Customerland
Customerland is a podcast about …. Customers. How to get more of them. How to keep them. What makes them tick. We talk to the experts, the technologies and occasionally, actual people – you know, customers – to find out what they’re all about.So if you’re a CX pro, a loyalty marketer, a brand owner, an agency planner … if you’re a CRM & personalization geek, if you’re a customer service / CSAT / NPS nerd – you finally have a home.
Customerland
Real-World AI in Retail: Talkdesk Leaders Reveal What Actually Works Beyond the Hype
The gap between AI hype and real-world implementation is creating both challenges and opportunities for retailers worldwide. Talkdesk executives Ed Durbin and Michael Klein pull back the curtain on what's actually happening in retail AI adoption, sharing candid insights from their conversations with retail leaders at NRF and beyond.
"We've spent millions on AI and we're still trying to get it to work," one retail vice president confessed to Ed during a recent conversation. This admission captures the frustration many retailers face as they attempt to transform customer experience with artificial intelligence while managing legacy systems, limited budgets, and talent shortages. Despite these challenges, consumer expectations continue to rise dramatically, with Talkdesk research showing 70% of consumers plan to use AI for shopping in 2025.
The conversation explores how retailers can move beyond viewing AI as merely the "shiny new toy" to building a solid foundation for customer experience transformation. The Talkdesk team distinguishes between three waves of retail AI development: predictive/interpretative AI (which has existed for a decade), generative AI (2024's focus), and agentic AI (emerging in 2025). They explain why purpose-built retail solutions offer advantages over generic platforms that simply "put a spin on existing technology."
Perhaps most valuable is their discussion of the evolving skill requirements for AI implementation. As AI becomes more capable of understanding natural language, the conversation suggests we may see less need for specialized "prompt engineers" and more emphasis on business professionals who understand desired outcomes. This democratization of AI could significantly impact how retailers approach technology adoption and talent development in the coming years.
Whether you're a retail leader navigating AI implementation challenges or simply interested in how artificial intelligence is reshaping consumer expectations, this conversation offers practical insights based on real-world retail experience rather than theoretical possibilities. Join us to discover how customer experience technology is finally evolving from cost center to profit driver through strategic AI implementation.
We have all these extra, all these people and it's expensive, you know, to have all these people, all these people developing AI technologies. And this vice president said Ed, you have no idea the amount of millions that we spent on AI and we're still trying to get it to work. That scared the crap out of me.
Speaker 2:Today I have the pleasure of speaking with Ed Durbin, who is Vice President, general Manager, industry Strategy for Global Retail at TalkDesk, and Michael Klein, director of Retail Product Marketing, at TalkDesk, and we have been, honestly, we've been trying to set up this conversation for what seems like months, even though NRF was merely a month ago, but having finally done so, thanks to both of you for joining me I'm really looking forward to this conversation.
Speaker 1:Thanks for having us. We had fun at NRF. Let's keep the fun going.
Speaker 2:Yeah, there you go. So maybe just to set some context for this conversation, ed, can you tell us about your role at TalkDesk? And then Michael, maybe you could do the same, and then you know, because there may actually still be people out there who are not familiar with TalkDesk. Let's just talk about what TalkDesk is all about right now.
Speaker 1:Well, thanks, that's a good question. My role here at TalkDesk is, like I said, vice president, gm strategy for retail. I you, president, gm strategy for retail. I like to say once a retailer, always a retailer. I have a very heavy background in retail. I've done retail technology companies, several startups.
Speaker 1:But what I find very interesting when I came to TalkDesk is, out of all the conversations that I've had with CIOs over the last 30 years, we've never sat around and said, hey, let's talk about what you guys are using in ZCAS, let's talk about your contact centers.
Speaker 1:No, we were talking about what's going on in the data center, what's with the database, what's going on in the stores, all that type of stuff. So I find it very interesting that with my background in retail, I've been able to take all that I've done in retail. And now I'm here at TalkDesk because I want to be able to promote what TalkDesk can do for the global retail marketplace. And you know I kid around by saying global, but let's face it whether I'm in Australia, whether I'm in Hong Kong, whether I'm in Los Angeles, whether I'm in London, whether I'm in, you know, south America, retail is retail. Retailers are trying to sell more stuff to people that come in the stores. So I have the ability of being able to leverage the technologies that retailers need today, especially in the customer experience area, to be able to help their customers in retail. My history in retail by being able to identify how talk desks can quickly get into and grow their presence across the global retail marketplace. So how was that?
Speaker 2:Well, done Great job, as if we would know different. Interesting point, though and I'm looking forward to kind of unpacking this with you a little bit later in the conversation but how universal some of these problems are around the world with retailers. I mean, some of the things that retailers are still trying to solve for are absolutely universal to everybody else in the world. So if we get to it, that's going to be an interesting part of this conversation. But before we even do that, michael welcome. Thank you, and maybe you can tell us a little about your role there at TalkDesk.
Speaker 3:Yeah. So thanks again, mike, for having us. I'm responsible for TalkDesk messaging go-to-market how we present ourselves, how we grow our business. So typical B2B marketing role really helping drive pipeline, understanding where the demand is, working with our ideal customer profiles, personas and really supporting our sales organization in helping them do a better job of communicating the great value that we deliver to the retail industry.
Speaker 3:Similar to Ed, I grew up on the shop floor and spent many years in retail.
Speaker 3:Then I spent 15 years at Adobe heading up their retail industry marketing. So very similar role at that point and Ed didn't bring this up. But Ed and I have been here for less than a year and we both joined because TalkDesk is very serious about our go-to-market from an industry perspective. We really recognize that in order to differentiate ourselves in the marketplace, we have to be able to speak the language and understand the challenges that our customers have in order to help solve for those problems. And that's really important for us because for both of us, having been in the industry for so many years, there's a lot of tech companies out there that talk about going to market from an industry perspective, but quite often that's just marketing putting a spin on existing technology and here at TalkDesk we have a full team in our product that are developing solutions. We call them purpose-built for retail and that's all they do. Our colleagues in product are only working on product for the retail industry and helping solve for those challenges with technology.
Speaker 2:Everything happened so fast there, Talking with the retail technology companies that were kind of customer facing. You're right, A lot of them have a really neat platform that they're looking to deploy kind of the same thing across industries and verticals and spaces. But they didn't seem to me anyway to have kind of a dedicated vision toward retail, many of them.
Speaker 3:So yeah, we're. That is a key differentiator for us and our three primary industries that the organization has decided to custom build for are retail, healthcare, life sciences and financial services. And when you think about those three industries, the needs of the customer can be vastly different in terms of the data, as Ed talked about, the foundation of what type of data that we're ingesting. Obviously, in healthcare, in particular, the idea of HIPAA and being HIPAA compliant. Financial services, with regulation and then with retail, when we've got so many sources of data and legacy. Retailers that have been leveraging CRM, loyalty, transaction, point of sale, and they have such a spider web of technologies that we need to ingest and communicate with those other systems. So it's really important to be able to build the technology to be able to solve the unique and distinct issues that each of these industries has.
Speaker 3:And that's just in going back to what you talked to ask, we have been typically thought of as a contact center as a software. We have made a shift. Tech center as a software. We have made a shift. And I think also, as consumers and retailers make a shift to understand that, while I need to run great customer service, there is a direct correlation to what my contact center, what my communication is to customer experience and also to the success of the business. How do I leverage artificial intelligence to automate, to be more efficient, to then be able to drive better customer satisfaction scores as well? As, obviously, the primary goal for every retailer is growing revenue and bringing margin to the bottom line.
Speaker 2:Right, right.
Speaker 2:Well, one of the great challenges, as I'm sure you know I'm stating the obvious here, but one of the great challenges for anybody operating in CX for the past, really ever, has been to try and create a direct correlation between CX efforts and a bottom line and lots of indirect ways you can point to the successes of CX and how they affect the bottom line, but directly, honestly, probably hasn't been real until fairly recently with our ability to process these mountains of data, largely using AI. And now, wait a second, you know you, you can actually see the cause and effect on the bottom line. So it's kind of a, it's a really neat moment to be in in the CX world for that reason. I mean, you're, you are, and many of your counterparts are, um, kind of emerging from cost center to profit center in certain, in certain ways. So and I attribute a bunch of that to what's happening in AI.
Speaker 2:Speaking of which, that's what pretty much the only thing anybody really wanted to talk about at NRF this year was what flavor of AI we're deploying and what you can do with it, which I don't know if you were at NRF last year, but everybody had AI, but it was kind of like the big message was we have AI and that was it that. You know, it was kind of the end of there.
Speaker 1:It's still the message.
Speaker 2:I mean.
Speaker 1:It's still a message, it is for a lot of people.
Speaker 2:I was happy to see that there are people like yourselves talking about how it's being deployed and what those effects are on the bottom line and on the actual customer experience real world experience at the end of the day, Well, you know, I'll give you an example.
Speaker 1:So we talked to a sponsor the CIO luncheon before their big meeting 60 some odd retail CIOs. So I was allowed to go up there and say a few words in front of the crowd and out of the 60 some odd CIOs, I knew half of them, which is great, which goes to validate the fact that TalkDesk is retail. I know a lot of people, michael knows a lot of people. But the interesting thing is is before I went up there, I asked the gentleman running it. I said what does everybody want to talk about After I leave the room? Of course they said all they want to talk about is AI. I said was anybody doing it? He's like no, they're really not. They're all struggling with how do I take this beast by the horns and what do I do with it?
Speaker 1:And then I have conversations with CIOs, and one CIO is a good friend of mine. He likes to equate it with something that he's been saying to me for 20 years that I've been selling to him. He's like Ed, don't talk about the shiny new toy, make sure I have a foundation to manage any shiny new toy. And so I think that's where there's a big plus with TalkDesk Because of our platform, our modularity, how you can use and not use it.
Speaker 1:We're helping retailers with a foundation to be able to effectively do CX, and then we're applicable and, where it makes sense, leverage in the AI that we have, in the modularity that we have. As the retail realizes that's an area that they need to go to, that's an area they need to raise the efficiencies of, and so I think we are me and TalkDesk, and hopefully Mike will agree with this we are in a position right now that none of our competitors have right now. So I think our biggest challenge and being on podcasts like this allows us to get the word out, because I think the more people hear it, the more people we talk to. I'm looking forward to our growth in retail over the next few years.
Speaker 2:I'd love to hear more about those conversations. It's funny they only take place in the sidebar conversations. The reality the real real of a company's situation and their struggles rarely happen from the platform, almost never happen at the booth. But the sidebar conversations is where the magic happens, and you seem to have had a bunch of those. So my question is when you're talking to these CIOs and they're telling you what their real struggles are, I guess you and I have both been around long enough to know that you can't just say I got your solution, call me on Monday and we'll take care of this. It's a much more complicated thing to do and oftentimes it has to do as much with platform and technology as it does with culture and expectations, which are really difficult conversations to have sometimes. But how do you do it? You're an industry vet. You've been doing this a long time. You've had these conversations. What happens after a CIO says yeah, I'm trying to figure this out.
Speaker 1:Well, I have a very easy formula I listen to the problems that the CIO is talking about, then you ask questions about what type of outcomes are you looking for? And then you figure out, well, how can we partner for success to try to get you those outcomes? And now let's talk about those technologies, because too many people I mean, I came from VMware and sales reps would walk in the door and go, hey, you want to buy some Kubernetes, you want to virtualize that? Well, would you ask me what my problems are? So I really think and with a lot of the CIOs that I talk to, they appreciate it when you sit down and listen to them and understand the outcomes. I've been using that word forever, but in the last few years, all of a sudden, oh, we got to talk about outcomes and making our outcomes. I really really feel that that's the key Listening, understanding, figure out how we can partner and then, whatever that technology is and again it goes back to we. You know, talk to us, we give you a great foundation and whether that, whatever that AI, that shiny new toy that I mentioned earlier, where that fits into the foundation is best going to help them. Because every CIO tells me says you know what, you come in and help me do something quicker, faster, under budget, and it's going to help my outcomes. I will use your pen and sign the deal.
Speaker 1:So it's easy to say that, but it does sometimes boils down to that easy. But you got to be talking to the right person, and the right person is usually CIO, cdo. I mean, you know, michael, the personas that you and I have been talking about. Where do you see in the angles of who you're talking to? Because he was on like one side of NRF. I was on the other side, you know, we were sort of like, you know, hitting it from both sides, but at NRF, what were the type of personas that you were talking to? That talk desk resonated. Is it OK that I was asking questions there, mike? It's.
Speaker 2:OK, it's OK. I'm going to step away for a cup of coffee. I'll be right back.
Speaker 3:Yeah. So, yeah, I have a couple of definitely thoughts as I was listening to Ed talk Definitely thoughts as I was listening to Ed talk, and I agree with that that we're in this phase of AI where we've gone from a lot of hype to what is this actually going to do for me and specifically to what Ed was asking. I was in the digital council where Ed was over at the CIO council. I was in the digital council where Ed was over at the CIO council. So those folks are more of the marketers, the chief digital officers, they're on the business side more than on the tech side and they're certainly looking for technology to be able to help them drive their key business outcomes. But I would challenge the conversation with this idea that we've been dealing with AI in some way, shape or form for more than a decade, and that being the idea of predictive and interpretive AI IDC. I always quote this. I was at an IDC event last year and by 20 and two years, we're still going to see that 70 percent of the AI that we use is still what one would call the traditional AI, that being the predictive and the interpretive, and shiny is all around the generative AI, which was 2024 was the year of Gen AI. This year, in 2025, we're going to see agentic AI as the big buzzword this year. And agents We've introduced agents into our solutions and TalkDesk. We've been doing AI in some way, shape or form since 2018.
Speaker 3:So I think that there is a bit of a learning curve of what is the AI that I have, the AI that I've been using isn't going to get thrown out because it still has value.
Speaker 3:And again, it's what is the right hammer for the right nail, and you can't just have one hammer for all the nails. They don't all look the same and that is where some of the challenges come in and the conversations that we have and I have in some of those sidebar is I've been doing this and do I need to change doing that? Because I have in some of those sidebar is I've been doing this and do I need to change doing that? Because I've already kind of played around with AI a little bit. It helps me understand where my inventory needs to go. It may help me understand next best offer. But with this gen AI and agentic AI, where does that fit into the equation? So that's kind of the other, I think, side of this conversation of how do I take what I've been doing and then adjust it to this new world of gen AI and agentic AI. Does that make sense?
Speaker 2:Yeah, yeah, very much. Thanks for that context. I'm thinking to another part of the conversation that we had we touched on anyway at NRF and that was not only how TalkDesk is deploying and how you're really seeing the idea of AI out there, your vision for it, but also how consumers are responding to it. You know it's a really big deal and I think your team had sent over some data that said I'm kind of pulling this out of thin air so I could be wrong here, but it was something close to 90% of shoppers that your team surveyed had engaged with some sort of AI technology during the holidays and the number was the percentage of those people that responded positively to it was giant, it was just giant.
Speaker 3:Again, I don't remember that number, but it was yeah, so I have those numbers if you want them.
Speaker 2:Yeah, yeah yeah, yeah. Throw them out there, save me.
Speaker 3:Yeah. So what we did in marketing and with our partners at Shift is we did a survey of both retailers and consumers before the holiday season to understand what they expected to do. And then what I always like to do is then I want to understand what really happened to do. And then what I always like to do is then I want to understand what really happened. So, right before NRF, we literally pulled aside a thousand customers, talked to them in the first week of January so we could bring this information to NRF, and what was really interesting is that seven out of 10 of the consumers we spoke to said that they're going to be using more than likely using AI in some way, shape or form for their shopping in 2025. And six out of 10 told us that, because of their AI experiences during the holidays, their expectations for customer experience have increased.
Speaker 3:So, whether it's the type of service I get the ability to find my order, getting to the right product at the right time using product search, all of these different capabilities and one of the primary reasons which again, no big surprise, no big aha here, but a lot of customers are now consumers are using AI to find the best deal on a particular product. Right consumer purchase decision is price, especially if it's an item that's available at multiple retailers or there's a like version somewhere else from another brand. So those criteria will always shift a little bit depending on the demographic and the category, but we certainly see experience, price and convenience continue to bubble up on why a consumer is going to shop from a particular retailer, and AI can help with all of those.
Speaker 2:I guess where that set of statistics pointed for me was that because an overwhelming number of consumers are now using AI for their own utility purposes they're overwhelmingly pleased with the outcomes there that it presents a pretty compelling case for retailers almost any vertical really to jump on this bandwagon and get it right quickly, Because those expectations aren't going to shrink, they're just going to continue to increase. And I, uh, there you go, you can't see this cause it's an audio medium, but, uh, Ed was making the universal signal, for it's all about the money.
Speaker 3:Uh, audio medium. Then why did I put on a proper shirt for this it?
Speaker 2:looks so good too. I'm just going to state for the for for the record here michael looks smashing so and so do you ed well.
Speaker 3:Well, he was on a call with me an hour ago and I wasn't wearing this shirt, so should I?
Speaker 1:should I go into that, my, my little money? Uh, should I go into that a little deeper?
Speaker 3:for you.
Speaker 1:Sound effect, yeah, so so so, talking to a customer, current customer, great customer, using our platform, trying to figure out again where our AI that we have today will fit in and meet some of their requirements, to get them some of their outcomes and we were talking about how TalkDesk has grown. We have all these people and it's expensive to have all these people, all these people developing AI technologies. And this vice president said, ed, you have no idea the amount of millions that we've spent on AI and we're still trying to get it to work. That scared the crap out of me To hear a retail vice president say we are spending millions on AI and we still haven't figured it out. It's almost like I wanted to go well, let's work together because we already have it, it's already working, people are already using it and let's try to figure out where that leverage is. And so I think that's an.
Speaker 1:That's where the opportunity is of going out to these retailers that are like I mean, because everybody and their mother, you know started getting budget money to do AI. So, like Michael said, we've been around for a decade what they're trying to do. We know retailers. They don't do things fast. They usually spend too much money. It takes way too long. So I think there's a great opportunity if we get the word out correctly and we position this correctly that retailers that might have some type of AI today that they could work with us to leverage our fuller package of AI to help support and make their AI more efficient, and maybe they'll spend less millions of dollars developing it and maybe spend it with me. Well, there you go.
Speaker 3:Since I'm the survey guy, a couple of I'm doing a survey.
Speaker 3:Yeah.
Speaker 3:So our pre-holiday survey we did speak to several hundred retailers and a couple of interesting data points from those surveys and those conversations is that 56% felt that AI was going to give them a competitive advantage, but there weren't 56% using AI yet.
Speaker 3:So we wanted to know first, who thought that this was an advantage and we still only see one out of two, so not everybody has bought into this. We still only see one out of two, so not everybody is bought into this. And then we wanted to know well, if you haven't started to invest in AI for your business in some way, shape or form, why not? And the two things that clearly came out of that were budget and resources. And resources being having the skills internally to be able to take advantage of the technology. So a recognition that A I might not have the resources or the budget of some of the big boys and we know who those are you know the Amazons, the Walmarts, the Best Buys, home Depots, et cetera and also understanding and having a self-awareness that I might I don't have the right people in place to do this now and that I need to figure out first before I go and invest in all this technology.
Speaker 2:You know the answer to this question would solve a lot of stuff for a lot of people out there. But if it is a people question, if it is a people problem, we don't have the right people to manage, deploy and fully leverage this technology. Who are those people? What is it they're looking for? Because you don't just go out and put out we need an AI specialist. That means nothing. So in your world you're talking to CX leaders, you're talking to people who are trying to optimize for the full customer relationship out there and they need to solve for a people problem, to get the right people there to manage these kinds of things. What are those positions? Who are they looking for? What would be the job title or job description for somebody like that that could come in and solve for that?
Speaker 1:That's a good question, because you have some retailers out there that love buying and you have some retailers out there that build everything, and you probably know who some of those are. I do know that a very sought after position is data scientists to be able to have that as a core, have that as a core. But I mean these titles of you know, ai guru, ai Sherpa they're just popping up all over the place. But it also pops up all over the place depending upon the industry. When it comes to you know graphic arts, when it comes to you know transportation logistics. So I guess it's you know. It boils down to I do not know what the criteria is to say I am fully knowledgeable on agentic AI, please hire me. But, by the way, I cost a lot of money.
Speaker 1:I think a lot of retailers are taking the people they have today saying, hey, all that stuff you guys did on agentic AI, well, guess what? Now we need to move into agentic. So I hope I have the knowledge base that can move into the phases of AI to get us where they are today, and I think there's a lot of that going on versus going out and hiring, bringing people in. Of course, there's plenty of the GSIs out there to be happy to come in and do it for you at an unbelievably extravagant price. Not to talk down about our friends. But no, I think a lot of that has to do with. I think most retailers are just pulling that knowledge base through what they already have today into creating more of what they have.
Speaker 1:So I'd like to add that there's.
Speaker 3:I think this is going to evolve and I'm going to give a very specific example. So last year, and at the end of 23 as well, there was all this conversation around the job title of being a prompt engineer right, so being able to understand how do I actually speak to the AI to get the results I want out of it. And then I had to be very knowledgeable of how the AI actually behaved in order to get the results that I was looking for. And today at least, what you're seeing in 2025, when we think about agentic AI, there's conversations, and this was a very specific sidebar conversation I had at nrf is this idea of prompt engineering may become extinct because, as ai becomes more intelligent and more natural language and more agentic, I can just be a business person that actually understands what I want to have as an outcome and I can speak to the AI. That way, and as the AI evolves, then I'll be the outcomes I want, because it now is all about natural language.
Speaker 3:I always love to use the example of 10 years ago, when Amazon came out with the Echo and Alexa. It was a really poor experience, and that's because we didn't know how to speak to Alexa and now, because if you weren't very specific it was like sorry, I don't understand. But now you just have to speak to it, naturally, and it understands what you're looking for. So I think there's going to be a constant evolution of what those roles are. You know and maybe this is me thinking of just you know my own career and how things have evolved. If you understand the business, you're willing to learn and you're willing to expand, those are probably the kind of people that you're looking for for your business, that you know there's the idea of being an athlete going to best suit your business as things change and evolve, and you don't need a bunch of data scientists everywhere. Of course you need some, but I would say that's not the future of the general growth path of the people. Yeah, I mean, look when web developers was huge.
Speaker 1:Before the web there was no such. When web developers was huge, before the web there was no such thing as a web developer. Then all of a sudden it was like internet, oh, I need a web developer. And you know a lot of those people moved into those positions. But that evolution, I agree, and I think there's a lot of retailers out there today that have great people on their staffs that are hungry to learn for that, and those are probably some of the best people that they have.
Speaker 2:Yep, yep on their staffs that are hungry to learn for that, and those are probably some of the best people that they have Yep, yep Ability to learn, grow and morph as the situation dictates. Well, ed Michael, I really appreciate this. It seems like, because of all the effort we expended to make this conversation happen, we should be talking for the next two hours, but I don't feel like that's legitimate or kind to your schedule. So, but I would like to get together, maybe as a follow-up give it another six months or nine months or something like that Because I think it'd be really interesting to hear how talk desks, approaches, have evolved over that period of time, because the evolution in AI, the capabilities of it, are expanding so fast that what we see now, we could all be prompt engineers of yesterday in six months from now. So we'd love to see how that looks when your times and schedules permit, but for now, thanks a million. Really really appreciate this.
Speaker 1:Well, thanks for having us and I look forward to our next conversation so we can talk about all our successes.
