Supporting Human CX Agents In An AI Era

What does great customer experience look like when artificial intelligence is handling half the calls? For Danny Wareham, Founder and Lead Psychologist at Firgun, the answer has less to do with the technology and everything to do with how leaders treat the humans left holding the line.

In a wide-ranging conversation with CX Today, Wareham argues that too many organizations are approaching AI deployment from entirely the wrong direction. "Decision makers purchase AI tools because they feel that if they don't, their competitors are going to have some sort of lead on them," he says. The result is a wave of reactive, FOMO-driven implementations that fail to solve real operational problems, and leave frontline agents feeling like change is being done to them rather than with them.

That psychological friction, Wareham explains, is rooted in self-determination theory. When agents feel their competence is threatened, their autonomy restricted, or their role uncertain, they revert to familiar behavior, bypassing new tools, creating workarounds, or simply disengaging. The organizations that beat this pattern, he says, are the ones that swap compulsion for curiosity.

His go-to example is IKEA. When AI tools reduced the company's inbound call volume by 47 percent, the typical response would have been to cut headcount accordingly. Instead, IKEA asked why 53 percent of customers were still calling. The answer revealed an appetite for design advice. The company retrained its freed-up agents as amateur interior designers and generated an additional one billion dollars in revenue from the insight buried in their contact center data.

"Your contact centers are the source for where that rubber meets the road," Wareham says. "If your CFO's mindset is around OPEX savings, you tend to miss all of the value that's potentially there."

Wareham also introduces his Constellation leadership model, explored in depth in his book. Rather than confining people to rigid job descriptions, Constellation asks what individuals can contribute toward a shared goal, freeing teams to bring their full range of skills to complex, high-stakes situations. The model draws on military frameworks like mission command used by the British Army and the US Navy SEALs, where shared intent replaces hierarchy the moment situations become unpredictable.

"Constellation is not leaderless," Wareham clarifies. "Culture and our behaviors become the leader."

For CX leaders navigating the AI era, his message is clear: invest in understanding why your people are resisting change before trying to overcome that resistance, and start treating your contact center as a source of strategic value, not just operational cost.

Transcript

SPEAKER_02 0:07

Hello and welcome to CX Today. I'm Rob Wilkinson, and today we're looking at what it really means to support human CX agents in this new world of AI-enabled contact centers. There's plenty of noise going around about automation right now, uh, efficiency, uh, future of custom service. Uh, but the bigger question, I think, is the human one. So, what should agents be doing more of? What should AI quietly take off their plate? And how do leaders get that balance right uh without leaving customers frustrated, obviously, um, or the teams feeling exposed. And uh might sound a lot, but we're gonna get some help to unpack all of this. And I'm joined by Danny Werren. He is the founder, director, and lead psychologist at Fergon, and we're gonna be talking about what great human CX looks like in an AI era where organizations are getting the balance wrong and what CX leaders can do in sort of the next 90 days to better support their people. Uh, we're gonna explore Danny's idea of leaderless leadership and his constellation model. Um, so basically, this is where leadership shifts to a person uh with the clearest perspective in a moment. Uh, and it might sound simple, uh, but until you're dealing with high-stakes customer issues and someone needs to make a decision, then that's where things really come into play. So uh, welcome Danny. Really excited for this one. Thanks very much for joining me.

SPEAKER_01 1:35

You're quite welcome. A pleasure's all mine because it's always uh always a privilege to have a chat with you, Rob. Oh, you're too kind. I'll pay you later.

SPEAKER_02 1:45

So let's kind of sum things up a little bit for the audience and in and kind of just set the scene a little bit. Um in kind of a single sentence, what should great look like for human CX agents in you know, like an AI enabled contact center today?

SPEAKER_00 2:04

So I I think to put it in one sentence AI or or any other automation should remove the routine so that advisors can focus on what humans do best, which is understanding context, showing judgment, building trust, and helping the channel to become a value add rather than a cost centre.

SPEAKER_02 2:31

Nice, really concise, and and delivers, you know, makes it clear. So in your world, um where are you seeing teams getting this balance wrong at the moment? And I guess what's the cost of that, whether that's customers, to agents, or yeah, online.

SPEAKER_00 2:52

So it's it's gonna be a bit of a bit of a rambling answer to this, I think, because often the teams that are using AI are not the decision makers or purchasers in it. So they are they're a recipient of this new process or this new technology. Um so actually the real problems that I tend to see are decision makers purchasing AI tools either because they feel that if they don't, then the competitors are going to have some sort of lead on them. So they need to grab the latest shiny shiny. Um so there's almost a type of FOMO that is driving some decision making, and that's why we're seeing a lot of rollbacks, you know, famously Klarna that let 700 people go, etc. Um, so actually starting with what's the pain you're trying to solve, or what's the issue that you're trying to resolve, or what's the what's the current process that you have that's a little bit clunky or doesn't have the data or insight that AI might be the right tool for that one specific use case, that is where I've seen AI be really successful and actually deliver much better CX uh impacts. From a team point of view, though, they're they're the recipients of this technology. So it's the same pain that we have when we decide to change the GUI on the on the front end of our CRM, or it's the same when we decide to launch a new product and we haven't engaged the front line. That um feeling that the change is being done to you rather than you being um part a partner in that change, that rarely, on a psychological level, rarely leads to people adopting the change. We we tend to see threat responses and you know guarding of information or weaponising of values and and other things that go on because we weren't brought on the journey with them, we were just um had it foisted upon us, and we we don't really understand what it's doing, other than telling me how to do my job or threatening my employment or creating uncertainty in my life, which we we don't respond to psychologically very well.

SPEAKER_02 4:54

You're trying to say that we don't always engage the front line before doing new stuff.

SPEAKER_00 5:00

Yeah, that that was much more succinct, but but there's there's two parts. There's there's this weird le leaders have a way of focusing on compulsion. So what I mean by that is that when we when we introduce something new to our contact centres and there is resistance, we see that resistance as something that we should overcome. And we do it by trying to compel people that this is a good idea. So we have arguments like um this is going to make our lives easier, or it's gonna simplify processes, or it's gonna make uh it's gonna give you more free time, or we spent a lot of money on it. Um, or we put targets in. You know, you need to use this knowledge management tool on at least 10% of your interactions, for example. So we're very good at going to a place of compulsion and trying to compel people why this change is good, or why this tool is good, or why this process is good. The companies that have been really successful in managing change, though, are the ones that go to a place of curiosity. So rather than seeing that friction as something to be resolved, they see it as so they just see it as a point of data to try and understand why it is that you're not using this knowledge management tool, why is it that this change is causing you um friction that we didn't anticipate, and understanding and using that curiosity to then help people be lead you through the change to adopt. Um, so yeah, the the the organizations that have traditionally failed at um at change management tend to be from a place of compulsion, whereas those that manage change effectively tend to be far more curious in the what are we seeing, what are the conversations we have in? Why are people avoiding using this system or bypassing this process or they've got their own workarounds? So yeah, compulsion is not as good as curiosity. Too long didn't read.

SPEAKER_02 6:48

And also, people buy into stuff if they're engaged early on, if they're part of the decision-making process. Um, and and you know, if you if you want to kind of get in there first and and make sure that people are gonna buy into it, then then get them engaged, get you know, and they will help you shape it better as well. It will it will it won't just land badly with them, it will land badly with customers when it doesn't hit the note. And our frontline are the people speaking to our customers, so they get what's going on and where the problems are.

SPEAKER_00 7:18

Absolutely. And the things that kind of motivate us as individuals, that sort of self-determination theory, as we call it, um, the things that give us more joy in the work that we do are you know our autonomy, uh, our competence, and our relatedness. So, relatedness is how do I feel safe? Do I connected to the people around me? That tends to be really good in contact centers, um, particularly in-person contact centres, because we're all kind of trauma-bonded by the same uh customer interactions, by the same systems, etc. But also the good stuff that goes on, the employee engagement activities and so on. Autonomy is a bit of a double-edged sword that we want people to use their initiative to make the right decision, to find the right um resolution for this particular customer within a set of frameworks or set of guide rails. Um, however, you can't have autonomy without competence, and competence needs to be trained and practiced. And often with AI tools, we see it as this genius intern that we've employed that can do everything super fast and super accurate, but then we haven't trained people really how to get the best out of that intern. And that intern knows more than we do. So actually, it's a threat to my competence, it's telling me how to do my job. So you have these opposing factors within people's kind of own motivations of I someone's telling me how to do my job. I don't know what that means for the future and how it's going to evolve my role or if my role's going to be there in the future. So, what I'll do is I'll over-index on the relatedness and we'll just huddle together like scared zebra and uh ignore the threat that's over there. We'll bypass the system, we'll we'll not follow what's been bought in, and we'll just hope everything's going to be okay. And this is one of the reasons adoption rates across all industries is not quite as high as you would expect when we've introduced this technology and people are just not using it or resistant to using it or finding reasons not to use it.

SPEAKER_02 9:13

And you're right, it it's not just AI, we've learned this lesson time and time again. Well, have we learned the lesson? Probably not.

SPEAKER_01 9:19

That's what it's the same mean, different gravy.

SPEAKER_02 9:22

Yeah, with like, and and and you know, you and I have had conversations around what I think about people who roll tech out without kind of doing their homework before not you know, many a times. Um and and and also because you're coming from uh you know, you're not just a psychologist, you're coming from a contact center background for many years working with big companies doing really cool stuff. So I guess asking you to put that CX leader's hat on right now in the modern day, um, there's so many cool tools, right? And I I I I love the opportunities that all of that presents, but it's almost like spoil for choice, and we need to kind of think about which ones are going to have the kind of the best impact. If you could um, you know, if for in the next 90 days, you've got two things that you could change if you were a CX leader to support your agents. What would those be and what would you prioritize?

SPEAKER_00 10:16

I I heard a really good example of uh just a few days ago um around the motive for introducing these tools. So many CFOs and and CXOs are seeing AI as an operational efficiency tool. So um if we can deflect calls or if we can handle calls within some sort of automation, or if we can take off some of the more transactional things, then actually we won't need as many people um on the front line. So it's seen as an OPEX saving. Actually, the organizations that are doing really well are the ones that are seeing just what value is in that data for CX. And the example I heard was IKEA. So IKEA brought in an AI tool that effectively deflected or handled customer queries, and they managed to reduce their inbound call rate by about 47%. So 47% of the calls that were handled by advisors, now no longer handled. Now, what would normally happen, what's kind of typical, is we would reduce our headcount by 47%. What IKEA did is they looked at it completely the other way. Why are 53% of customers still calling? And that rich data, what they realized was the customers that were remaining wanted advice on what kitchen table goes with this cabinet or what bedding goes with. So the questions became about design. What should I design my house? What goes with what? So what they did is they used that data from the AI tool, effectively a form of CX sentiment analysis, to retrain their advisors who had been freed up from this 47% of calls that were now being handled, and they retrained them effectively as kind of amateur um interior designers, so that now they can add value to the interactions with customers. And from that, they created a 1 billion with a B billion dollar additional revenue. Because the insight that's being provided by your contact centers today, we don't tend to look at. Or we we tend to see if your CFO's mindset is around OPEC savings, then you tend to miss all of the value that's potentially there, and your contact centres are the source for where that rubber meets the road, and you get to hear all of those reasons for the calls. So, yeah, the if I was to change one thing, it wouldn't be the recipients of the change. It would be how do you influence um the rest of the organization, particularly the CEO, the CTO, and the CFO, to understand the value that the contact centre is now able to bring to your business.

SPEAKER_02 12:55

That is so important because with the evolution of the technology, the evolution of the role comes, we've seen it time and time again. Evolution of the training needs to happen. That's really key. That's the thing that people seem to miss. But by doing it right, like you've just given a great example, there's some real significant improvements there. And that's not just a monetary value, that is better customer experience that is underlying that monetary value. Not only that, but those agents' roles have evolved, they felt supported. And you know what? They're now in a role that is way more fun, probably. So, you know, it's a thing to be embraced, and you know, to prove a point, if that had been something that they'd been given an idea of was going to happen to their role early on in the decision to go down that route, like we've talked about being important to do, then you can guarantee they'd have been on board from day one.

SPEAKER_00 13:46

But um completely, completely if your contact center lost 30% of its calls, then the natural reaction is to potentially look at well, I can save headcount, save FTE, etc. All IKEA did was look at it a different way and say, why are we still getting calls? Why couldn't they be deflected? And in being curious rather than compelling, by being curious about that, they're able to leverage an extra billion pounds of revenue.

SPEAKER_02 14:13

So I want to shift over to um your your new book. Well, your book that was new last year, but it's uh going to be new to some of the yeah, newish newish. Uh and um, congratulations, it's got lots of uh kind of amazing feedback and awards. Um but it really it's really relevant to this conversation because your constellation idea um it suggests that leadership can be shared or is shared uh and should shift to the person with the best perspective in the moment. What does that look like in practice, though, in a CX operation for our for our for our watchers?

SPEAKER_00 14:51

So for for a bit of context, so the way that leadership normally works is what are we here to achieve? So a leader or a leadership team will help design kind of what's our mission or or what's our vision, where we what's the aspiration of where we're heading. And based on that, we will then break it down like blocks of Lego. So what functions do we need? What people do we need in those functions? What are the job roles within those functions? What are the KPIs or the OKRs within those functions? But in doing so, what we're doing is chiseling off bits of what people could bring to the table. So if you look at your job description, for example, if you're in one of these sort of organizations, you'll have a list of 10, 12, 15 skills and attributes and so on that you need. But the chances are you've got a hundred other attributes that are not in that job description. But because you're in that environment, you'd need to stay in your lane. You can't bring all these other skills and experiences. So, for example, I'm a I'm a psychologist, I've spent a lot of time in contact centres. Um, I am also a really good cook. I also spent some time at a side hustle as a web designer, so I know my way around colour theory. Um, I've never had anybody in any role ever say to me, Could you cast your eyes over this and let me know what the layout looks like and what would you tweak? Because that's not my role. I was a contact center manager, I was an employee engagement manager, etc. So we unintentionally kind of shackle people because of how our leadership is set up. What consolation leadership does is it looks at this slightly differently and it says, What are we here to achieve and what can you contribute towards that? And what that does is it allows people, providing we create the environment that allows people to bring all of their skills, we create an inclusive environment. It allows people in a skill shortage era to find ways that they can empower themselves and job craft and create ways that they can contribute to this greater good. And the results, all of the data and the research suggest that where there is ambiguity and complexity and change, this type of model outperforms every other style of traditional leadership across every measure, whether that's employee loyalty or um NPS or um tenure or every every measure. Where the old style of leadership works well is where things are predictable. So if you've got a production line or you know what's coming through, or you have an area of your contact centre that is very maybe an order-taking line, for example, then it might work better if we still have hierarchies and okay's and KPIs. But where we don't, liberating your people from these systemic shackles and saying, what can you do to help us towards this North Star? And that's all consolation leadership is. And that that works great in CX because people there is a balance between people getting what they expect and that kind of transactional do you do what you said you were going to do. But actually, great customer experience and positive experience needs ambiguity and flexibility as well as that bedrock of predictability. And consolation allows you to do both.

SPEAKER_02 17:59

So it sounds like and a phrase might come up that says leadership, sorry, leaderless leadership. Um and I guess for some people they'll hear, well, there's no accountability there.

SPEAKER_00 18:12

It's a commune. We're all sit, we're all sitting around singing come by all. Yeah, that's not what it is, no.

SPEAKER_02 18:18

How do you keep clarity then? Ownership as well, kind of make decisions quickly during high-stakes customer incidents in an environment like that.

SPEAKER_00 18:28

So constellation leadership, it can on first glance it can it can feel a little bit kind of alien and like the the lunatics have taken over the asylum. Um, but actually it's a model that's used in high-stakes places already. So things like Mission Command, which is used in the British Army and the uh the Navy SEALs in America, for example, what they do is there's kind of three stages to their leadership. The first is what you would you would maybe conjure in your mind as traditional army leadership, so very hierarchical, very um descriptive of this is what we're going to achieve, these are the tactics, this is who's on the squad, etc. And also regulatory, you know, don't shoot unarmed prisoners, for example, and uh and so on. There's specific rules and guardrails after the mission. Exactly the same thing happens. So it's still traditional leadership where we have a debrief, um, what went well, what didn't, who who will make the next team kind of thing. What those do though is they set a very clear motive and intent. So the intent is what are we trying to achieve? And the motive is why are we why is it worth achieving? Why are we here? And because everybody in the group knows that, when things go off piste, when when things are happened that are not in the plan and we need to flex, we don't have to down tools, stop the project and go back to the product owner, for example. Um, what we can do is well, we know why we're here and we know why it's important, so now we can that this leadership becomes a process rather than a person, and it becomes malleable, and that might mean that somebody that you wouldn't necessarily think of as a leadership person can step up. So it consolation is not leaderless culture, our behaviors become the leader, and from an accountability point of view, you'll know does that behavior that you're demonstrating now, or that thing that you're trying ad hoc and and on the on the hop, does it help us get closer to that goal? If it does, fill your boots. If it doesn't, then the group actually self-pleace. The group will say that's not going to help, or that behavior is inappropriate, or that focus isn't quite right. So you you don't need policies and processes and um lots of complexity in terms of catching people doing it wrong. It it will work really well if you're very clear about why we're here and what we try to achieve.

SPEAKER_02 20:56

It sounds phenomenal. Uh, and I know you know I could I've read the book and I know I could go talk about this all day, but unfortunately, that's all we've got time for today, Danny. Um thank you so much though. Just before we go though, um for anyone who's watching this who maybe kind of wants to learn more about this approach or uh just check you out and see what you're see what you're all about. What what's the best way for them to do that and to get into touch maybe?

SPEAKER_00 21:22

There's a there's a couple of ways. I'm very active on LinkedIn. Um you can find me Danny Wareham. Um you can also find me at DannyWareham.com. Um and you can find out anything about Constellation if you Google Danny Wareham Constellation. It will appear in your local bookstores, in Barnes and Noble, in in Waterstones, and on of course on Amazon. So just check me out on LinkedIn or at dannywarum.com. Great.

SPEAKER_02 21:46

And don't forget that you can catch up on other stories like this, other videos like this, uh at cxtoday.com. Um but for now that's all we have got time for. Um, thanks very much for watching and um see everyone soon.