Melina Palmer: Welcome to episode 599 of The Brainy Business: Understanding the Psychology of Why People Buy. In today’s episode, I’m excited to introduce you to Richard Chataway. Ready? Let’s get started.
Announcer: You are listening to The Brainy Business Podcast, where we dig into the psychology of why people buy and help you incorporate behavioral economics into your business, making it more brain friendly. Now, here’s your host, Melina Palmer.
Melina Palmer: Hello. Hello, everyone. My name is Melina Palmer and I want to welcome you to The Brainy Business Podcast. Oh, my goodness. Before we talk about anything else, can we take a small moment to just acknowledge that this is episode 599 of The Brainy Business Podcast? How cool is that? More than eight years of this show and I’ve had the opportunity to connect with so many amazing people over the years, whether that’s hearing from listeners like you or the amazing guests and researchers. So just thank you so much for being here. Whether it’s your first episode or if you’ve been here for all 599 or somewhere in between, I appreciate you more than I can say. And one person I’ve been so lucky to get to know over the years is actually my guest today, Richard Chataway. Richard is the author of The Behavior Business and this is his second time on the show.
Melina Palmer: He’s a behavioral scientist who’s been working in the field for ages. He’s had some fantastic roles over the years, including Director of Behavior Change Strategy at Universal McCann, Head of Ogilvy Change at Ogilvy, UK CEO at BVA Nudge Consulting, Director of Behavioral Science at Concentrix, and more. Today’s conversation leans into AI and human behavior, specifically as it applies to call center interactions or working with chatbots. And while we will be talking a bit about the customer side of this, it’s more about the way to think about the strategy as you set up your teams to be profitable and a good investment for your organization, which includes how you hire, how you set up goals and time, and look at all those roles moving forward because a lot of companies are getting it wrong and that can be costly. Even if you don’t work with a call center or experience center in your organization, I encourage you to stick around because the insights impact all areas of the company and can shed a lot of light on your own reactions as a customer. It’s enlightening stuff. I know you’re going to get a lot out of it really quickly. Before we get into the conversation, I want to be sure you know that there are links in the show notes for my top related past episodes and books, ways to get in touch, and more. It’s all waiting for you in the app you’re listening to and at thebrainybusiness.com/599.
Melina Palmer: Now, let me jump right in. Richard Chataway, welcome back to The Brainy Business Podcast.
Richard Chataway: Thanks for having me. It’s great to be back.
Melina Palmer: Yes. Well, and so we have chatted quite a bit in between the last time you were on the show. We even met in person when I came over to London and went to Nudgestock a couple years ago, saw you there, and we were able to connect outside of that, and had plenty of other interactions. And so you just had said you sent me the updated bio and headshot because it was five and a half years the last time you were actually on the show. Whoa.
Richard Chataway: Crazy. Yeah, yeah, I know. Crazy, crazy to think about it, yeah. And we got to meet face-to-face in Seattle as well, which was very exciting when I was there last year. So that was—yeah, it was great to catch up.
Melina Palmer: Yes, so amazing. So, I of course know you quite well, but for everyone who’s listening who doesn’t yet know you and the work you do, can you share a little bit about yourself?
Background and the Evolution of Behavioral Science
Richard Chataway: Sure, sure. So, yeah, I’ve been working, I guess, in applied behavioral science one way or another for getting on for 20 years now, really. But that wasn’t necessarily what I was doing or I knew I was doing at the time when I first started working in the field. I think when I was on the show before, we spent quite a bit of time talking about my early work in the field, which was on anti-smoking campaigns. So I sort of started working on anti-smoking campaigns for the Department of Health in the UK, running marketing campaigns, marketing-led campaigns to encourage people to quit smoking. And we were bringing a lot of principles from what we now know as behavioral science, but at the time was referred to often as social marketing or social psychology, to inform the work we were doing on that campaign. And one, it was hugely effective, but two, for me it kind of was the lightbulb moment to focus on that for the rest of my career. So from there I went to work for central government in the UK, and then led strategy for the Federal Government of Australia for a few years, working for an external agency based in Sydney and developing best practice around behavior change in communications, marketing, and customer experience. In the last 10 years or so since I returned from Australia, I’ve been very focused on those applications to customer experience. So I’ve been running my own consultancy businesses as well as working for organizations like Ogilvy and Concentrix, large multinational organizations working on behalf of lots of big brands that you would have heard of, to help them use behavioral science to deliver better experiences for customers and for employees as well. And also, last time I was on was just after my book came out, which is called The Behavior Business, which is all about applying behavioral science in business, which covers a lot of different areas including technology and AI, which I guess is one of the areas that’s hugely changed in the five and a half years since I was last on the podcast.
Melina Palmer: What’s that? Well, I haven’t heard of it.
AI Predictions vs. Modern Reality
Richard Chataway: It’s a new thing. Well, that was the thing when I wrote the book in 2018, 2019, and I interviewed 25 experts in different fields in different areas to inform the book. And there were some people, some technologists and people specializing in technology and products—tech products—and people who worked at places like Google, for example, who were saying to me, “You’ve got to cover AI. This thing’s going to be huge.” And in 2018, 2019, you know, no one really outside of those businesses was talking about AI or even knew what it was. And so, you know, when we talked about AI in 2018, we’d probably be talking about it in the context of video games and non-playable characters and things like that, rather than anything that, you know, the way we use it today. And so I wrote a section on AI in the book and behavioral science in terms of our understanding of AI. And a lot of people at the time were a bit puzzled by that, including me, if I’m being totally honest. You know, I learned a lot in the writing of the book, particularly about that. And there were some—I’ll be honest, it was the hardest part to write because one, you know, we didn’t know what the future was going to look like. But two, it was a new area really to me and one where I was very heavily reliant on external expertise. But I’m very glad I did because obviously, as we know, things have moved on, times have changed. It’s become something that’s hugely—a day doesn’t go by without seeing some major news story about AI. At the time of recording this, there’s big news stories that broke about the company OpenAI over the last few days, which everyone’s been commenting on. And also, as I say, it was something where, in the work that I’ve been doing over the last few years, it’s become increasingly important and relevant. And also, I’ve been able to build on, I guess, those hypotheses and some of the things that I predicted in the book with more recent studies, research, and writing.
Melina Palmer: Yeah, it’s one of those where you look back and are thankful to the publishers saying, or like you said, some of the other interviewees that said, “Oh, you definitely need this AI thing in there,” and you go, “Okay, I guess I’ll put that in there,” and then go, “Oh look, look how smart I look.”
Richard Chataway: Well, yeah, I mean, there is a risk to it. A few people have said to me, “Do you want to go back and rewrite that bit because it’s a bit out of date now?”
Melina Palmer: Huh?
Richard Chataway: Well, yeah, and it’s interesting because I reread it—not that recently, but over the last year or so. It’s always difficult going back, and I mean, you’ve written more books than me, so you’ve probably experienced this more keenly than I have. But I find it difficult to go back and read stuff that I’ve written previously because you’re naturally going to try and pick apart things and go, “Oh, I’d probably write this slightly differently now if I were writing again.” But I think when I did reread it, the bit that was gratifying, if that’s the right word, was that I think the conclusions that I drew in my book wouldn’t change, but the use cases and examples that I would use—obviously the use cases for AI now are very different. So, one of the things that I said in the book was that AI was already being used by a number of businesses globally to influence our behavior as consumers. So one of the examples of that is around recommendation engines. When we buy something on Amazon, we’ll get the recommendation afterwards saying, “Oh, other customers who bought that product also bought this,” which I think—I’m sure you’ve talked about on this podcast before—is the kind of social proof mechanism that’s being used there, which we know is very powerful. And by some estimates—it was difficult to get the exact numbers for this—I was told that about 25% of Amazon’s revenue directly comes from that recommendation engine, which is huge.
Melina Palmer: Yeah, not nothing, for sure.
Richard Chataway: No, well, absolutely. Yeah, I met someone at Amazon at an event not long after my book came out and said I was told it was 25%, and he said, “Well, I can’t confirm or deny it, but I can tell you it’s a lot.” So, that recommendation engine that’s used on the Amazon site, which we’ve all seen, and similar tools at Netflix, for example, where they recommend shows to you to keep you watching the platform—which makes you a more valuable user and therefore more valuable for advertising—that is an AI-based tool in the sense that there’s no human involved in terms of what content it serves you. It’s basically making recommendations based on machine learning algorithms that are continually optimizing based on data. But it’s not generative AI; it’s narrow AI, as the phrase was often used. When I wrote the book, I was saying AI is already a thing, it’s already being used to influence our behavior and deliver business value, and it’s using techniques like social proof based around behavioral biases to drive outcomes and revenue. But we may soon live in a world where AI is used more extensively and where it’s generative AI, starting to generate content autonomously without human involvement. That suddenly could open the floodgates to a whole host of things. And that’s exactly what’s happened, I guess. My central hypothesis holds: technology moves and evolves quickly, but people don’t. We are still fundamentally, in psychological terms, the same creatures we were 3,000, 5,000, or 10,000 years ago. So the behavioral biases that govern our behavior don’t change. But the way technology is used to leverage and potentially exploit those biases does change. And that’s the bit we need to be thinking about in terms of the ethical dimension. There’s lots of debate about that happening at the moment, but also about how we can make sure these tools continue to work for customers who remain human. Our customers remain humans; we’re not in a situation yet where we’re building products to serve AI and sell to AI. If we get to that point, there’s a whole load of other implications. But also, our employees in organizations remain human as well. The more use they make of AI, the more we need to recognize the uniquely human traits that govern our behavior. For these tools to be effective, they need to recognize that as well. So as a central premise and hypothesis, that holds true, I think. It’s more about the use cases changing, but also the evidence and research that we’ve got now and the experience that I’ve had in the work I’ve been doing over the last few years. Obviously, there’s a lot more insight from that, which is why we thought it’d be good to speak again at this point.
Integrating AI into Customer Service Workflows
Melina Palmer: Yeah, yeah. Oh, and when we spoke the first time, and in reading your book, I know there’s an example that I end up associating with you a lot. I know I featured it in at least one of my books—it is hard to remember as time goes on, I think maybe even two—that has to do with call center experience and thinking about these counterintuitive things, which we don’t have to talk about now because there’s a whole other episode where we talked about it, which we’ll link to in the show notes. But as we were reconnecting and thinking about things like call centers, optimizing customer interactions is one thing and very important. But we were also talking about people who are thinking strategically about these consumer-facing elements, right? Like you said, our employees are human and we’re having to incorporate a version of AI teammates that people work with. So what do we have done by AI? What do we have done by people? What implications are there in getting these things wrong? What are the choices people need to make and things they should be thinking through? That’s really the bulk of the conversation today. We’ll probably mostly talk about it from a call center dimension—or chat, online messaging, emails—however, people will be able to see how this applies to other areas of their work as well, even if they don’t run a call center.
Richard Chataway: Yeah, I’d hope so. Because I think, obviously, that’s an area that I’ve done a lot of work in over the last few years, which is why I’ve got examples and that’s the use case where it’s most salient for me and where I can draw on that experience. But also, it’s one of the industries that has been—and again, this was something I predicted in my book—most impacted by the use of this technology, both for us as customers and users of those services, but also for the people working in those environments as well. It’s been hugely transformative and disruptive to those industries, but that doesn’t mean it’s not going to be transformative and disruptive for other industries as well. I think a lot of the learnings from that are going to be relevant for people outside of those industries.
Melina Palmer: Yeah, yeah. So let’s dig in a little bit with maybe an example or two. I know we’ve talked about some really interesting stuff that I’ll prompt for if you don’t get to, but I think you will.
Behavioral Shifts: Human vs. AI Interactions
Richard Chataway: Sure, yeah. So a lot of the challenges that I’ve been working on over the last few years have been very focused in the call center environment. In my most recent role, I was leading behavioral science globally for one of the largest providers of customer experience services and call centers for brands around the world—a business that had nearly half a million people working in call centers all around the world. The work we’ve been doing was around how we could use behavioral science-based interventions or nudges in order to deliver better customer experiences, encourage customers to buy more in terms of sales conversion, improve customer satisfaction rates, and encourage them to adopt and use digital products and services. Those techniques are very effective; we’ve got some great case studies and results that we delivered. But obviously that industry is one where AI is increasingly being used as a tool to facilitate those interactions. We’re moving from a world of human-to-human interactions—a customer on one end of the phone talking to a customer service representative on the other—to a human-to-AI interaction because you might be talking to a chatbot fueled by AI. Or even when you’re talking to a human on the phone, what they’re saying to you is being prompted by an AI assistant that’s prompting them on things to say to you as a customer. So indirectly, you’re talking to AI. Now, what that means in behavioral terms is interesting because you might think, “Well, it’s not going to have any implications around how I behave. Whether it’s coming from an AI bot or a human isn’t going to change how I respond.” But actually, the evidence we have is that it changes it quite a lot. The set of biases that govern a human-to-human interaction—for example, this conversation that we’re having here today, hopefully it’s clear that it’s two humans talking, not two AIs or aliens having this conversation. But if what I’m saying to you today, Melina, were being delivered by an AI avatar of me, even if every word were the same or if it were a bot that I trained, you would respond to it differently simply because the biases that govern the interaction between two humans are different than between a human and an AI.
Richard Chataway: There was a really interesting study that demonstrated this effect quite neatly. It was conducted in China a couple of years ago. The way it worked was that participants were recruited for the study and told they were going to be interacting either with an AI or a human through a chat interface. The task set for them was very simple: they said the human or AI had preselected a number between 1 and 10, and you had to guess what number that was without any other information. It’s effectively a random guess. Then participants were told what the correct answer was, and afterwards, they revealed what their guess was and how accurate it was. The thing about this scenario is that there’s an opportunity to lie because you’re told the answer and then asked to reveal your guess. The answer was always eight. Because people are guessing a number between 1 and 10, the average answer should mathematically be around 5. Interestingly, when people were guessing against a human, the average guess they reported was 6, showing that people lied a little bit about how accurate their number was, even though there was no financial gain. But when people were guessing against an AI, they reported 6.8 as an average. It showed people lied a little bit more simply if they were guessing against an AI rather than a human.
Richard Chataway: Where it got really interesting was when they recreated the scenario and put some skin in the game regarding financial benefits for individuals. The scenario was created where people had purchased clothing items online and wanted to return them. The only way to get your shipping fee back for returning the items—which is a relatively small amount, about five or ten dollars—was to lie about your reasons for returning the items. You had to say the items were faulty rather than just that you didn’t like them. Again, you had this interaction through a chat interface with either a human agent or an AI agent. When speaking to a human agent, 12% of people were willing to lie to get their delivery fee back. But when talking to an AI agent, 62% of people lied to get the delivery fee back. So 50% more people were willing to commit fraud effectively simply because they were talking to an AI rather than a human. That demonstrates that our social norms around reciprocity, honesty, and fraud don’t exist in the same way or to the same extent when dealing with AI compared to a human. So if we’re increasingly using AI tools to deliver customer experiences and people know they’re talking to AI, they are more likely to be fraudulent. If you are a bank, for example, that’s a huge risk to your business if you’re simply trying to recreate human-to-human interactions using AI instead. To your point about integrating AI as colleagues into existing teams, that has huge implications. We need to think: Is AI a direct replacement for a human? Should it become a colleague, or should AI be integrated into existing human workflows and processes rather than directly replacing humans? I would argue the evidence is very much in favor of the latter.
Melina Palmer: I love that example so much because it’s very stark. It’s pretty obvious the issue. The argument people make all the time is, “It’s cheaper, it’s more efficient than hiring a person to do the thing.” But if 50% more people are going to claim a refund and get the delivery fee back, that’s significantly more expensive! It’s not like everybody’s going to behave the same like they always do because that’s what humans do.
Richard Chataway: Yeah, absolutely.
Melina Palmer: Those ripples matter, right?
Data Biases, Output Ethics, and Counterintuitive Truths
Richard Chataway: Yeah, and you’re 100% right that simply looking at it in terms of the cost savings of removing that human role and replacing it with AI ignores that the cost to the business in that example is going to be hugely more than the cost saving made, resulting in a loss of revenue and profitability. It needs to be looked at more holistically. Like any machine, AI is dependent on the inputs you give it, and then it produces an output. Like any manual piece of manufacturing, you put something into a machine and it pushes something out the other side. When we think about inputs, one thing we know—and again, this is something I wrote about in my book—is that any data that humans touch will contain traces of human bias. The training data we use for these AI tools will be influenced by those biases, and there are lots of well-known examples of the implications of that. One example in my book was when Amazon introduced a tool for screening resumes for people applying for jobs. Very quickly, it screened out all resumes from women because it had historical bias against recruiting women in the training data. It just manifested that bias more quickly, efficiently, and obviously than perhaps they had been doing in the past. One really important aspect of any data we use to train these tools is to make sure it is representative and that those biases have been controlled for so you don’t get a negative outcome.
Richard Chataway: Now we’re in a world where we’re using large language models to help us—whether OpenAI’s ChatGPT-based tools, Gemini, or whichever tool people are using—because it’s gathering information from the entire internet, there’s a whole host of biases potentially going into that training data that we’re not necessarily going to be aware of. When you’re using that in a corporate context, you need to be ring-fencing the data you’re using and making sure it’s pulling from the right sources, and that those sources are accurate and can be read adequately by those tools to deliver the right outcomes for customers. So that’s one aspect from the input side. On the output side, we’ve talked about the implications for how people respond to outputs from AI differently than from humans, making governance of those solutions critical. But again, much like when we talk about how my teams have used behavioral science to deliver better outcomes in contact centers, a lot of it was quite counterintuitive. Some of the ways in which interactions between human and AI differ from human-to-human interactions are also quite counterintuitive. We talked about the fraud example, and a lot of people might say, “Well, that makes sense if we think about trends in retail environments moving toward self-service checkout machines.” One of the consequences that retailers have found is that people are stealing more because there are fewer humans around, less supervision, and people are able to game the system by putting high-cost items through as low-cost items. A lot of retailers have had to rethink their strategy around using those machines because they found they were losing money as a result, even if they were saving on the cost of staff. You might say that’s the same as what you see in that Chinese study with the fraud people were willing to commit: “I’m dealing with an AI here, no human is involved, no one’s really losing out by me stealing something, so I can get away with it.”
Richard Chataway: Where you get into interesting areas is when you have, for example, the sharing of sensitive information. In most contexts, most people would say, “Then you definitely need to be speaking to a human. I’m not going to tell an AI about my medical history or sensitive personal information.” Actually, that’s not necessarily true. There was another study that showed that when people were talking to a human doctor, they disclosed 11% to 12% less critical information than they did to an AI. People were more willing to talk about sensitive, medical-related information with an AI than they were with a doctor. That has huge implications for how we use these tools and in what context. The hypothesis behind that is that if we disclose sensitive information to a doctor, there are consequences—we might feel embarrassed or worry they are judging us.
Melina Palmer: Judging me.
Richard Chataway: Yeah, exactly that. Whereas with an AI, we know it’s not human, we know it’s not going to have that same emotional response or potential judgment. Also, one factor that’s probably not taken into account is that we know doctors are very busy, stressed people, and we worry about wasting their time—certainly I do when I see the doctor. Whereas with an AI, that’s not an issue. So we’ll maybe take more time, be more thoughtful about our answers, and disclose more as a result. That’s a very counterintuitive finding, and one that has huge implications for how we use these tools and in what context.
Strategic Workforce Management and Changing Roles
Melina Palmer: What I think is really interesting about that example is that it shows a balance: hearing something like that could help enhance an experience both for patients and for doctors and staff. If you say, “Hey, we have this AI assistant,” as a first step 24 hours before the appointment or when you book it, where you explain what’s going on, the AI can be trained to ask, “Well, that’s interesting, do you also have this other symptom?” You can get into depth on surface-level stuff deeper than what you would have time to get into with the doctor during quick appointments. For a lot of doctors too, I’m assuming they don’t love asking the same four questions of somebody every single time. Solving the puzzle on a deeper level feels like getting to do more of the work they signed up for. If that doctor is able to get those summaries before appointments and take a couple minutes to look over them, they can have a more detailed and in-depth conversation with their patients. That feels like a better experience all the way around. What a lot of people do in other types of call centers is say, “We’re going to let all the easy stuff be handled by the AI to take those quick questions off the plate so people get off the phone faster.” A lot of that looks at the customer experience side. But there’s a problem with that approach that a lot of people don’t necessarily think about. Having been an agent in one of these spaces before, I already knew what the problem was going to be as you started to say it. So let’s talk about why that doesn’t necessarily transfer and why we want to take a step back to understand what it means for our teams. How do we think about the people on the team and the problems unintentionally created for them that have real physical and time costs? Do we have to train different types of agents, hire different people, or give them more breaks?
Richard Chataway: Yeah, absolutely. You’ve hit the nail on the head in terms of how, if we use the technology correctly, it allows the humans in the business—in the case of doctors—to focus on the things they are uniquely good at, things that are potentially more motivating, interesting, challenging, and things that AI can’t do. You achieve that perfect balance: AI takes away the tedious, repetitive work—with added benefits like obtaining greater depth of information—and the human handles the difficult parts that require empathy and creativity in a customer service or healthcare context. The challenge with this is that it fundamentally changes the nature of the job. I saw this with multiple clients I worked with over the last few years where they successfully integrated AI tools. It meant they were able to handle a lot of fairly mundane, repetitive contacts—people calling with a change of address or asking, “What’s my account number?”
Melina Palmer: Yeah, those kinds of things. What’s your routing number?
Richard Chataway: Yeah, all of those! We’ve either been able to successfully channel those people to websites to get information or automated it through AI-based tools. What that means is the calls getting transferred through to humans now are all the more difficult, complex calls. If you’re the person working in a call center, it used to be a few years ago that maybe 80% of your day was dealing with fairly mundane, basic queries that required a few keystrokes, giving the answer, and moving on to the next call. Whereas now, 80% of your day is probably these difficult, complex, longer, more emotionally draining, and stressful calls. All the customers you’re dealing with are potentially angry, frustrated, or have vulnerabilities and particular needs that are more challenging. Your job has changed and become more demanding; it requires different skill sets. That’s something a lot of businesses haven’t taken into account. They haven’t upskilled people to deal with those demands, and they are experiencing consequences like burnout, attrition, and recruitment challenges. If you need people to have those skills, you’re going to have to recruit in a slightly different way using different criteria. Particularly in large-scale industries like call centers where a lot has been offshored, that presents new challenges around how you recruit and train people because they’re dealing with more complex queries.
Melina Palmer: Well, and so much of call center metrics are those efficiency pieces of data, right? It’s calls per hour. If you’re not hitting—depending on the industry and role—12 calls an hour or whatever it is, you’re in trouble. That is bad; you have to be hitting those numbers. Then you don’t necessarily get breaks when they are more difficult calls. If you’re trying to hit targets, that person becomes very stressed. If the people setting up expectations are just looking at metrics while trying to balance customer service, quality can drop if agents are stressed and dreading the next conversation. If you’re not working in breaks or giving space to know these calls take longer without needing to be extreme, it creates a huge disconnect. If you’re not hiring the right team or planning for how that role is changing as you roll out AI, you have a real issue. As we wrap up—which I hate that we’re already at the time to start wrapping up—if you were going to offer advice for people doing this work, how should they think about it? What key questions should they ask, and what are your top tips?
Actionable Advice for Leaders
Richard Chataway: I think one of the issues, as you pointed out, is thinking about how you are recruiting and assessing the capabilities of the people you recruit. One of the things I talked about in my book was how we can use more scientific ways of recruiting and assessing people to see if they already have capabilities around “softer” skills—empathy, active listening, those kinds of things. Give those equal, if not greater, weight than traditional measures like literacy or academic credentials during recruitment. If the job is fundamentally going to be about those capabilities, recruiting people who already have them will help with those challenges. Also, to your point, they need to be resilient. If your day in a call center now is spent mostly being shouted at by angry customers, you need resilience and the ability to manage that.
Richard Chataway: I think of it as analogous to a project we did last year for an organization that provides emergency services. A lot of callers are in states of distress and need support right away. We helped train their human agents in techniques to deliver a better experience and get customers less stressed in that context. That had two great effects: one, customers were happier generally with the outcomes and less likely to escalate complaints to managers. But the other interesting aspect—which wasn’t the primary objective—was that absenteeism and attrition levels came down significantly. The people in the call center felt better equipped to deal with those customers and situations, making their job more pleasant and tolerable because they weren’t spending all day being shouted at. You can look for people who naturally have the capabilities to deal with that, but you also need to provide them with the tools and training. Behavioral science is a great way of doing that. That’s a significant implication for workforce management. There are also big operational issues around making sure people have adequate breaks and support networks, recognizing that fundamentally the job is a lot harder than it used to be. I think of it as analogous with emergency services: you’re constantly dealing with people on one of the worst days of their lives, or certainly a day where they’re going to take their frustration out on you, and you need to be able to deal with that.
Melina Palmer: Yeah, definitely. Having worked in the customer relations department of an airline, you get not-happy people with that sort of experience, for sure. As you’re talking, I’m realizing the best use of AI or other systems is making sure all systems talk to each other so you’re not asking people for the same information over and over again. That frustrates people! “I called and gave my account number and date of birth, then it transferred me to a chatbot that asked for it, and then it transferred me to a person who asked me a fourth time!” No! People are so irritated already. Analyzing where friction and “sludge” exist and fixing that for people is a much better use of technology.
Richard Chataway: Well, absolutely. But the interesting thing is there’s generally a reason from the business perspective for doing that: security.
Melina Palmer: Security, right. But it doesn’t feel like that to the customer.
Richard Chataway: No, from a customer viewpoint, it’s just irritating. One of the most effective ways to deal with it—if you can’t reduce the number of times you ask for information—is simply to explain why. Tell people why you’re asking them again for information they’ve already provided. It really frustrates me when talking to a call center and they ask for information I inputted into the IVR. You just need to say, “I’m asking this again for security reasons, just in case you passed the phone to someone else.”
Melina Palmer: Right! Then I feel all right with it. “Oh, okay, yeah, of course. That makes sense.” Then I’m on board! It’s an easy thing you can train your teams on. I love it. I think it’s a great spot to end—a good, practical piece of advice for people to implement. We’ll have links for people ready to follow you, get more advice, and buy the book. What’s the best path for people to find, follow, and connect with you?
Richard Chataway: I’m on LinkedIn, so please follow me there. I try to be active there. There’s also a website for my book called The Behavior Business at behaviorbusiness.com—that’s “behavior” with a “u”, I should point out, using the UK spelling. Associated with the book, there was a podcast where you very kindly appeared as a brilliant guest, Melina. Listeners can hear you there alongside other authors and contributors to my book. Those are the best sources for more information. As you can tell, this is something I’ve been working on for a number of years. We’ve got great experience and case studies, and I’m advising leaders, particularly in the CX space, on how to address these challenges. If any listeners are interested in learning more, please get in touch.
Melina Palmer: Amazing. As I said, we’ll have links in the show notes to make it easy for people to do that. Once again, Richard Chataway, thanks for joining me on the show. It was great to catch up with you today.
Richard Chataway: Thanks very much. Great to see you.
Episode Wrap-Up and Final Thoughts
Melina Palmer: Thank you again to Richard Chataway for joining me on the show today. What got your brain buzzing in today’s conversation? For me, the importance of zooming in and zooming out on a problem is really apparent in this discussion. It’s so easy in business to get myopic, look at one thing, and think solving it won’t change anything around it. But that often isn’t how it works in these interconnected webs we call companies. Setting up a chatbot to work with customers may seem great in a vacuum, and someone might celebrate cost savings in that decision. But if you don’t realize that people are going to be way more willing to lie to the AI to get free shipping and factor that cost into your plans, it could be catastrophic to your numbers.
Melina Palmer: Sure, it seems easy on the surface to eliminate the easy stuff with AI—that low-hanging fruit that gets simple calls off the phone faster. That’s good for customers in many cases, but when you don’t factor in how that fundamentally changes the job, you may not hire the right people, you could set unrealistic goals, or you could end up with massive turnover as you burn out your team. That’s not a good place to be. Plus, there are often good cross-selling opportunities in those little, friendly calls, or opportunities for someone to relate to your business through a positive interaction with a real person who cared about them. When you remove that element, how much loyalty is lost in those little moments of interaction? This really ties in with the conversation I had two weeks ago in episode 597 with Will Leach, where we talked about the efficiency trap people find themselves in when they start to apply AI. AI efficiency isn’t always the thing you need, even though it may feel like the most immediately measurable thing. As Richard and I talked about, changing those measurements may be needed as your experience changes as you incorporate AI into whatever aspect of your business you bring it into. It’s definitely not something you want to figure out as you go and just hope it works. It is worth investing and getting it right before you lock into a system, train your team on it, and change all your processes.
Melina Palmer: Whether you hire Richard or someone like him to help with that, you’re going to be so much better off if you tackle the strategy aspect beforehand instead of trying to fix it after the fact. As I say in my book, What Your Employees Need and Can’t Tell You, you can’t unburn the popcorn if you put a bad experience or communication out there. It’s really, really hard to draw people in when they have that burnt taste in their mouth and it lingers, impacting the effectiveness of everything you do after that point. So what about you? What resonated with you as you listened today? Is there anything you’re going to apply right away to your strategy around AI as you incorporate it into your teams? Whatever it is, come share it with me on social media. You’ll find me as @TheBrainyBiz pretty much everywhere and as Melina Palmer on LinkedIn. There are links in the show notes to make it easy, as well as links to my top related past episodes and books, including The Behavior Business, ways to get in touch, and more. It’s all waiting for you in the app you’re listening to and at thebrainybusiness.com/599. Thank you again to Richard Chataway for joining me on the show today. It was a delight to chat with and learn from you. Join me next time for episode 600 of The Brainy Business Podcast. It’s going to be a lot of fun—you don’t want to miss it. Until then, thanks again for listening and learning with me, and remember to be thoughtful.
Announcer: Thank you for listening to The Brainy Business Podcast. Melina offers virtual strategy sessions, workshops, and other services to help businesses be more brain friendly. For more free resources, visit thebrainybusiness.com.
