisations and ultimately the people whose behavior we’re trying to nudge in the direction we want to nudge are humans. They’re consumers, they’re shoppers, but they’re human beings, right? And so the AI will get us to a certain point to Jonathan’s, you know, 50, 80, they’ll get it to a certain point, but we’ll still need deep human psychology. We’ll still need to understand the drivers of behavior.
Simon: Well, gentlemen, thank you so much for your time. Really thoroughly enjoyed chatting to you both today. Thanks so much for being on The Persuasion Game.
Nick: Thank you very much.
Jonathan:Thank you.
Simon: What a brilliant discussion. I went through all sorts of emotions during that conversation, I’m not going to lie. But I’m leaving, feeling really optimistic, really encouraged with a strong sense and conviction of the really exciting potential that this has for the work we do, the impact we create, but also the role for humans within it.
The thing I keep coming back to in my head, I’ve got this visual, Nick’s visual that he talked about -about inverting the bell curve. That currently so many of our existing processes we’re stuck in the middle, doing the insight work, doing the analysis, et cetera, et cetera, and we don’t as an industry spend enough time upfront on the framing and at the backend landing.
It is so, so crucial and the transformation effect that AI can do to free up that time, to enable us to actually have those really challenging, difficult, knotty conversations I think is just fantastic. And as I said before, honestly in the forties experience at least, that’s the area that our clients enjoy the most, and I’m thrilled to think about how much more we can do of that to really solve those big knotty challenges and drive fantastic growth for all of your brands and businesses.
Before we go, just a quick message, The Persuasion Game podcast is going from strength to strength, and we really couldn’t do it without your continued support. If you can subscribe to the podcast and leave us a review or share an episode with a friend or colleague, we’d really appreciate it. Thank you for listening.
