Beyond the AI Hype: Why Execution Will Determine the Winners in Customer Service

Everyone. This is Vickie Harrell, Executive Director of SWPP, and welcome to our webinar today. It’s a very popular topic. We are going to get some great information today about AI myths and realities. And we have two great speakers with us today, and Jen Lee from Intradiem and Donna Fluss from DMG Consulting. And this is gonna be a great hour, and I hope you guys are ready for some fast paced, great information coming right at you today. So just a couple of housekeeping announcements. We will be recording this webinar, and I will send you a link to the recording after the webinar. Several people will ask if they can have a copy of the slides, and we’ll send those out as well. We want this to be very interactive, and so we want you to use the chat on the bottom right hand side of your screen to interact with Jennifer and Donna. So if you would click over there on the bottom right hand side of your screen and see the chat bubble, click on that chat bubble, and then just send us a message in to get us all started and tell us where you’re listening from. And, personally, I would like to know if you were going to be less than ninety degrees today. Because here in Nashville, I don’t think we’ve been less than ninety degrees all summer, I don’t think. So let’s see where everybody is. Oh, seventy six in the Bay Area. That sounds great. Sixty one and raining in Oregon. About to rain in Oregon. Seventy in Minnesota. You guys are making me very jealous today. So, Albuquerque is only eighty nine. That tells me something. Right? Hundred and two in Tulsa. Okay. I feel a little better that we’re probably only ninety eight today. So Oh, Rebecca, I’m in Phoenix too, and it’s nice to be under one hundred, isn’t it? It’s beautiful. And it’s a dry heat. Right? Yes. That’s what we tell ourselves. Yep. Alright. Well, looks like everybody has found the chat. Michael, I think Long Island at seventy one will be nice. So I’m gonna go ahead and turn it over to Jennifer and to Donna and let them get started. Thanks for being with us today. We’re looking forward to all the great information you’re gonna be sharing. Wonderful. Thank you so much, Vicky, and hello, everyone. So excited to be here. Looking forward to a great discussion for sure. Just really quick before we get into intros, during the presentation today, if you have any questions at all, please do drop them in the chat. We absolutely want this to be interactive. We will answer them at the end. So we’re gonna hold for questions and make sure we answer as many as we possibly can at the end. Please drop them throughout, and we’ll get to them. So my name is Jen Lee. And as Vicky mentioned, I’m the president and co CEO at Intradiem. Just a quick little bit about me. I started my career twenty six years ago taking customer service calls for a big computer company back then. Worked in call centers and back offices and cut CX operations the majority of my career, and I’ve now been with Intradiem for eleven years with the pleasure of leading as we help our customers implement their own automation and AI to orchestrate their workforces. So happy to be here, particularly happy to be joined by Donna. And, Donna, I’ll pass it to you to introduce yourself. Terrific. And I didn’t know you started as an agent. That really does explain the depth of knowledge because that foundation just can’t be replaced. And so I’m also thrilled to be here with you guys today and to be speaking with Jen to share some of their thoughts about the realities of applying AI and CX organizations. Now I am I should introduce myself first. I’m Donna Fluss, and I was thinking about Jen’s background. And I’m the founder and president of GMG Consulting. And I was as I was saying, we specialize in CX, both front and back office. Jen, over to you. Wonderful. Thank you, Donna. It’s gonna be a great conversation. We thought we’d start out with a couple of poll questions just to kinda gauge where everybody is in your journey today and some of the things you’re seeing. So if you could just drop your responses in the chat here, I’ll be watching. So the first question is just very straightforward. Where are you in your journey? Are you not started early, midstream, fully embraced, like, in, Seeing a lot of kinda early. Okay. Some fully embraced coming in, a mix of b and c. Pretty consistently in the b to c range so far is what I’m seeing, which maps pretty well to what I’d say we see within our customers. Also, if you have a different answer that’s not up there, feel free to drop that in the chat as well. Good to kinda get a gauge for where everyone’s at. Looks like we’ve got a few, Donna, that are further along, which is is interesting and good to see, but the vast majority what’s AI? Love that. Well, we’re about to find out. That’s for sure. Somewhere in the a to c range seems to be. So, you know, I’m gonna say probably b to c is where the majority seem to fall, and that really resonates. So thank you for everyone for answering that. Incredibly helpful to see and see where we’re all starting off from. Now this question, if you’re, you know, in the early stage, maybe think about what you expect to encounter. But as we’re gonna talk about some of these things, curious to hear what are some of the blockers or challenges that you’ve seen so far or if you’re in the very early stages again that you expect to see. And, again, feel free to add something else into the chat if there isn’t an answer here that really fits your experience so far. Seeing IT starting to come up as the early ones, fear or lack of adoption, real values, and save savings of AI. Yes. Absolutely. We’re I think we’re gonna talk about that some today. A c d. Wow. We’re seeing all of them. Yeah. B c d, all of the above. Yeah. I think, the more I’m seeing coming in here, the more I’m seeing a lot of challenges and blockers and, across the board, multiple things that are getting in the way. So I think this will be hopefully a helpful discussion. We’re gonna talk through some of these and as we go. Oh, HR, not on the list. Great addition to the list as well. All of the above. Alright. Great mix. I think definitely this one seemed to be spread and, inclusive of just about every answer, which just tells you kinda how complicated the landscape is right now. Okay. So thank you all so much for that. We’re gonna go ahead and jump in, and I’m gonna hand it over to Donna to introduce us to the topic here. You know, Donna, you and your team wrote this great research paper, and maybe just talk to us about why and why you felt you needed to do it and what we’re gonna talk about today. Terrific. You know, those other slides really do look like the football slides given that we’re getting into the season with the little twiddle twiddles all over the place. Yeah. This report is truly actually started as my sense of humor. So I’ve been an analyst for, over twenty five years. And prior to that, I also started in contact centers, but I have to which weren’t called contact centers. As a matter fact, they were called ACDs, and then it went to phone centers. So you you got some of the idea here. Although, I did start on the management side because nobody had any idea what to do with these things. And, actually, what’s what’s interesting is I think we’re back to the future here In regard to in regard to AI. We wrote the report because we kept getting asked these questions. So it wasn’t that it just it was a one off or two off. There are a lot of concerns, as we saw from the responses here, about really what people should be doing with AI, or shouldn’t they be doing, or should they be waiting? And, you know, what do they do about IT or the c you know, or the center of excellence that wants to own the operation. Then, you know, how do you shake HR and get them to do what’s right for what’s happening today, not what was happening a couple years ago? So a lot of, a lot of concern and a lot of, strong, strong interest in it. Now there’s no question, Jen, that AI is the most transformative technology to enter the market. And this is in a period of time, the last fifty years have been amazing from a technology perspective. Right? The Internet, about nineteen ninety six, ninety seven, followed by cloud computing computing, both of which had been there, but all of a sudden, they got attention. Social media. I mean, I can imagine a world without social media. I happen to think it would be better. But, I mean, you know, think about how everyone’s walking down the street, not looking at the sidewalk or the car’s coming, but looking at their phones, you know. And then smartphones, right, that came out around twenty you know, in two thousand and six. But none of these technologies, and they’re all great, have the potential to transform the world as does AI and as will AI. But of course, there’s always two sides. The challenge is that AI in its many forms, which includes generative AI that we talk about as the missing link that gives the insights in the world of CX. We’re talking about agentic AI. We’re talking about predictive AI. I’ve been talking about predictive AI for twenty five years because that always was going to be the future of contact centers, but we didn’t know what it meant back then. Now at the same time, the vendors, and this is where the fun comes in, are all striving to sell their solutions and products. And there aren’t a hundred of these or two hundred. There are thousands. There are thousands of vendors just alone in the bot or conversational AI category, just just to mention one. And so in order to differentiate themselves from a lot of products that look and feel the same, they’re marketing. And that marketing translates into miss. Now it doesn’t mean that all of what they’re saying is not true. And some of this may one day become true. But, you know, to give a bottom line before we even get into this, what I strongly suggest to all of our clients is do not wait because, actually, it’s gonna be more costly to wait than not to wait. But please, folks, do this to quote a movie with your eyes wide open. And it’s that reason why we wrote this executive brief beyond the myths, the reality Realities of AI and CX, that Jane and I I just called you Jane because I was just talking to Jane. But Jen and I are gonna discuss today, and this report is on and available to you on Intradiem’s website. That’s right. Yeah. I I thank you, Donna. I just have to say I was so interested in this report because, you know, I I have the privilege as part of my role of talking to all of our customers, so executives at organizations of all different sizes, global operations. And I hear a lot of the same themes come through from them when they talk about their challenges and their thoughts around AI. The first one is exactly what you just talked about. It’s incredibly hard to see through all the noise and hype. There’s so much noise coming at them and so much hype and so many promises. And trying to figure out what’s real is incredibly challenging. The second one is we are seeing customers have early pilot success, but then when they go to scale, the success doesn’t translate. It doesn’t scale. And they’re realizing that something that works in a small part of their organization doesn’t just automatically extrapolate and bring value within the whole organization. And then the third one is AI is bringing them more insights than ever before, which sounds like a good thing, but can actually be really overwhelming if all you’re getting is insights and you don’t know what to do with them. Yes. Information overload. That’s exactly it. So I think I’m excited to have this conversation. I’m grateful you and your team wrote this paper because bringing some clarity and bringing some calm to that noise, I think, is what today’s conversation is all about. So, with that, let’s let’s start busting some of these myths. This is the this is the fun part. So the first myth here, we had to start with this one, of course. AI is going to eliminate the need for human agents. Donna, is it? Well, this one was a real painful one, but I also understand where it comes from. Right? Contact centers are some of the most people intensive organizations and enterprises and for the many, many decades that I’ve been in these environments. And that’s over over four, I’ll I’ll admit to. This has been a top goal of executives. And what happened was November twenty twenty two, OpenAI introduces something that it claims is new, generative AI. Right? ChatGPT becomes available to all of us. And the AI leaders, I mean, these brilliant people are out there saying generative AI is going to replace the need for human agents. So the executives, the boards, all bought into this and put tremendous pressure on contact centers in order to cut the staff. Now those of us, you know, with who are on the ground know that this wasn’t gonna happen when other similar nothing similar to AI, but other things like this came into the marketplace in the past, and it’s not gonna happen now. Unfortunately, there were quite a few organizations who did let go of their staff with the exception of one almost I mean, all of these organizations have now hired their people back. In some cases, they’ve hired more. And in a lot of cases, they actually had to pay them more. So this was really a costly issue. But I also have to point out this isn’t a yes, no answer. No. It’s not gonna happen. Yes. It is gonna happen. There’s a lot of, if I make goodness, in what we are learning with AI. Its primary use in CX organizations today is to augment agents, to be a little buddy on their shoulder, providing them with the context, right, as we make sense of what Gen AI is delivering. And this technology is also allowing these operating environments to scale without adding as many people. But, you know, bottom line is, is it gonna eliminate all people? Absolutely not. If only because whether it’s today or five years from now, customers at times wanna speak to humanoids. So lots of complexity with this one, but I’m just gonna bust the myth. Yeah. I completely agree with you, Donna, in there. You know, we see in our within our customers deployment of this technology, and the reality is we’re not seeing a mass elimination of CX roles. We are seeing some reskilling, reshifting, repurposing, and changes within those roles. Absolutely. And we’re seeing a lot of, to your point, the internal facing AI assistance having real impact. I think what I would say to those joining us here to be thinking about is maybe start asking a different question, not, you know, which of our people is AI going to replace, but more as we implement AI into the work that our people are doing, how does that make the rest of the work harder? Right? What’s left behind? So you think about call center agents where we do see customers having success with some of the voice AI or the conversational AI is in the more kind of rote, what I would call easy back when I was an agent, transactional call types. And as you take those away, what’s left behind for the humans are the harder ones and more of the harder ones, which means I think the job of the human agent during the age of AI is actually gonna continue to get harder. And if we don’t focus on that and focus on how we prepare our people and handle that, then we’re we’re looking at the wrong problem here. But totally agreed. We’ve not seen this kind of big, big shift that was predicted. To your point, that doesn’t mean that this trend won’t become a trend at some point and won’t grow. Right? The technology will continue to take off, but, just simply not seeing this in the field. Yeah. Absolutely well said. Okay. So let’s take a look at another myth that is confusing the market. And this one, the only reason it’s out there is to confuse the market. So the myth is AI is the same as automation. What do you think, Jen? Oh my gosh. This is one we run into all the time, frankly, because we do both. We have both AI and automation, and it’s so easy to confuse. And to your point, I think it sometimes is made intentionally confusing because if something says AI, it might get popular notice right now. The way we think about it, and this may be overly simplistic, but it just kinda helps me. And and when we help our customers think about what they’re implementing, we think about it as automation is really the execution portion of it. AI senses, detects, recommends, does a lot of that other work to kinda surface information, assess insights, make predictions, that type of thing, and then automation takes an action. And they’re actually incredibly complementary, and frankly, I think neither one is as valuable on its own as they are together. So the power to me is when you combine that sort of sensing, deciding, reasoning, predicting capability of AI with the execution capability of automation. I like to use an analogy here that is that of, like, a brain and a nervous system. So if you think about it, you can have the world’s most powerful brain that is sensing, detecting, distilling insights, drawing conclusions, making recommendations to you. But without an actual nervous system to drive those through the body and take action or, in this case, through the organization and take action, they’re just insights. It’s just a lot of thoughts. It’s just a powerful brain, but the outcome isn’t really there. So combining the two, I think, is where we see the most value, and that’s how I would describe the difference. But, Donna, what did your research show, and what’s your your expert take on this? Well, actually, you said it really well. The one piece I will add that kind of adds to the confusion is that automation is workflow. Right? So a lot of organizations have had workflow in their operation for decades. I mean, back back, you know, twenty five, twenty six years ago when I was at Gartner, I think there were nine different types and levels of sophistication of workflow. Now what’s also the case is that while some vendors offer both, other vendors often want off offer one or the other. So, again, I mean, part of the purpose of a myth is to make things a little bit less clear in some cases or to market. And I think that’s a bit of what we’re seeing in in the marketplace. Yep. I completely agree, and great call out on the workflow, especially as we talk about AgenTic. That’ll come in more into play. Alright. The next myth, this is an interesting one that so, actually, sorry. We’re gonna transition into talking about what it really requires to be successful with AI here. So we’ll go to the next myth because there’s a lot of myths about how to achieve success, and everybody’s searching for success with this stuff. So the first one, I hear this one myself. I get pitched this one being asked to buy solutions, which is basically AI can fix data quality problems. Doesn’t matter what your data’s like. Just drop it on, and AI will work, like magic. So, Donna, is it true? Well, this one is, probably you know, so we rank different myths, and I’m this is up in the top five of what is most dangerous, and I literally mean dangerous to organizations. And by the way, there have been some interesting comments about those AI leaders who really did believe that that was gonna be the primary use. I I will say that by the end of twenty twenty three, the whole tone of the marketplace had changed. And a lot of the same leaders said, nope. We were wrong. Great. I’m really glad you’re putting this out there now given literally some of the damage that was done in the marketplace. Exactly. And, yes, the people those of us who run these operating environments, those of us who speak to the people running the operating environments, really, we’re in the know of what what was gonna happen. And as I pointed out earlier, this isn’t the first time something like this has come along. But the myth that you were just referring to about that AI can fix data quality, you know, that always brings me back to I have a bridge I’d like to sell you. It’s the Brooklyn Bridge. I mean, actually, I’m a Chicagoan, which is why I wanna sell you the Brooklyn Bridge. It would be really nice, but it’s absolutely not true. What is the case is that data is the fuel and engine for AI, and it has to be kept accurate, updated, and this has to be done on an ongoing basis. Right? But AI cannot fix bad data. Right? We can we can use AI to improve our knowledge bases, But what the knowledge base is doing when we use it that way is surfacing up things that are redundant or complicated so that they can be reconciled. But, unfortunately, when organizations use that data, like, they just take the SharePoint database that hasn’t, you know, that hasn’t been updated in two decades because, you know, people are really very busy. And you use that data. And what happens with AI, and it’s great at doing this, is that it amplifies it and propagates it. And that’s one of the reasons why bad data is one of the underlying and underappreciated causes of drift and into hallucinations. AI AI models learn from the data they’re given. So let’s sum it up this way. When the data is flawed, incomplete, and we all have incomplete data, outdated, it comes from all kinds of heritage mergers and just migrations, when it’s biased and contradictory, which is almost always the case, and a really interesting series of discussions about, you know, the industry about what to do about that, what will happen is that the model is gonna base its output, its answers, on this data. And it’s gonna reproduce those flaws at scale. And we’ve seen some wonderfully damaging and actually wonderfully wonderfully funny examples of this. So, Jen, what are your thoughts on this one? Yeah. I completely agree, Donna. And, you know, it’s the thing I would add to what you said is that this data effort, it’s not even just a one time effort. You might think, okay. We need to clean up, structure, label, you know, and and and warehouse our data, and then we’re good to go. No. It’s ongoing. Data hygiene is something that has to be a continuous effort because it’s very easy, to your point, for it to start to erode and slip over time. And so I think reminding people that there’s this work that must be done, it is foundational, and it’s the only way to really see success and consistent success is important. And then I would also say, when I talk to our clients and they’re asking, you know, how do I approach this? What should I be using AI for? I push back on that too, and I say, I think, you know, that might be the wrong question. The question isn’t what can AI do in this area. The question is, what data do we have in this area available? What data do we already have available today? How recent and normalized and clean is that data, and how reflective reflective is it of the way work actually flows in this area of the business? And I say actually flows because you don’t wanna look at the way you designed work to flow. That’s not always how it plays out. You wanna look at how work actually flows through your system and through your operations. Ask those questions first, get really solid on those answers, and then look for how AI can help. If you do it in the other order, you’re never gonna get to the ideal outcomes or, frankly, you, to your point, might get to some, very much less than ideal outcomes without doing that work up front. Absolutely. You know, just tying back to one of the surveys, the number of people who said the biggest impediment was IT, I think that was actually the number one answer. And this is really an example of where it perpetuates because IT grabs. Yeah. They just grab a database. It could be the SharePoint, which is which always raises my blood pressure, because, you know, they think it’s it’s one one and done and all the same for everyone. Now related to that, now here’s another one that is really confusing organizations. And as a WFM person, it really bothers me at my core. Okay? So this myth is that AI eliminates the need for WFM. Jen, does it? Oh my gosh. This one. Yeah. Absolutely not. Based on what we’re seeing, based on my, you know, decades of experience, and based on what’s happening in the real world, AI does not eliminate the need for workforce management. And I’m so glad that we’re busting this myth. So from my perspective, what we see is that even when AI is implemented successfully so say we said this isn’t widespread yet, but it is happening. You’re deflecting a large number of interactions to AI, and it’s actually handling calls for you and work for you. The thing about that is it doesn’t just change how the work arrives by adding another way for the work arrives. It also changes the rhythm and the flow of the work itself within your operation, within your CX operation. So because AI is going to handle things differently, it’s gonna escalate things differently. One of the things about AI is it’s not one hundred percent predictable. That’s why it’s probabilistic versus deterministic. Right? So the outcome could be slightly different. It adds more potential volatility in, which, in my opinion, then makes it even more important to have workforce management. So what I think the the key here is not that workforce management goes away. It’s that workforce management has to continue to evolve and adapt, which we’ve been doing for decades, by the way, with the introduction of digital channels, chatbots, all self-service channels, all things you’ve had to adapt. This is another moment to adapt and evolve. And I think that evolution looks less like a focus on scheduling and planning and forecasting, which remain important but become almost secondary to real time orchestration of an even more complex environment with even more volatility with some probability mixed in, now into the mix. So I think, in fact, incorporating AI not only does it not eliminate WFM, but it actually makes WFM more important and elevates it to elevates it to an even more strategic level than it’s ever been before. That’s my opinion. Obviously, maybe a little biased having worked in these roles. But, Donna, what has the research shown, and what did you guys find? Actually, you’re spot on in everything that you’re saying. You know, to begin with, AI increases the need for WFM. Right? Because WFM is, without a doubt, the most important productivity tool in contact centers, but it’s also essential for managing the complexity. Right? I mean, if if you run an environment with, you know, more than a hundred people, you need to have WFM. That’s right. But and what AI does is is add that complexity. And I’m gonna, if I may, I’m gonna give you an example of this. There’s something that I’ve been talking about for a very long time called adaptive real time intelligent routing, and that is a predictive form of routing. I’ve talking about this for twenty five years because, you know, skill based routing has been great, but there are much better ways with the insights that we get from our our switches, you know, through GenAI, through lots of other things. Right? We know a lot about those interactions. The idea behind adaptive real time intelligent routing is that it ignores the established routing, which means it’s it’s ignoring what was forecasted and scheduled. Exactly. And it is doing its thing. So let’s say that Jen is really hitting it on terms of, you know, closing sale after sale after sale after sale. But, you know, after a couple hours of this and having those interactions routed to her because she’s doing such a good job, we see that her numbers drop, and all of a sudden, Vicky’s up there. So totally changes, and everything goes to Vicky. Now there’s lots of issues with the example that I just gave, but what it’s about is the ACD ruling the house. Now those of you who’ve heard me speak on on WFN in the past know that I talk about the concept of real time adaptive intraday management. And in my humble opinion, and we all know what that means, once the day kicks in, the intraday management should become the heart and soul of the operating environment. As a matter of fact, this actually has to happen because it without that, AI is gonna add too much complex complexity to the environment, and all the routing is gonna not be effective. And so what you really need is an agile That’s right. Intraday management application that can handle the shift in the demand patterns in real time. Okay, Jen. Back to you. Completely agree. Thank you, Donna. And such a great example with the routing of how something that we think on the surface reduces complexity actually adds to it. It doesn’t mean it doesn’t bring value, by the way. Like, just for the record, for the group, there can be tremendous value. That value comes with complexity, and your operations have to respond. And by the way, guys, great conversations happening in the chat, so continue to share. And I see a couple questions we’ll come back to. It’s just really interesting to see all the perspectives. So we’re gonna shift now to another category of these myths that I think is incredibly important, and I touched on it when talking about at the beginning what our customers see, which is what AI actually delivers. So very interested in getting into this one, and we’ll jump straight to the next one here, which is AI will reduce operating costs by forty to fifty percent in one year. I’ve heard this. I know our customers have heard this. Donna, what’s the reality for this one? You know, when I first started to be asked about this, I thought people were kidding. But there’s a rule of thumb that we have. The first time, I probably do kind of go, isn’t that interesting? I hear something. The second time, I’m like, okay. The third time, I go back to being a Midwesterner. And, obviously, it’s something that somebody needs to do something about. This goes back, and this claim goes along with the one about agents. Right? So if you can eliminate your agents, you can eliminate a lot of your costs. And by the way, you’re not gonna be eliminating all your costs, which is interesting too. Now related to this, I’ve been on hundreds and hundreds of contact center sites. And as a process engineer, I’ve never been on a site where I couldn’t take out twenty percent of the cost over a period of couple years. Now that isn’t what I’m typically there to do, but there’s just some easy gives, right, of things that you can do. And, you know, any of us who spend a lot of time on these operating environments know what they are, not necessarily something you wanna do. But when you see something like this and you have the leaders of AI making these statements, folks, it’s dangerous. You know, there was a comment earlier on which, yeah, those of us in the field, those of us who run these operating environments, we see these things and we just cringe because it’s really very, very silly. Now but there’s truth in some of this. Right? There was a comment that went through in regard to workforce management that it can be good for things like real time adherence. It absolutely can be good for real time adherence because anyone who’s had the pleasure of sitting there and watching their ten agents, and this goes on a lot at BPOs, to tell them when they’re one second late, first of all, who wants to be told that? That’s absurd. And secondly, who wants to be the person to tell them that? Right. Right? So AI can be very good at a lot of things, and it will change the environments. Right? We’re gonna go from agents to CX specialists who are gonna handle the type of calls that Jen was describing. We’re gonna automate activities where systems do a much better job than human beings. Right? Much better at giving balances, much better at reminding people, excuse me, you’re two seconds late. And again, as you can tell, I’m not a big fan of RTA, but if it’s gonna be done, let’s do it with an AI bot. But in terms of really being able to reduce forty to fifty percent in one year, okay. I’ve got another bridge I’d like you to buy, and it’s the tap and z bridge. Really, this is something that I do not foresee happening in the next five years, although there will be some potential benefits from AI. Jen? Absolutely agreed. And I I think, you know, as a leader of a business, this is one of the most dangerous of the myths to your point, because it really set a trap for business leaders. It put a target out there that then boards and shareholders and business leaders all set their sights on that’s just wildly unrealistic. I mean, we all know, those of us who have been running operations for a long time, to your point, know there are ways to incrementally drive tremendous efficiency throughout an organization. AI absolutely provides new, unique, and compounding incremental opportunities to drive that efficiency. But to throw these kinds of numbers out there set a lot of our customers up for failure because they were given budgets that reflected this. And now here we are down the stretch of the year entering the last part of the year, and they’re having to scramble because it’s not producing it. So, you know, I think it again, it’s a trap, I would say. It’s one that is easy to see and get excited about if you’re a business owner or an operator, but it just is not coming about. And even if at some point we do achieve this level, which, you know, who can predict what’ll happen a decade from now, I still think these are the wrong questions to be asking. The question we should be asking is not how much will AI save me. It’s how will AI reshape the way that we do work? What will it allow us to do that we’re not doing today? And I think that ties into some of the discussion I’m seeing in the chat here about it’s absolutely true that AI can eliminate certain tasks. That’s what it’s built for. Right? And identifying those tasks is the prudent way to proceed with implementing AI, but it’s not going to eliminate everything. And it’s also, as it eliminates some things, going to open up space for new things. So thinking about this from kind of a pure operating cost reduction, think, is not only dangerous, but as a business leader, it’s shortsighted. Because instead of thinking, how can I cut those costs? You should be thinking, what can I do with that extra capacity that I am getting? You know, even though, again, it’s not forty to fifty percent, but I’m getting this extra capacity and time. How can I leverage that to do more, to do more for my customers, to do more for my employees, to do more for my shareholders? All of those things. So it just encourages asking the wrong question, and I’m so glad that we’re bringing this one out. I don’t I’ve not heard from a single customer that we work with that has seen this type of value. It just doesn’t exist. Yeah. I haven’t I haven’t seen that either, and I’m forever asking organizations about it. One other thing, though, that we’ve also learned, and this has started to happen significantly this past year when it comes to AI, it’s also highly effective at helping organizations increase revenue. And that that has been a tremendous win for a lot of organizations. I will say, as a CEO, I would rather increase revenue all day than cut costs like that. You know? For one thing, it’s more fun, but also that should be the target. Every company’s goal is to grow. Right? So absolutely great point. Great point. Alrighty. So let’s get into the last of the myths that we’re gonna be discussing today. And I can’t do this one with a straight face. So I’m gonna just smile my way through this, not because I think it’s funny, fine, but because I think it’s funny. Okay. AI can handle ninety percent of all customer inquiries. Jen, Janet? Oh, absolutely not. We have, as I said, many customers who are working to deploy AI to handle contacts, and we they are seeing some success for sure. Absolutely. You know, I can think of one customer in particular who shared that they’ve been able to get AI handling sixty percent of one specific customer call type. Of their many call types, sixty percent of them in their in the payment category are now being handled successfully by conversational AI. That’s wonderful, by the way. That’s a great savings that, you know, gets that deployed as another way to handle those, and customers seem to be happy with that. But that’s the biggest success story I’ve heard thus far, and that’s one call type within a broader scheme of call types. Right? Nowhere near ninety percent. And I think we talked about this, you know, earlier as well, which is customers are thinking about it now a little bit differently or or rather, I guess, I’m encouraging customers to think about it a little bit differently. Because once again, I’d say, like, this is the wrong question to be asking yourself. Don’t ask yourself how many of our inquiries can AI handle. Ask yourself two different questions. The first is how many should they handle? How what percentage do our customers actually want to deal with this type of technology for? Because your customer experience still matters, and understanding that journey still matters. And then for those that they are handling, I would challenge people to think about a different metric, which is what I actually call the resolution rate or the handling rate, or or I’m sorry, response latency. So in this case, what I’m talking about is what is the time frame between when you determine the AI is not successfully handling some type of interaction and how long it takes you to see, sense, and address that. So if you have AI deployed and it’s handling calls, but you determine at some point it’s not resolving calls, which all of us in the industry know are two very different things. You can handle a call. It doesn’t mean you’re resolving the issue. How long does it take you to respond to that? I challenge you to think about that because during that period, that response latency time, your customers are suffering. Your calls are being handled. Their issues are not being resolved, and your customers are suffering. So even as you deploy, even if you start to see ten percent, twenty percent, sixty percent of your calls handled, make sure you’re looking at whether it’s actually being resolved. So a little bit of a rant there because I’m really passionate about this one, Donna, so apologize. But curious to hear what you guys heard. Like, I’m not seeing it. Are you seeing it? You’re really spot on in terms of everything that you’re talking about, but I’m now gonna turn into a Midwestern storyteller. So way, way back when IVRs were first introduced to the marketplace, I took over what was at the time one of the biggest call centers in the world. It was in financial services. And my second day there, my boss came up to me and said, your IVR is down. And I said, okay. If you tell me what it is and where it is, we’ll look at it back up. Alrighty. So using plain old common sense, we actually back then, and this was in retail banking, got a nine got a sixty five percent true displacement rate. But what we also got was the highway effect and the overall volume of inquiries into our shop more than doubled in a year. So really interesting from that perspective. Now as as you said, Jen, what’s really important when it comes to conversational AI, when we talk about conversational AI, we’re talking about Gen AI enabled, self-service. And, you know, as as Jen mentioned at the beginning, you know, Gen AI provides us with tremendous insights. What do you do with those insights? So, what’s very important when it comes to conversational AI self-service, JNI enabled self-service, is to identify the appropriate use cases. Right? So what specific call types? What specific digital types? What specific, you know, social type of interactions can it handle. And it’s great at informational, transactional, giving balances. As a matter of fact, bots can give balances better than people because we don’t trip over our words. Right. And so what you want to do is find those and one by one automate them. Now most of us already have IVRs already. So take a look, find, start with one, make it better with the conversational capabilities of CAI solutions. Right? So you’re always going to be wanting to increment up. Right? Very importantly, and again, this is midwestern common sense. Please don’t lock your customers into these things. You will get more success by letting them come out. So what’s great about this whole category is that there are already a lot of best practices for these type of implementations because in the early days, starting at at the end of twenty twenty two and going through twenty twenty three and twenty twenty four, there were so many failed implementations. But with the right best practices, right, identifying the right use cases, you are going to be able to do quite a bit with this technology freeing your people to add really add value. So let’s wrap up. And, Jen, would you like to share some final thoughts? Sure. Thank you, Donna. So, you know, I think here’s where I’d leave it as we close out. From my experience myself and from talking with customers, the biggest recommendation I can give those on this call really builds on what you just said, Donna, which is before you drop AI or look to implement AI into what you’re doing, particularly customer facing, but, honestly, this applies to any implementation, take a step back and look closely at your customer journey as it exists today. Look at your data. Look at your workflows. Take the time to really evaluate what’s already working. To your point, if the IVR is working, do we need to do something different? That’s a question to ask. Dig deep. Don’t just, you know, say automatically yes because it’s new technology. Many, many things today could still be resolved really effectively using non AI tools, and and that’s okay for us to admit as an industry. They don’t need to go away just because there is a new tool. So if your customers are happy with how they’re being served, that’s worth thinking through and being incredibly thoughtful. And then the reminder I I always like to give, but I think is more important than ever for those of us who have spent our entire careers in CX, there is a human on the side, on the other side of every interaction that you are shaping, that you are orchestrating, that you are dictating. And remembering that human on the other side, I think, can make us much more thoughtful in our approach here. AI, I think, to your point, at the very beginning, I completely agree. The most transformative thing we’ve ever seen is going to continue to drive transformation and change at a pace which, frankly, as I tell my team, we just have to be comfortable with because it the pace is the pace. But it still requires a thoughtful, nuanced approach, ongoing discipline. All the things that we know as operators and as business leaders have been true all along remain true. You have to continuously improve. You have to prepare. You have to be thoughtful. There’s no easy button. It’s not a magic easy button. How about you, Donna? Any last thoughts to leave the group with? I agree with everything that you’ve said, and I’m just gonna reiterate a little bit. All CX organizations, contact centers, back office operations, and for that matter, almost every function in enterprises should be adopting AI. You shouldn’t be waiting. But please, folks, and this is I’m gonna beg you, approach it as you would any significant investment. AI is technology. So please do your homework and build a business case, an ROI model for your environment. Don’t take those generic numbers that vendors will give you. Ask them for the models. They all have the models, and they base their models on the one fantastic implementation, which is good to know they had, but doesn’t mean that it’s gonna happen again. And, basically, think about it this way. If it sounds too good to be true, it is. Jan, back to you. Perfectly said. Perfectly said. Vicky, time check for us if we want to take a couple of questions. I didn’t do a good time keeping job here, so but we certainly keep it. We are good. We have about eleven minutes, so we’re good. Okay. Perfect. We’ll go over it looks like we have gotten a couple questions throughout. And if you have more, please feel free to drop them in the chat here. So I’m gonna go back to a question asked by Patrick earlier. And, Donna, maybe we can tag team this one. So the question was, should we be concerned that some AI leaders talk about AI eventually replacing human workers even if companies today are positioning AI as augmentation? I think you touched on this a bit earlier in that there is a shift that’s happening, but I I’ll I guess I’ll start, and I’ll emphasize that, Patrick, that there is starting to be a shift. There’s certainly still a lot of talk that I would suggest, is under the category of kind of scent creating a sense of urgency to drive implementation of these tools and purchasing of these tools. So there’s some myth busting that needs to happen under there and think really about the intention of those statements, but we are even starting to see those that made the boldest statements start to walk them back. Predictions are slowing down. I think reality is really starting to hit. So I can’t really answer whether we should be concerned. I think that’s to each individual. I think what we can do is bring our own discernment, and I’ll continue to ask the right questions, dig underneath, and look at the kind of motivations of those making statements like that and make our own assessments. But I will say I think the industry is starting to to quiet in that regard a bit. Donna, not sure if there’s anything you want to add there. We can go to the next one. Well, just to support what, what was Patrick wrote, I’m also where some of my clients who have to take cuts in their budgets on people for implementations that didn’t work. And that the managers said they wouldn’t work, but the senior executives. So this concern is still in the marketplace. And, unfortunately, right, you know, once the cat’s out of the bag I have a cat. She’s very cute. She does like bags. But once she’s out, it’s next to impossible. She figured out that it’s dark in there, she’s not going back in. So this remains a problem within the industry. And if I may, there’s always some silliness within the industry. But that’s where if you build your business case and you actually show reality and you’re talking numbers which they relate to and resonate with them, you have a better chance of making your point. What’s also the case, and I’ve seen this, is that AI is creating a lot of new jobs and new opportunities for people, particularly people in contact centers who have a very broad knowledge. So, again, two hands, give and take. Jen? That’s right. Oh, love that. Yeah. I’d emphasize that your domain expertise, your experience and knowledge and skill in whatever role, know we have a variety of roles here, is more valuable than it’s ever been in the age of AI because AI gets better when informed by humans with solid foundations and domain expertise, knowledge, skill, experience, etcetera. So totally agree, Donna. That actually speaks kinda to this. The second one that I thought we could answer was from Michael, and there was some discussion in the chat about this, which was even if AI doesn’t replace WFM as a whole, could specific WFM functions like traffic monitoring be full fully automated by AI? Donna, you wanna start with that one, and then I’ll I’ll give my thoughts as well? Sure. So I alluded to this earlier. There are certain things that AI can really do very well. And to begin with, WFM has dependent on is an AI application. Right? It uses machine learning, and it uses rules. There’s there’s a debate whether rules is AI or not, but let’s just put put that aside for now. There’s a lot of areas of WFM where AI adds value in regard to, you know, real time adherence and and being that person. I was joking about this, but it’s a great solution and actually much quicker. And, you know, if you ask any RTA person who has to sit sit there and stand there and do this job, it’s no one’s favorite job. You know? Basically, it’s one of the most hated things, and I also think it’s ridiculous, but that’s a whole another discussion. But there are many areas. It can improve forecasting. It can improve scheduling. It improves intraday management. It improves many, many different areas of these applications. So, yes, WFM should absolutely apply AI where it adds value. And by the way, another place that it adds value is it is helping the vendors, release software software much more quickly, although that also partially depends on the underlying architecture of the application. So AI has a lot of value. And the the new, you know, adaptive real time, intraday modules, a lot of that is based on AI technology. Totally agreed. No. The only point I’d reiterate here is as AI does replace the need for humans to do certain tasks, that creates tremendous opportunity to do other things, things we’ve never even dreamed of before, things we’ve always wanted to do but couldn’t do, things we couldn’t get to as far as, you know, from a strategic perspective. And so in my organization, we call this thinking from an abundant mindset versus a scarcity mindset. Instead of what is what does AI take away? It’s what does AI enable us to do that we couldn’t do before? And thinking differently about it, I think, is how workforce management continues to evolve. What could workforce management become when not bogged down by the things like what you just described that AI can do? We’ve got one last question in here, and I think just enough time to answer it, and probably another one that’s good for both of us. It just came in from Grace. It says your primary guidance was to embrace and implement now. I think some of us are on board, but we lack the guidance and potentially the support in how to do that. Do you have any advice on how to organize or implement AI in a way that is scalable and productive? Any insight into how to get a business case seriously considered? I’ll start with a couple things, and then Donna would love to hear, you know, what you work with your clients to advise here. So the first thing I’d say, Grace, is that as we touched on this, I don’t know if we talked about it enough, is it doesn’t have to be all AI that you’re looking at doing, by the way. AI plus automation, workflow driven, finding opportunities. So that’s what I would first say is you don’t have to go out and try to find some big, quick hit, big bang, easy button AI as we’ve just kinda dispelled the myth of that doesn’t exist. So start looking at the solutions that are available and look for solutions that have credible proof points, history, and scale as a place to start in doing your research. When I think about when my team comes to me to make a recommendation, I want to know that they’ve done their research, they’ve vetted it, they’ve got they’ve read case studies, they’ve you know, all the things that human discernment bring to it, frankly, that help inform it. So broaden the spectrum as far as what you’re looking for to do. When it comes to the business case, you have metrics. WFM, in particular, is the most metric driven part of any organization. I know because I came from that world. Right? Everything you do is based on real tangible metrics that inform the cost of your business. So when you build your business case, make it very specific to the metrics you’ll impact. I saw somebody talk about AI that helps reduce handle time. That’s a real tangible money making, cost saving metric that you can impact through assistive AI. That’s a great example of some place to start. Start with something that you know ties directly into your costs. Again, typically, your labor costs when it comes to WFM. Handle time is a great place to start. It’s where many of our customers start is how can we reduce handle time using AI and automation so that we can be more efficient with the headcount that we have and build your case based on those real metrics. So I think that’s the key is grounding it in reality, not counting on any of the hype that we talked about. Donna, obviously, you work with a lot of clients in these areas. So, you know, what’s what’s your best advice to Grace and others who are wondering this? I’m just gonna build on this one, and go back to metrics. Years and years ago, the first time I wanted to buy my first toy, which was a multimillion dollar, ACD, I based it I know, newly minted MBA. I based it on quality because I was running a really big contact center. And in less time than I thought than I thought it could get through the internal mail system, we used to have those envelopes, it came back denied. Like, okay. So I didn’t do my homework. I need to do my homework. What I discovered, and this is the case in most organizations. And by the way, and I’ve never had an investment denied again ever, you know, for my clients ever. I mean, because he who plays the game with Beth is winneth. So the first thing you wanna do after you do what Jen talked about is reach out to your office’s CFO and find out what the thresholds investment thresholds are. And you need to build your business case so that you exceed those investment thresholds when it comes to payback, IRR, NPV, whatever it is the numbers that you’re using. And is that gonna take some guts and some risk maybe in some cases? I think possibly. But if you combine a variety of things into a a business case, you’ll be able to work it out. And, obviously, talk to vendors like Jen. They have models for that. You know? Feel free to reach out to us. We can always help. So back back to you guys. Thank you, Donna. Right. Well, we are just about out of time. We are down to the one minute mark. So are there any last remarks you wanna say before we close this down? Can I I just really wanna address Michael’s comment about not all CEOs understand it’s meant to enhance replace? More and more are coming on board. I talked about the false promises made, the big promises. Those are shiny light billboard things that can be distracting. So, yeah, there’s there’s people still distracted. But I think more and more people are are kinda coming back down to earth a little bit with reality, Michael. So, hopefully, you’ll start to encounter that more. And, Donna, anything any last comment? I agree. Alright. Well, I will be sending out the slides and the recording to this. Cheryl has already put in the chat a link to the to download the report, which I think everybody should make sure you download that report. Yes. But we’ll also probably send something through SWPP as well, so be looking for that. But thank you guys so much for all of this. It has been a very interesting hour, and I hope everybody has gotten as much out of it as I have. So thanks again to you and to Intradiem and Donna for presenting this, and have a great rest of your day. Thank you all.