Why Contact Center AI Projects Fail After the Demo

Max Clark: [00:00] Let’s see here.

Uh, usual spiel here. Um, I should actually just write this down and read from it from a script every single time so I don’t forget anything. But, uh, if it makes noise, it will make noise. So turn it off. Um, team Slack, email, outlook, calendar, phone, doorbell, dog, um, you name it, it will, uh, it, it will, it will make noise.

My dog is completely deaf, so she doesn’t even realize what’s happening anymore, so she’ll be over in the corner until she wakes up. And then when she wakes up, she’ll want to talk to everybody. So, we’ll, usually it’s, usually it’s not a problem.

Jim K Tennant: Interesting. Interesting. Yeah. Lemme know if you hear anything on your side.

Um, but yeah, I.

Max Clark: We didn’t get any, uh, I don’t think I got [01:00] a case studies here, so you’re just gonna get in the hot seat, bring it on. Um, you wanna do any burpees or pushups to get, uh, get the, get in the mood.

Jim K Tennant: All good? All good. 7 0 2 here. Cup of coffee. It’s, uh, actually a little soup bowl to Philadelphia.

Where are you based? Where do you live? Uh, Northern Cal. Northern Cal. Just outta, just outta the bay area of Walnut Creek.

Max Clark: Okay. Yeah, there I, uh, I was working for this.com and we bought a company that was based in Concord.

Jim K Tennant: Okay.

Max Clark: Yeah, we we’re

Jim K Tennant: two minutes from Concord.

Max Clark: Yep. And I, uh, they wanted me to relocate to Concord to take over operations up there, and I said, no, thank you.

Mm-hmm. And it’s just,

Jim K Tennant: it’s a different little city. Different little city. You know, Walnut Creek have been here. Uh. But almost 31 years. Um, I’ve seen it grown. Uh, we like the weather, proximity to, you know, mountains to the city. [02:00] Um, it’ll go warm here, but we’ll, we’ll go, go East Bay, go, uh, go west, I should say.

And then, uh, hit the city and it’s 20, 30 degrees cooler over there.

Max Clark: Well, to to date myself, um, this acquisition happened, uh, right before what you would really consider like the dotcom crash Uhhuh. So as I was going back and forth, the Bay, San Francisco went empty. And so I was driving from Oakland Airport.

I would come up and then drive through the, you know, through the tunnel to go into, uh, into Concord. Yeah,

Jim K Tennant: yeah. Like

Max Clark: the tunnel. Yeah. And the Bay Bridge would, I mean, there’d be, there’d be no, there’d be nobody out they bay. It was like an apocalypse, you know? There’s nobody on the freeway, nobody on the Bay Bridge, you know, and it was, it was very eerie.

Um, yeah, that was so long ago. It’s crazy to think. Um, what else should I tell you here, uh, on the left side of your screen. So, how would I do this? So this is my right. So I guess [03:00] over here you’re gonna see a little icon with a number and it’s gonna say like a little purple box with uploading. It’s gonna say like, you know, some, some percentage.

Correct. That’s the, um, that’s the important thing when we’re done, the platform records locally, so if you, like, if the video gets pixelated or whatever, you’re just getting a stream of me, not the actual recording. So don’t worry about it. Mm-hmm. Ignore it. Just go about doing what you’re doing.

Jim K Tennant: Mm-hmm.

Max Clark: Um, when we end and I stop, all you really have to do at that point is just wait for it to say a hundred percent.

And that’s, and that’s it. So your, your, your duties will be over at that point.

Jim K Tennant: Yeah, it looks, it looks pretty current. It’s, uh, bouncing from 97 to 96. 97. Yeah.

Max Clark: I’ve only had a couple of times where somebody was way behind, so it’s, it’s, it’s not a, it’s usually not a problem. Um, let’s see here. Um, trying to think what else I should, I should

Jim K Tennant: tell you, warn you [04:00] about, um, we doing any, uh, slideware shows?

It’s just mostly audio. Okay. Nope. Okay.

Max Clark: Nope. Well, I mean, the video’s recorded, so the video goes okay. But I don’t, uh, I’ve stopped doing slides. It turns as soon as you go into slides. It’s just people reading from the screen and it’s, yeah. It’s, it’s kind of like, what’s the point here? Um, any questions, comments, concerns, things?

Do you,

Jim K Tennant: uh, edit once, I mean, it’s, it’s obviously it’s recording, but do you go back and cut out or edit or shorten or, yeah,

Max Clark: we will, uh, go through and clean up. I had an editor that was really aggressive, and now it’s just about like major, taking out major stuff. It’s not about trying to condense or shorten.

Uh, typically, typically we go for about an hour, hour and a half. Just depends on what kind of cadence we get into. I find unless you’re start jumping around to different, like, like different topics where you’re then having three or four conversations mm-hmm. Like one conversation after about an hour kind of runs.

[05:00] Um, but yeah, it’ll go, it’ll go into queue. Um, I don’t know what the current pipeline is in front of you. We publish once a week. Denise will ask you for your headshot and a bunch of other information for you. Okay. And then when it gets like two weeks out from actually being published, it’ll be a FYI. This is getting, getting ready to go.

Great. Great. Um, I’m trying to think what else I should tell you. Yeah. I mean, otherwise it’s pretty straightforward. Great. Let’s have fun. All right. Well, Jim Jink, thanks for joining the hot seat here. Um, I.

I, I, I will rail about marketing buzzwords. It’s just, I’ve been in tech for too long at this point that everything just turns into a buzzword at some point. And before chat, GPT launched, and we have this kind of current version now of ai, there was a lot of people branding AI that were in this like machine learning space.

And, and the tech nerds, you’d be like, that’s not really ai. And you’d be like, well, you know, it’s whatever. And the marketing department, it’s like ai, ai, ai. But you know, recently we’ve had this just [06:00] really just hockey stick explosion. It feels like tech has just gone vertical. In now what is the AI space?

Jim K Tennant: Mm-hmm.

Max Clark: And observe. AI has been in this for a long time now, and I’m, I’m, I’m curious to start, you know, from your perspective, how have things changed for you? I mean, we can, let’s, let’s call this like, since like 20, early 24 to now, you know, late 25.

Jim K Tennant: Mm-hmm. Mm-hmm. Mm-hmm. Great, great. Great question. Um, having a chance to be here in the, in the early 2020 fours, uh, having, coming from another, call it AI company, I think the fundamental change we’re seeing Max at the technology.

The technology in itself. Um, and when I say that used to be, Hey, we have our own proprietary LLM, that was a good thing, right? Companies said, wow, you really have that. Mm-hmm. To today, you hear words like optionality. You’ll hear, [07:00] multimodals, you’ll hear, um, you know, multiple. Uh, LLM models, LLM models talking to each other.

And then last year, AgTech came out and you talked to, and if you’re, you’re, you’re watching Nvidia, CEO, talk about the next gen, you know, ai, what, what does that, what does that mean? How do I apply that? Um, not every use case would facilitate using AgTech, but to your point, that hockey stick that grows, um, what I might say today to a client and say, yeah, I’ll show you my roadmap, um, might change tomorrow.

I mean, it’s that, it’s that scary, right? So the old days where you used to say, Hey, let’s do QBR, blah, blah, blah. If I did a monthly business review, uh, we’re still talking about, um, new technology that’s there. And because of that, I think what we’re gonna see, or companies [08:00] that have digested ent. And the foundational pieces of new technology move, much swifter provide, uh, perhaps a better consulting practice to their clients.

Uh, and along with that comes customization. Um, but yeah, we, we can talk about a number of things, customization, scale, how you do that for everybody, best practices. But fundamentally, I think the technology has changed and continues to advance. Um, and I would predict, I mean, we’re not gonna see that slow down.

So it’s, and it’s, you know, fasten the seatbelt and brace. And how, how do we do that? How do, how do we do that? How do we scale with that?

Max Clark: I’ve lived through a few hype cycles. Uh, well, a few I’ve, I mean, it, it feels like everybody’s trying to create a hype cycle, right? Like, is 5G really a, a transformational technology or not?

You know, who knows, right? But, um, the. [09:00] The early, you know, accessible access to the internet in the late nineties Right. Was transformational now early entrance into the space. And what they thought was gonna actually be the dominant, you know, business outcome wasn’t the case. You know, we went through like you saw the.com boom and bust.

Um, but it was, it was transformational. And I feel like, um, you know, search, going from indexes and lists and discovery to relevance with Google mm-hmm. Was another really big transformation. Mm-hmm. And it, it’s, you know, redefined how we interact with technology. And what’s exciting to me about AI has been not necessarily all of the, like this is, you know, the end of end of work and into this and end of that.

It’s, it’s really this new interface that we now have to interact [10:00] with. A platform, right? I’m not gonna say a computer. I’m gonna say a platform where, you know, going from having to learn how to search on Google to being able to have a natural conversation with an AI tool, and then have that AI tool be able to understand what it is that you’re actually looking for, and then go out and be able to do work, and then be able to refine that conversation and have, and then have that across different modality has been, uh, has, has been incredible to watch.

And you, you know, and, and I get, I’m very optimistic, like I’m an optimistic technologist. Like this has changed the world. We don’t know what’s true and what’s not true at this point, but we know it’s changed the world, in the world of like, you know, um, in the, in the business world, the B2B world, we talk about service providers and we talk about platforms that are coming after attacking business problems.

You know, what, what are business problems that we’re trying to alleviate and improve? Business efficiency, productivity, revenue, generations, these different things. And you know, AI has been out there. Especially in the [11:00] contact center for a lot of space. Mm-hmm. And so now I look at this and I, I kind of immediately go to like, how does somebody start separating fact from fiction of what’s really happening and what’s a hype cycle?

What’s not the hype cycle and what’s, what’s everybody saying we have AI versus, you know, oh, they’re coming for your jobs. You know, that’s not really the case, but like, you know, what, what is, what’s flushing out right now, um, with that for ai?

Jim K Tennant: Uh, great, great segue. And you talk about, we’ve been doing this a long time and you talk about the hype cycle from, uh, prem to cloud, the internet in general, this thing called unified communications, uca.

Oh, you know, hosted phone systems. Slowly people adopting that, right? That turned into, to your point, contact center, um, contact center, the cloud. Well, what does that mean? Doesn’t mean that premise is going away at all, but the adoption of, do I embrace it? I think it’s a challenge for [12:00] every SMB mid-size enterprise client, and as we get into this new say, realm of the next generation of ai, everybody wants to say at talk it.

Um, you know, you and I talk to technology advisors every day and what I can continually see is, is how do I embrace that as a ta myself to be able to say to my existing clients, Hey, it’s okay. This stuff’s growing. Um, it’s okay. Meaning I’m here for you to talk about what I’m seeing in this growth of all the suppliers that are behind me that I can represent.

We’re seeing absolute traction in the contact center space because you think about. Let’s, let’s don’t just don’t talk code here. You think about anything that we do in any given day. Who are we calling? Who are we talking to? My auto insurance company. A [13:00] vertical, a healthcare, uh, financial. Uh, I was, I was just online earlier talking to somebody else before, before this call today about some finance stuff.

We’re experiencing contact centers as individuals every single day. And one can say, oh, AI’s gonna take people away. And yes, we’ll hear stories of, you know, Salesforce riffing people because he wants to replace with ai. Everyone, it’s, it’s fine. What are we gonna see? Some of that. I think the idea is that you’re augmenting, you’re helping particularly, I mean, our history in eight years of this has been we’re gonna augment that, that agent, right?

We’re gonna augment that. We’re gonna help that agent be a better agent. Uh, I might be ramping up in my new role. And, uh, I get a call. And how nice is it to get a little popup screen? Hey, you’re talking to Max. He’s called in X amount of times. He’s called in about this order. Hey, max, are you calling about this order as a [14:00] person, as a human being?

Max Clark: Lemme ask you. Uh,

Jim K Tennant: I can open eight other windows and scrape all that and try to figure it out. Hang on one second, max. Let me, let me, let me see if I can find this. Mm-hmm. Or, you know, I’m, I’m seeing a bunch of things about your account. Which one are you talking about today? Right. How do I make that agent a better agent?

Uh, and, and, and in a post world, it’s, um, as a supervisor, I can now see how Jim was handling that call. You know, is he compliant? Is he asking the right questions that we want him to ask? How, how was the emotional sentiment analysis, right? On that call? Is there an opportunity for some coaching? There are all my people doing the same thing on a Friday afternoon when they’re tired.

AI doesn’t necessarily have to be in, in the front end of the contact center. But we are seeing that as a monumental shift of acceptance. Okay, let me say that in a different way. We’re seeing so [15:00] many suppliers and technologies and why? Because people are creating these, you know, I’ll say it, the voice bots, right?

They’re tired of the legacy IVRs. They’re tired of pressing one, two, like, right? So, you know, you might experience a bot after you hit two or three because a client’s trying to see what AI would be in their world, right? There’s opportunities now where you don’t have to rip things out and you wanna augment what you have without going in full board.

So you see that more and more in the context center space.

Max Clark: So I wanna talk about, I wanna talk about virtual agents mm-hmm. And, and this, but before, I wanna get, I, before I get there. Uh, something very common I see when I talked with a company about their contact center is, you know, globalization has already hit most organizations running contact centers.

And they’re, what I mean by that is the BPOs are already deployed. You see staffing that’s already changed, where, you know, commonly they’re running like a level [16:00] two, level three in-house and they’re running 10 20 x the amount of staff through A BPO and, and, and they’re doing that. And so, you know, for my viewpoint, when we start talking about like, when people have this, this, you know, AI’s coming for people’s jobs in the context center, our conversation, you know, the first thing that comes to my mind is like, which job?

Right? Like is, you know, and, and it, you know, I don’t say this to be like really controversial or hyperbolic about it, but, but it, it, how much of this is really the risk to the BPO? People staffing inside of the BPOs. And how much now is this really a replacement risk for companies that have contact centers that already have, you know, like a tier two, a tier three in-house escalation path?

Like what? Yeah. You know, and I, I’m okay, I’m gonna leave it there and I’ve got like a few follow-ups. We,

Jim K Tennant: uh, we, we, we see exactly what you’re describing. Um, yes, there, there is some, there is some attrition with that person up front, uh, being replaced [17:00] by a voice bot, if that voice bot can do anything and everything now that human can be.

But I would be willing to say that, you know, more than, um, 20, 30% of these companies are putting that voice bot up front so that the agent can actually be more productive on the back end. If I. Authentication and mundane tactical questions, and that can be him. And I can, you know, have calls routed to me that best fit my skillset.

I’m gonna be a better agent, and quite frankly, I might be a happy camper and, and stay with that company longer because I’m not doing this monotonous work all the time. So take that away, not taking my employment away. Right? But you’re helping me by be a better agent by taking some of those mundane tasks away.

And quite frankly, you know, people we’ve seen, and even you, you talk about generations, um, Zs exes, uh, hey, I’m a boomer. I’ll admit it. I’m the guy that’s like agent, agent, agent, right? Mm-hmm. I’m not, [18:00] I, I’ll be honest with you, I talked to one the other day, beginning of the business, and I’m like, okay, I’ll give this thing a try.

It sounded so human-like. And I’m pushing the envelope. I’m, I’m interrupting it. Yeah. Yeah. And all these background noises are happening, and it was a, it was a solicitation for a donation. Because of the bot. Because I even said, are you a robot? I mean, it’s, it was talking to me. I made a donation. I mean, was that, it was for good thought.

This

Max Clark: is, yeah, this is the funny thing about it, right? Where it’s like, I, I’m, I’m, I’m not too dissimilar here. I think we’ve had very similar, I’ll have, like, we can, we can go into some rants about this stuff. I, I think the missing piece for a lot of people is when you call a co you know, when you call service, you call a sales line, like you’re calling to do something, you need something, right?

And, and what I care about is can it get com, can I get what I need and how fast does it take, you know, how long does it take to get there? And [19:00] the stuff that makes me crazy. And all these ones, first off, companies that don’t have virtual hold at this point, I hate you, but, um. Which also probably a good argument to have, you know, some sort of, you know, AI in line, but it’s like, it’s like when you have the fake agent, right?

Where you’re just like, you know, you’re talking to a robot and then it gives like the fake keyboard clicking in the thing. It’s like instantaneously, like, I’m checked out, I’m done. You know, or, um, that was great back in the

Jim K Tennant: day, right? That was like, Hey, this is kind of cool. Yeah. But it’s acceptable now.

Max Clark: Yeah. Or, or you see these like half integrated systems where it’s like, if you ask me to authenticate and I have to jump through like five hoops to authenticate, and then I actually get to a human, and then it’s like you have to re-authenticate the whole thing. It’s like, what are we doing here? You know?

Like, and I, I feel like a lot of these things, like customer satisfaction, the customer satisfaction bars are relatively low, like, like really simple things that landmines you’re stepping on, that are causing problems that, that affect your CSAT at this point.

Jim K Tennant: Correct? Correct. Yeah. We, uh, when you talk about csat, you know, it makes me think of, you know, what are the key metrics of [20:00] success?

Right. So, so you and I both read the articles. I think MIT came out with one not too long ago, month ago. X percent, like 90 plus percent of these, these AI projects fail. Well, you have to dig into that. They’re failing. You know, to your point, am I looking at the right metrics for success? Everyone wants their CSAT to go up, but if I expect that from the bot that I haven’t really identified what I want it to do and set that up, then it’s gonna fail.

It’s gonna fail. It can be the best bot in the world, right? Um, so the key with that we have found is not just provide clients information of if they’re saying the right things, but, but why dig into that a little bit further. What, why is that happening? Um, um, you know, why aren’t they saying the right things?

Uh, dig into the, the behavior of that. And then do changes internally because of that, then that [21:00] will of eventually you, you talk about the metrics on csat, that’s a whole nother half hour plus. Um, is it, is it, is it, you know, answer time, right? Handling time, whole time. How many times do I have to transfer it?

Why are those clients dropping off if, if we don’t know that as a, as a, as a client? Those are, those are key things that you have to, you have to identify.

Max Clark: Okay. So before we get to back to deflection here, um, you, you, you touched on this briefly, which, you know now I think universally we’re talking about like agent assist, right?

How do you use AI to assist the agents and what can be done? Right. And, and, and I’ve always loved this approach to contact center because when you start really like digging into contact center metrics and the operations of the contact centers and like the business of the contact center. You get into some really interesting stats and numbers and it’s like, you know, what’s your retention rate for your agents?

You [22:00] know, like how, how long does it take you to hire, train, and make an agent productive? Mm-hmm. How long do you retain that agent once it’s there, what’s your churn rate within that? And you, and you look at these things and, and of course there’s a lot of technology that then comes out that, you know, WFO and WFM of like, you know, and, but that doesn’t solve that problem.

That just tells you like stats, like how many people do I have to have online and then how do I adjust for seasonality and these sorts of things. But, you know, I’ve, I’ve always kind of thought it was really interesting when you start digging into, you know, seasoned context center leaders that really understand like their fighting churn as like their number one thing that they’re really dealing with inside of the organization.

Because if you can take, you know, an average churn rate of 30% and, and cut that in half, like it changes everything for you. Correct. Correct. Um, and so the assist is really. The assist is always fascinating to me. And the second part of it is, um, like, is onboarding and training because, you know, like observe AI didn’t start with a, we’re gonna give you a virtual agent.

You guys started in a very different [23:00] space and have kind of moved forward there. Can you talk about that a little bit?

Jim K Tennant: Absolutely. Absolutely. Appreciate you bringing it up. You’re right. Agent assist is one of those. If you don’t know what you’re looking for, how do you, how do you coach that agent live?

How do you coach in real time? Um, observe eight years ago started, started with post post analytics, uh, with the mantra of, uh, guess how many clients are recording the calls? You know, we’ve all heard that this call is being recorded for training and really, really, are you really monitoring it? Are you really gonna go back there?

And, well, some were, but we found, and reports would, would confirm this. Less than 2% of those calls are actually being monitored. Someone was actually going back, looking at listening to and now technologies, and there’s the gongs of the world that kind of look at those and analyze those calls. Our whole mantra is built on, give me your call recordings and continues to be, uh, and I’ll QA those, right?

I’ll provide quality assurance. [24:00] So much data was coming back that grew into auto qa that grew into summarization. You know, I, I don’t have time to read the data that I have now that I didn’t have before. This is great. Now what do I do right? Much less where do I store it and who has it, and all those other things.

Um, but we grew knowing that we wanted to eventually get into the agent assist, and then now early this year, launching the voice piece to create that entire infinity loop. But we, we built that on the basis of Mr. Client. This isn’t me trying to guess what, what your agents might be saying or what your customers are calling in for.

These are your calls. These are your calls. So, so tell you what, why don’t we do a proof of concept? What do you mean? What does that mean? Gimme your call recordings. Give, gimme a good set of call recordings. I can ingest to my, my model and let’s, uh, gimme a couple weeks. I’ll come back to you and, we’ll, we’ll [25:00] do a report, a readout.

Mm-hmm. We’re seeing these elements, we’re seeing these behaviors, whether it be per day or per metric, or whether it’s insurance, hospitality, services. Right? So what we’re doing all along that is building our parameters for the next one in that vertical. Right? So we’re getting more defined in every vertical because of.

What we’re seeing and, and specifically designed for the contact center space, um, not for a car dealer space, not for, you know, banking industries, credit cards. I mean, we, we, for contact centers specifically, not sales teams. Now, who knows where the future might go, but right. Contact centers. And that grew into the agent assist that we’re talking about where now, whether it’s summarization, uh, something we call knowledge ai, where it’s scraping different CRM systems within the system.

So you’re not having that lag of, hang on, let me, let me go check. I got eight windows [26:00] open here. Lemme go check to see where that order is. Lemme follow it through. Imagine, you know, proactively all that things, all that information is popping up for you on your screen to help your, help your client quicker.

Call right higher satisfaction. Um, shorter handling time like you mentioned. You know, they want to get in and get out. Uh, they don’t want to get in there and talk about weather and sports and all that. Just gimme away information. I’ve got a busy day and you know, I gotta go kind of thing.

Max Clark: Um, I find an expectation that, oh, we’re gonna go out and buy an AI tool, and they we’re just gonna turn it on and it’s like, it’s like, oh, you know, it’s Thursday and we sign a contract and, you know, Monday we can just flip a switch and it’s just gonna take over for us.

And I find still people really surprised with how much configuration and training and, and calibration, you know, that you have to do on these things. And. Now it’s getting a lot better. And I’m, I’m curious when I want, wanna talk to you about this more, [27:00] but, you know, a few years ago, like you had to go through and you had to like, create these giant flow charts and process and decision trees and like input output.

Like if they asked this, give it this response and, um, the amount of effort that that would take to actually go through and try to train or just understand your own internal knowledge base of what knowledge you were exposing to this tool of like, is this accurate? Is this inaccurate knowledge? How is that changing and how, how are you tackling that with customers that are onboarding?

Jim K Tennant: Mm-hmm. Mm-hmm. Great question. Um, again, you know, eight, eight years ago, uh, post analytics, uh, qa, auto QA was born based on call records. We had sound technology continues to be, we continue to advance and what kind of data we want to output there. Something we call moments, Mr. Client, we’ll, we’ll create moments for you specific to different events that are happening in your world.

To your point, we’ll, constantly we have a team that’s on the backend calibrating what those moments might be. [28:00] Um, how’s that data? How are you using it? Right? So to your point, a lot of clients needed coaching on and or QBR is on how are you using the data? What are you doing with it now advanced to the front end?

What you’re referring to, moreover, is not even probably two years ago, um, what would be a two, three minute conversation would take weeks to build. Mm-hmm. Um, the coding on it. I mean, I even remember looking at the coding for a two minute conversation. 1200 lines plus of code. Right. Oh, you didn’t want it to say this accent.

Oh, you didn’t want to say this word. Well, okay. Gimme a week or two and I’ll get back and I’ll get that corrected.

Mm-hmm.

Today, today, um, without code, I can go in. Um, you know, I’m, I’m in the West Coast, a big San Francisco Giants fan. Um, I created my own. Bought my own voice, bought five minutes by simply, Hey, make it sound like the, [29:00] the, the sportscasters.

Um, I just need to know my prices, my seats particularly, you know, any, any kind of giant dodger game. Um, and, um, you know, best price and seat, blah blah within five minutes, it’s combing the web, it’s combing everything that, uh, I may have given it. I could feed it documents. Um, I didn’t do anything. I didn’t feed it anything Max.

And it, it came back in a, in a sportsy voice, um, told me the seat price, the availabilities, um, I didn’t have to write code for that. So, to your point, what we’re seeing, and part of that is AgTech and part of it’s the use of multiple large language models. So, so companies that are providing these services need to articulate to their clients.

How they’re doing that. Which ones use what because the client doesn’t know Is, is Claude better than Anthropic? That’s better than [30:00] chat. That’s better. And to, you mentioned platform earlier and I appreciate that because what, what point systems do are one thing what platforms could do or a number of things, but you’ve gotta be able to articulate to your client, look this way too much to digest.

Okay, let, let’s start where the biggest hemorrhaging point is, right? And then the whole crawl, walk, run theory. Let’s get your confidence level up with what we can do, what that can provide. And then you grow with that client. You grow with client. Yeah.

Max Clark: Issue with LLMs is still hallucinations. We still see.

It, you know, data leakage.

Jim K Tennant: Mm-hmm.

Max Clark: Uh, we still see if you start talking about agentic, right? Like you’re gonna, you’re gonna have your AI go out and connect and, you know, I mean, look, agentic with MCP, there’s some, some amazing technology of it there, but now you start talking about if you’re running a brand and you’ve got customers calling into your [31:00] brand, like there’s another set of, of safety rails that you’re gonna want to have on these tools.

You know, like, is the LM you know, hallucinating, you know, talking to somebody, or can they break out of the rails? And you see these examples of like somebody, you know, chatting with some bot that convinces it to sell ’em a car for $1, right. Or, you know, like it, and then of course, do you trust the information that the model is, is being fed and looking or interacting with, and then building.

So how do you, you, how do you go and, you know, take the, I mean, I love, I love the like, crawl, walk, run of like, you know, we’re gonna do. You know, QA QA and recording and analysis of existing calls, and then we’re gonna take and feed that into assist. Mm-hmm. Because then you’re trading on your own opus of information.

Right. You would hope that after whatever period of time, you’re gonna be able to have like a repository of like, we think this is good interactions with good information mm-hmm. To then start, you know, um, you know, [32:00] providing some, some element of either assist or then ultimately like a deflection for, but you still have to deal with like stuff changes, information changes, products change, service changes, policies change.

Or maybe it was just bad data to begin with. Like how do you, how do people how create safety rails for themselves with this stuff?

Jim K Tennant: Yeah. You, you, you, you have to build it with, with the idea of the safety rails in mind. You, you have to be able to tell a machine when it goes a certain direction to come back or that that call is escalated or, you know what, I don’t know that answer.

Let me get you to a live person, uh, as opposed to trying to make things up. Those, those have to be documented inside the machine. Um, those guardrails have to be written for that. Something we’re doing to, to help with that is, um, you know, we talk about the voice agent up front, and if the client doesn’t want the backend or you know, they’re not ready for the agent real-time assist right away, we’re still QAing [33:00] those voice agents on their behalf.

So we’ll come back to them and say, we’re seeing trends in this area. Um, you know, if something changed in the business, should we change this voice bot? We’re seeing these things because to your point, they’re not gonna call into that 800 number every single time and have the identical experience every single time.

You know, they, they wanna know when it goes and maybe pushes those rails where things need to be changed. And to your point, um, compliancy rules, regulations, they’re gonna change, uh, on a daily basis. And the expectation is we’ve gotta get that word out to everybody today, not two weeks from now, but you know, the new message has to be delivered today.

So yeah, those, those are all things that, uh. Clients need to watch out for, to make sure the providers can look at those hallucinations and what those risks are and uh, be able to identify how far the conversation goes before it’s okay, you know what? Hey, that’s a great question. How am I gonna handle that?

Does it get escalated [34:00] or does a bot freeze up? Does that, when it stalls, do you get hang up and lose someone or you know, do you take the time to actually, hey, that’s a great question Mr. Client. Let me, uh, lemme get you to someone who can help you, you know, and say it better than I can today. You

Max Clark: know, if you go back to the stone ages of contact center, right?

You had a PBX in an office, you had a different platform doing call distribution and queuing. You had a different platform that was doing IVR. You had a different platform that was doing call recordings. You had a different platform, right? There was like, you know, so it started out as just like, we have a really big PBX and we have to like, plug all these things into this PBX to make this stuff work.

And then as time has passed, that functionality then started integrating into the PBX. The p BX is all one IP native. Then we went to cloud, right? And so you go multi-tenant cloud and, and so now we have this, this aggressive consolidation, right? When there was there and, and there was this a split, there was like a UCAS vendor and there was a CCA vendor.

And now there’s like, you know, the UCA and CA platforms [35:00] mm-hmm. Are all coming together and trying to be this unified experience. How are, do you think about this in terms of like being a solution that overlays on top of another platform and gives you the ability to integrate with any platform and take on a customer regardless of which platform they’re on.

So there’s advantages to that. Mm-hmm. But then the flip side of that becomes these platforms are trying to move horizontal as quickly as possible to have everything verticalized within them. Not saying that that’s the best solution or that’s the best technology, but that bundle becomes more and more compelling just because.

Default decisions have already been made. Um, how, how does this landscape continue to change and, and how do you view yourselves changing with it?

Jim K Tennant: Excellent, excellent question. Um, you know, I look at, uh, the history at Ucas, CCAS and I can’t help think, uh, um, there are some people that are still in that stone age, um, um, and I say that with all due respect, right?

[36:00] They’re, they’re out there, they’re absolutely out there, and there’s reasons why they need to stay there, because whatever their data is or their business is, right? Mm-hmm. They, they don’t wanna go in the cloud and they wanna hug it and see it. And, you know, we used to tease people about that, you know, back in the day, not going, not going to the cloud.

It’s not safe. It’s not the, um, but no, you raise a good point. If, if I’m a contact center provider, um, and they’re, you know, wildly successful, providing native solutions, right? Um, growing organically. Be able to expand my portfolio with products. I wanna be able to do everything to everybody and every client.

Mm-hmm. Right. Whether it’s, you know, W-F-M-W-F-O summarization, the voice, I mean, we see it every day. Um, you and I can name off half dozen of ’em now that are wildly successful clients look at that and go, well, to your point, if I need something else different and or this provider, this contact center [37:00] provider doesn’t have such product, then I’m back to the PBX mind of now I gotta go buy a point solution for that.

Mm-hmm. And we see more in the mid-size enterprise clients as their needs grow, their needs get more complex, other m and as happen. Um, other requests come in from their boards, investors information that they don’t have their challenge because their contact center provider may or may not have that. So what do they do?

They, they go out and look for different point solutions. So I see, I see a, a cycle of what we saw a decade ago, customers buying more point solutions today to fulfill those gaps. And from the supplier side, this constant ebb and flow of everyone’s agnostic and we want to be part of that contact center play and the contact center player saying, okay, [38:00] yeah, we want to have unlimited amount of people in our quote unquote market play.

So our customers have that choice.

Mm-hmm.

I don’t have that Mr. Customer, here’s a website. Go point and click and you know, full integrations all in. Right? Um, but I see with that a challenge at the surface level of if I’m the client, how much of that can I rely on my supplier to give me? The best of what I can.

And that’s why we, we dive into the technology advisors, the TAs for us provide that agnostic, neutral approach, uh, to be able to say, Mr. Client, you can, you can have what you need without a forklift, right? You simply have this as an overlay, right? Um, and having that as an overlay allows you to keep that data, keep that AI technology with you wherever you might want to go.

We, uh, we, we pride ourselves on just saying we’re, we’re [39:00] neutral agnostic. We don’t get into the, oh, you’re moving to this context. Great. You know, we integrate with all of them. Yeah. It’s not, we’re not asked of, well, who should I pick? Who should I choose? Hey, we’re thinking about moving off of X, Y, z. We, we really don’t get into that.

You know, we get more into where are your pain points? Where are the business initiatives? What are you trying to accomplish? Can we provide that for you? Um, and of course, you know, we get into those, Hey, compare yourselves against this voice ai, or hey, this provider saying they can do this kind of QA summarization or wfm.

Now tell me about yours. And it’s, um, um, it’s a constant day-to-day thing. It, it’s a constant conversation when we get into this proposal stages, for sure.

Max Clark: It creates an interesting conflict for the contact center, right? Because the contact centers for the most part, I mean, now we see disruptors coming in, but for the most part, contact centers exist by selling name seats, right?

Mm-hmm. Mm-hmm. How many licenses are we selling?

Jim K Tennant: Mm-hmm. Mm-hmm.

Max Clark: And there’s some really big ones that then come to [40:00] market with their own IVAs, and they license their IVAs on a, on its per seat count. Right. And it, it’s like this, you know, and of course the reason they’re doing that is because they’re trying not to cannibalize their own revenue.

Correct. And even though now they wanna offer this functionality if they’re being asked for, and that they know that the market is moving towards, it’s like, how do you offer this functionality without cannibalizing your existing revenue? You know, and it creates this really weird conflict internally, and I look at this and say, okay, if the promise of technology and computing, you know, especially in a cloud environment, becomes elasticity, you know, why, why are people not just immediately going to consumption based?

Right? Maybe, maybe as a customer. There’s examples where I want to have fixed capacity and I want to have capped capacity, but that’s a decision I’m making that says I wanna only spend $10,000 a month on this, or whatever the number is.

Jim K Tennant: Right?

Max Clark: Um, and that’s a different, that’s a different construct from saying like, oh, you know, we’re looking at, [41:00] you know, trying to do deflection for.

Password resets, which we know we get x amount of percentage of our calls or password resets, and we can put that into an IVA and we can do the password resets and the IVA or, you know, I mean, pick any industry, you know, pharmacies, right? You know, like, I wanna do a prescription refill. Like, you know, like there’s, there’s, there’s things that are relatively simple, straightforward that occur, you know, a million times a day for you that are really low hanging fruit to take away and, and try to automate.

But it, it, it does feel like there’s this disconnection of how a lot of this industry is approaching it that’s not really customer friendly or customer centric.

Jim K Tennant: Yeah. Yeah. You, you, you, you bring up a good point. Um, people are bleeding, hemorrhaging, they’re trying to fix different needs and problems. And with the use of new technologies, let’s stay on the conversation of IVAs.

We’re seeing more and more people use that data. [42:00] In a proactive way mm-hmm. To help the business in, in and itself. Lemme give you a story. I, I’d love to have, um, I, I love working on the yard, on the weekends, um, um, using all kinds of tools. And for me, any new gadget for the yard is a, is a toy. Um, we’re working with a client and the client was saying, um, to, you know, going back to your point on consumption model versus a, a license, the seasonality is important, right?

Businesses say, you know, I’m not gonna do that license. I’m trying to do the ROI on that piece, and I’m trying to do a TCO on that. That doesn’t work. Mm-hmm. So, so there’s more and more IVA clients and suppliers that are doing, doing a consumption model. We, we have our, we have our own. Um, but the idea being, okay, you use this amount of minutes, what’s happening?

Um, Mr. Client, we saw a spike in this particular month, thousands of calls coming in per day. Oh, and guess what they’re calling about a particular. [43:00] So it’s a yard company. They’re called about a particular sprayer that they purchase that clogs up after so much use. Mm-hmm. Right? So they’re out there doing their yard work or whatever, and, and the thing’s c clogging up, um, as they did the analysis, their q and a, these calls, they learn that this particular sprayer had a problem.

There was a part in this sprayer, some washer device in the sprayer that after so many hours of use was, was getting clogged. Now they could have changed the IVA to say, Hey, Jim, do you have this sprayer? This is how you do it. And reactively responded to all of those, which they did, but they took it a step further and they said, uh, how many sprayers have this been purchased over the last X period of time?

And, and what does our point of sale system say? Do we have that contact information of that person who bought that sprayer? Imagine proactively now telling a voice bot to go call Jim. And say, Hey, that sprayer you [44:00] bought a few months ago, have you had a chance to use that yet? Uh, no. It’s still in the box.

Oh, you, well, when, when you use it, we found some issues, blah, blah, blah. You can either return it, recall it, or I can send you owner’s manual. You, how about that, right? Yeah. The ability to now use technology and an IVA to proactively help their business ensure they probably lowered some of the calls coming in about the sprayer, but more importantly, they have a very happy client that, wow, I’m impressed.

You know, you’re, you’re doing this for me, letting me know ahead of time. That’s, that’s, that’s pretty

Max Clark: cool. The call recording analysis, right? It starts from a QA standpoint, right? Where it’s like, okay, maybe, maybe it’s a compliance requirement. We have to listen to our calls and confirm that we’re, you know, adhering to our compliance mandates.

Or maybe it’s, we’re just trying to, you know, train and, and track our, our, what our agents are saying on our sa, you know. As a training exercise, but then you, you know, [45:00] you take a step back from it and it’s amazing to me. ’cause you think about it, like, if you have a customer base that’s calling you and talking to you, like the amount of information that you can glean from that in terms of product demand.

Jim K Tennant: Right.

Max Clark: Or, or you know, and, and you know, I, I hate, I hate the T-C-O-R-O-I, you know, thing because it’s like so disconnected from reality for so many organizations. But, you know, if you say, you know, what’s, what’s the TCO, what’s the ROI on establishing that your customers are asking you for a product revision that you could then come to market with and then sell and generate additional revenue from, like, like what’s that cost?

You know, calculate that for your business. Right,

Jim K Tennant: right, right, right. It, it kind kinda gets into the whole, you know, who are the buyers for all of this now. Mm-hmm. Right. It moved from, you know, we talked about the cloud earlier, moved from it. Uh, and I’ll age myself here. I used to talk to, okay, Mr. Client, they had a data contact and they had a voice contact.

And the voice contact did all the pbx, the data person did [46:00] all the internet and it, right. What are we talking to now? We’re talking to COOs CIOs. You talk about brand max, I mean, it’s CMOs the CM O is now involved in, I wanna see that demonstration. I want to, I want to, I’m gonna, you know, cater to these moments when they’re talking about this color or that color, or maybe my competition.

CMO wants all that. You’re absolutely right. Um, so we’re seeing a shift in, um, the buying process, the procurement process. Um, the personas of who’s involved in that sale go from IT to other business units in, in the organizations. Um, and quite frankly, other, other use cases, multiple use cases in the same organization.

Max Clark: Right. How much of this is being deployed as a reaction to pain versus, you know, a transformation and, and, and, you know, altering [47:00] operations because there’s like a revenue or efficiency play, right. Well, efficiency I guess would be a pain response too, right? So it’s like, how much, how much of this is still like a reaction to pain and how much of this is a transformational ’cause We’re looking to do business differently and go after revenue differently.

Jim K Tennant: I’d, I’d say more than 60% today is, is based on pain, pain points. Um, that, uh, and to your point, those 60% then turn it into, oh, can we do x, y, Z more? Can we, can we now turn it into the brand or, you know, business efficiencies, operational efficiencies. But for the most part it’s, it’s acceptance of, and then where we, where is that pain?

And so long as that pain and the, and the solution can be measured. Um, you know, as much as, as much as we don’t want ROI, tso, those are all very, very much part of all of that now, and the mindset of they’re still trying to figure out the financial piece to it, um, to make sure that’s solid. But it’s, it’s majority pain.[48:00] Majority pain is they continue to accept the technology and move forward.

Max Clark: Yeah. Let, let’s talk about pain points. Uh, you know, the elephant in the room, of course, is humans, right? We’re spending a ton of money having people, actually, we should talk about pain points in three chunks, right? So I don’t wanna drive this too much in terms of agent replacement in VA, right?

But, but I’ll, I’ll start there because that, that’s the elephant, right? So the elephant is, um, we’re spending a ton on our BPO. We have a lot of staffing, we have a lot of churn. You know, we went off the hamster wheel of like trying to hire and retain, right? And that’s, and that becomes. Uh, a big budget, you know, a big budget expense line item that people are contending with across everything, you know, comes down to that.

So that’s, that’s the elephant. Right. So let’s, we can come back to that part, but what, what are other common pain points that you’re really seeing that people are looking at and saying, Hey, we really need to solve this problem, and, and they’re coming to fix, uh,

Jim K Tennant: a agent, uh, uh, [49:00] retention, churn, attrition is big.

Um, sure. The inevitable, uh, I wanna shorten the handling time. I wanna shorten the escalation time, but, uh, in today’s world, it’s, uh, majority, I would say, uh, uh, volume of calls coming in combined with the agent churn. So if you can help me, you know, train agents quicker, faster, retain agents, so I’m not constantly replacing whether I’m a BPO organization or not.

I might have multiple BPOs because I have better success rate on, on retention in itself. Um. All the other ancillaries of shortening the call, uh, shortening the average handling time, all that, those all become metrics. If you can, if you can solidify the, the agent churn.

Max Clark: Um, for somebody watching, you know, and having this problem with agent churn, like, you know, pre-state, post state, you know, what kind of stats and, and um, [50:00] wins are you seeing with your customers,

Jim K Tennant: whether it be financial insurance, healthcare, transportation, transportation, hospitality, I should say made a new word there.

Yeah.

Um, I got it. We’re, we’re seeing, we’re seeing, uh, 40, 50, 60% different, uh, type of overhead operational efficiency reductions, uh, based on the call volume, based on what we can provide. Um, we’re seeing, uh, success in, um, um, all those verticals. Um. Again, the, the percents, you know, vary from each one, what they see operationally.

Mm-hmm. To then that expands into, okay, if I can handle X period of calls more, um, you know, what is that agent retention gonna be? Um, what is that customer satisfaction gonna relate to be? Um, but we, we see all those, uh, every day, every day.

Max Clark: How, how, um, you know, I get really simplistic with a lot of these things in my brain, right?

And I think if, [51:00] you know, in the IT world, right, if you’re working an IT help desk job, and like your entire day-to-day life is like helping people reset their passwords. You have a very miserable experience. You do not enjoy your job and you want out, right? And, and that becomes, you know, it’s horrible for everybody.

It’s horrible for the person that needs to reset their password, and it’s horrible for the person resetting the password. And I think about that a lot with the contact center as well. When you talk about agents in the contact center and just the, you know, I, I don’t wanna say like happiness as a metric of the agent, right?

But what is the actual work that they’re doing, and is it enjoyable and fulfilling or is it not? Right? And how so? I mean, how, how much of this is just elevating what the person is doing on the calls that they’re ta, you know, handling as an improvement in their, in their metric and churn and everything else.

Jim K Tennant: So you see, uh, anything from, if I can handle the call easier because I’m being helped, [52:00] being coached, if my, um, as an agent, if my wrap up time, my ability to summarize that is shortened. Oh, geez. I haven’t even thought of that. Yeah. Um, is, is if my wrap up time and my, um, uh, my accountability and my, uh, actionable.

Next steps are more accurate, right? I might be tired and, um, I’m, my, my note taking is horrible. My call to action after that, right? I’m supposed to summarize that and, oh, that’s right. I’m supposed to put that into my CRM system because someone else is gonna look at that as well. I might have had the best call, um, but my, my after work, right?

Is fundamentally incorrect or accurate scale that scale that, and, uh, you know, what’s the results of that based on who, you know, who looks at that or what the action items are. If, if we can improve all of those, right? We rolled all up to operational efficiency. If we can, if we can prove that [53:00] to a client that we’re gonna be able to help them with those pieces to that.

And then you turn that into a tangible action of what’s the value of doing that? That’s, that’s where the rubber hits the road. Um, I mean, agent attrition, churn. Uh, handling time, those are yada, yada. Those are all table stakes now. Now it’s a matter of what am I doing with that information? Mm-hmm. We keep talking about operational efficiency, Jim, so what, what does that mean?

It’s, it’s taking that data and putting it back into the business to have the business better perform based on information that is coming from their clients. So it’s not, it’s not, oh, I’m, I’m guessing at what I should be doing next. It’s taking the raw data, uh, inaccurate or accurate and, and doing something with it.

Max Clark: So, I mean, does the customer still have to go out and have their own data analytics team, data science team start, I mean, mining, by the way, I wanna come back to the whole wrap up, but to, to mine all this [54:00] data? Or have we gotten to a point where you’re now surfacing things or saying like, Hey, you know, by the way, here’s your dashboard and, you know, we’re seeing, you know, like, here’s, here’s a trend line of, of topics that we’re seeing that’s increasing.

You know, how much this is actually, um, intelligence in the platform versus, you know, I can go find it if I’m hunting for it.

Jim K Tennant: Excellent question. What, what we’re seeing too is a big trend. They’re, they’re getting this information, they’re getting this data, um, what, wherever they store it, we’re seeing, we’re seeing clients repurpose agents and saying, Hey, you, you, you know this data, you were an agent before, but let’s give you a different role.

Let’s give you a promotion where you’re gonna fundamentally look at all this data in this QA and help us, right? And, and, and be a, be a better. It’s not even a supervisor, right. It’s now someone, it’s almost a QA person looking at the data. Mm-hmm. We’re finding that we, we work with clients and I talk about the account management team afterwards, almost like a, a behavioral change, [55:00] meaning I’m gonna act differently if I know a robot’s watching me or if someone’s summarizing what I’m saying or missing.

Right. I might have a different behavior mm-hmm. Because of that, because I know my supervisor now is, you know, gonna evaluate me and, and quite frankly, I might be paid differently based on my accuracy. Um, but change management is big. So people are looking at that data, how they use that data, um, putting people in roles to review and qa that data, uh, to, to better the, the business.

Right.

Um, uh, so we’re, we’re seeing a trend with more and more like that because a lot of times we’ll have QBR and it. Wow. Didn’t, didn’t know that was happening. Thank you. Um, and you know, they’re writing notes. I’m gonna go talk to Sam about this, that, and the other thing. Um, but, uh, no, the reality is we’re gonna continue to see that, um, companies will handle that in [56:00] different ways of how they want it.

Some will ask us to manage that for them, continue to show us those trends, but what they do with those trends and that data is different for every client.

Max Clark: I read something a long time ago and it, it, the point was the success of the franchise model is being able to scale

Jim K Tennant: mm-hmm.

Max Clark: The business and the experience of your customer.

Mm-hmm. With the, you know, with to scale the experience and the consistency of the experience with the lowest paid workforce available, basically. Right? Like, if you’re talking about fast food, you’re gonna go out and put up fast food restaurants. You’re not talking about a high paid workforce, right?

Mm-hmm. But now you have to wanna scale and have a consistent experience across every single one of those locations. Mm-hmm. And. You know, like understanding that and thinking about that was, is really like, there’s so many layers that onion you can peel through, but then you think about it, I mean, it’s a little big brother has to say it, but if every interaction across everything that you’re talking about from terms of a customer, you can track, record, monitor and [57:00] improve the experience and the consistency of that ex interaction with that customer, you can scale your business in a very different way.

Right? Because now you can have a consistent experience and the customer can have, and you can find abnormalities, right? You wanna know when a customer has a bad experience ’cause you don’t want them on social media talking about their bad experience. Right? Right. You wanna be able to proactively be like, Hey, I’m sorry you had this bad experience, you know, let’s, we’re gonna, we’re gonna deal with it.

Right.

Jim K Tennant: Right, right, right. You’re, you’re, you’re talking about being able to get information that used to take much longer period of time. To digest it and then to react to that. And that goes into the, you know, comment we made earlier in the show. And that was, um, you know, what do I do with the data and then how quickly can I respond to that?

And how fast this technology is growing, um, to the point where, how do you slow it down? Mm-hmm. Or it’s going to, you know, I wanna partner with someone who knows where it’s gonna go two, three years from now, [58:00] and not someone who can just help me with my needs today. But, you know, we’re seeing clients say, you know, vendor of choice will be someone who can help me do my journey three, four years from now.

Um, so I don’t have to keep switching these point solutions and other vendors. Right. Uh, they wanna know, okay, hey, I’m doing a gamification for my people. Can you help me with that? You know, what insights, what workflows? Um, so we have to have data that can support what they wanna do and what changes they want.

Max Clark: It, it really is an interesting switch when you start talking about like, who then owns this technology and this stuff internally inside of an organization. Mm-hmm. And, you know, uh, contact centers, customer support organizations, pr, they’ve had to deal with this for a while now, which is, you know, maybe you had your, your frontline, you know, uh, agents or BPO, whatever it was, right.

Answering the phone, or maybe it was email electronic or you had a chat bot or something. You had chat that you were doing. But then social media comes out and becomes this like thing that you have to monitor and you have to [59:00] respond to, and your response is now as public as the initial complaint. Right.

So now, so then you end up with teams and people have a dis dedicated teams for social media. Well, who’s responsible for that social media team? Is that social media team a customer support, customer service function? Or is this now a marketing, pr branding function for the business? Right, because you wanna have positive public interactions with people.

Um. And, and, but all of those interactions, every people end up on social media after they’ve had some other experience. They were at your store, they were at your building. They were, you know, they were, they were at your, you know, bank. They called in, they some, they were online. Whatever it is, there’s some other interaction that then prompted that first, right?

Jim K Tennant: Yeah, yeah. I mean, you, you think about all of us have the experience, uh, on our phones and it’s kind of scary. You’re talking about, you know, uh, uh, a 1965 Ford Mustang, red cherry, blah, blah, blah, manual stick to your point, and [01:00:00] your phone all of a sudden starts popping up ads for used cards. Um, but you raised a whole nother question and, and commentary of chat websites, social media, um, everything that we’re doing, whether it be, you know, CCTV, everything we’re doing is being watched by social media and AI to the point where.

People are, Hey, you hit our website the other day. You didn’t finish that order. Or you’re getting a message from somebody, or you don’t wanna do the website because they know they’re gonna come after you on the phone. The circle of right, the infinity loop of wherever all this information is, you know, so many different solutions out there today.

Organizations are looking to try to, how do I combine all that? How do I take all that information? And then to your point, what do I do with it? Who handles that? You know, who manages that moving forward? So it’ll be a, a constant challenge with everybody. Right. And, and, and what’s more important than hitting the website [01:01:00] versus listening to the phone calls to how do we proactively tell our clients what’s, what’s happening?

How well do we know our clients?

Max Clark: Yeah, let’s talk about the, the, the, the expansion, the evolution here a little bit. So. I’ll give you a scenario. Let’s say, let’s say a customer starts with, uh, an with, with, um, the recordings. And they’re going through and they’re using this for a, you know, to monitor and get a hundred percent QA coverage across all their phone calls.

And after doing that for a while, they realize like, hey, there’s, you know, we’ve gone through and we’ve done training, we’ve, we’ve doing process improvement internally, and now we’re gonna expose this in terms of agent assist, right? Mm-hmm. Because now we have this opus of information. Mm-hmm. And we’re gonna, we’re gonna link in our knowledge base and one of our CRM tools and we’re gonna, we’re gonna present the agent on the call with a screen that gives them as they’re talking, right?

So now they’re talking and they’re getting, you know, the AI is giving them information in real time in front of them. They don’t have to go out and search for it. They can find it. What’s the step that takes you from that [01:02:00] into, um, into the, into the AI actually interacting? With a, a customer in real time.

Like, you know, is it we’ve identified things that we think we can deflect? Is it we’re trying to you, you know, like, what, what becomes that? Like, that like line, it’s, now we’re ready for it. Now we should do it. This is how we tackle it. Let’s start with this thing in our, in our IVA first. Like, what’s, what’s the process?

Jim K Tennant: So, so if I hear you, so you have my data, I have my post analytics, I have my qa, been there, done that. Yep. I’ve moved now into real time. I’m coaching my agents real time as they speak. At what point does that say, okay, now I’m ready. We talk about growing, right? The journey. Now I’m ready for that virtual agent up front.

Yep. When you see that real time agent handling those mundane things, those repeatable, um, um, tactical things that can be taken off their plate. Mm-hmm. You talk about, you know, resetting my passwords, um, that so common in the [01:03:00] financial world, right? Mm-hmm. How many times in real time is someone saying, great.

Yep. Let me, lemme reset that for you. You’re getting more accurate data in that real time widget, um, than you would if you were not using that at all. Um, you take those trends and figure out, okay, does my, does my IVA just need to have a, a bot off one of those nodes because I just spent enormous amount in replacing my IVR with an IVA.

But you know, that IVA is not doing everything that I need it to do, but based on the data that I’m getting real time from my agents, I need to either add a node on my IVA, I need to tweak my IVA with an address, a direction. Um, maybe we need to be proactive with the IVA, the, the, the gardening example. Um.

Using that data all combined, uh, we’re seeing these trends and working with a supplier that can [01:04:00] kind of help you see what those trends are and the impact, the business impact of those trends. Right. The call volumes, the repeatable information.

Max Clark: Yeah. So we should really just look at this from a standpoint of, you know, you’re gonna get into a scenario of continuous, you know, improvement and you’re just looking for this like flywheel to start spinning of, oh, we see a lot of trend here.

You know, can we automate this? Okay, automate it, come back. Okay, what’s the next big trend here? Can we automate it? And you just kind of keep spinning around in the circle a little bit.

Jim K Tennant: Exactly. Exactly. Yep. Uh, um, you know, IBA can be changed and the beauty of, you know, what these new gen NECs are doing, it could be done on a daily basis if you wanted to.

Right? I mean, you could literally go in and, Hey, this week we need to do this, but based on LA last week’s results, or, you know, Monday, Tuesday of this week, we need to change this.

Max Clark: You know, you call into a, a company and you get their IVR or you, or you get an IVA, right? You get their, you get their recording, right?

And [01:05:00] it’s like, your call is monitored for quality assurance purposes, which, you know, to your point earlier, like what percentage of them actually are. But then the second thing that you hear immediately is like, depending on the type of business, it might be like, press one for store hours, or you know, you’re, you know, whatever, right?

Maybe it’s regional or dial in your, you know, but then you’re gonna get like, press two for sales, press three for service, press four for this, press five for that. You know, and you get these, you get these, you know, you know nodes, you get these trees that are gonna take you down different paths. And as part of it was for routing, right?

Like if you’re going to a sales person or to a service, a support person, when do the IVAs get smart enough that that goes away completely. And you call in and it says to you, you know, thank you for calling, thank you for calling it broker.com. How can I help you? Right? And you get this, you know, just, just what sounds like a person.

Answering your call and you could say, Hey, I need my, you know, I’m having a problem logging in. My password isn’t working, and the system just goes, oh, okay, great. I can help you with that right now. Or you say, I need help with this other thing. And it says, oh, hold on a second. I’ll transfer your [01:06:00] call. You know, and, and, and that, that gating of that experience changes to be more human.

Yeah.

Jim K Tennant: That’s, um, that’s actually available today. Um, what we’re seeing as you call in and what you wanna do is your multi-factor authentication. We validate that it’s Max calling in. Let me ask you a couple questions, blah, blah, blah. You know, get that through, right? Two years ago it was, let me validate you Max, and then get you to the right person.

Then it’s transferring you, right? Mm-hmm. Mm-hmm. But now the IVA, it can, it can MFA you multi-factor authentication. You, however, design the client wants. Matt, are you calling about that order you did the other day? And you’re like, no, no, no. I’m not talking about that order at all. You know, I, I, I saw something else on the website that I want.

Mm-hmm.

But it knows today, and you can. Tell it to proactively bring up different topics. Um, you know, did you want sales service? I see you have this history with us. Uh, it gets into that. We talked earlier about how long do you want that conversation to be? [01:07:00] You know, the, the bot can, can rattle off a bunch of stuff that the client’s like, Ooh, that’s pretty scary.

How, how much do you know about me? Uh, versus how long do you want that conversation to be? Or, you know, what do you, what are you, what are you calling about today? And quickly, you know, get in and get out. Um, yeah. That’s available today.

Max Clark: So what you’re telling me is in addition to my rage for companies that don’t have virtual hold, I should have an equal amount of rage when I have to call the airline.

And then it prompts me through this impossible, you know, tree first. Yes sir. Yes sir.

Jim K Tennant: Yeah, yeah. Now convince that, you know, same, same organization to imagine for a moment if you did call in and say you call about this flight or that flight. They’re the one that we delete. Yeah. Yeah. They don’t want to go down that path.

Yeah. But it’s available. Technology can do that today. Yeah, it’s, uh, I would love my airlines to be able to do that.

Max Clark: You know, like it’s, change is scary, you [01:08:00] know? I get it. Change is scary. Like, and especially when you start talking about change at scale is scary. Yeah.

Jim K Tennant: Um, well, it’s a great point. I mean, you know, we haven’t said any brand names at all. We’ve been very careful about this, but I, I really like calling into my 800. I’ve got their number memorized.

I like it because they’ve got a lot of agents behind that phone. And, uh, I never talk to a robot. I never talk to a robot. And, uh, I can be authenticated by a real person. We can get in and get out. Now, could they adopt change? I’m sure they are looking at, they have to be able to, um. But the cost of having humans available for all of that right, is a decision that businesses need to make.

Um, because inevitably we will always have those clients that say, we will never do that. That that is not our, you know, we have to have a, a person behind that. We’ll [01:09:00] continue to hire four or five BPOs, but buts in the seats that, that’s, that’s okay. That’s, we’re gonna see that those are the same people potentially that, you know, will never go into the cloud that have their phone system and their PBXs and their hardware and everything in that back office that they’re paying AC for to keep it cool.

Yeah. Um, we’ll see that. But, you know, we will also see the conversa that, um, hyperscalers, you know, going, going full board and, uh,

Max Clark: adopting

Jim K Tennant: that

Max Clark: change. We’ve, so far, our conversation has been very externally focused, right? Company that has a call center that’s supporting external customers. But what we haven’t talked about at all is supporting your internal employee base, right?

Mm-hmm. And I don’t care how big you are, you have, you know, a customer service and support function for your internal teams, and then as you scale and grow and get larger and larger, that becomes a bigger and bigger issue for you. So how, how, how should companies be thinking about this and approaching [01:10:00] this also just on their internal side of things, very internal.

Jim K Tennant: That, that’s great. Great comment. We have a, we have a retail client that, uh, a very large retail client, uh, and, and any enterprise client, X amount of employees, right? Uh, seasonality of, uh, renewing your health plan, uh, at a, as a compliancy piece, right? Um, whether I make changes or not. I have to go in there and say, yeah, no, everything’s fine.

Wife’s the same, kids are the same, yada. Yep. Click, click, click, click, click. Or maybe I wanna make some changes. But either way, let’s say I didn’t wanna make a change at all. We’re seeing companies explore ways that AI can help them internally with internal employees, because my HR staff is only so many people.

Mm-hmm. And, um, I need to do this compliancy, healthcare revision X times a year. I’m very seasonal. Right. So you’re talking about I’m not gonna pay a license for something that I need for 60 days. [01:11:00] Right. Um, so companies that size are looking to see what I can automate through AI that, uh, without having to buy new systems.

Right. But, but so I can be compliant. Um, and, you know, the cost of doing that versus lemme go hire a bunch of people for 60 days and then, uh, you. Riff ’em, lay ’em off just because I need something else handled. Um, so that’s happening today too.

Max Clark: Are the use cases for that inside of an organization? I mean, across every department, right?

Like this also becomes an interesting thing ’cause it’s no longer like a siloed thing or one team owns this platform, right. It is gonna have stuff they could have help with. HR has stuff that they need help with, used to talk about, you know, insurance renewals have stuff that could help, you know, like this is a long list of stuff.

Oh yeah,

Jim K Tennant: yeah. Imagine just, you know, having, having in our own company, you know, talking to a bot and tell me about what promotions do we have, what events do we have coming up? Uh, where can I find this [01:12:00] information? Right. Call, calling, calling a robot internally and having the bot say, oh yeah, check this website, or check this site or this application, whether you’re using back office systems or what.

Um, so using, using a self service tool with the your own company to help navigate orchestrate, Hey, I’m a new hire. Where do I go get information about X, Y, Z sales process or what have you. Imagine calling an IBA with your own company to help you navigate through the waters of that. I, we’ve

Max Clark: been, we’ve been using like, language of call and it’s been very like voice centric.

But the great thing about these tools as well, and, and what you’re doing is, it’s not just voice, right? I mean, modalities of being, it’s, you know, text as well as voice, which then means it can basically be any channel that can write on top of text, right? Correct.

Jim K Tennant: Correct. Yep. The idea with AI isn’t necessarily to, to be just the voice in the real time.

It’s, you know, I don’t want to talk to someone and I want my preference to be text multimodal, right? Yeah. Um, so all that’s there, and [01:13:00] then you have to have those systems talk to the other system so that if I do text and SMS, but all of a sudden now I hit the website or I do call in. How about that now?

Now the bot on the voice site says, Hey, I saw you hit the website a couple times. Hey, we’ve been texting back and forth the last few days. Really, you know, that, you know, I’m coming in at entry point somewhere else. Clients want that seamless network so that, in regards of where that client is looking, seeing coming in, um, to have that, that 360 visual, you know, experience.

Max Clark: So there’s a, there’s an interesting, um, there’s an interesting dichotomy with this, which is, you know, one hand you’re gonna have the tinfoil hat, right? Where it’s like, oh, how did you know that I was like, doing this already? And like, why do you have this information? But then you have the other side of it, which is just the reality of, oh, I don’t have to repeat myself for the 15th time to try to get a pro, you know, whatever accomplished I’m trying to accomplish.

And, you know, in, in my world, like. Like, I know I’m being [01:14:00] monitored 24 7 for, to like, serve me advertising. Like, it’s just a fact of, of, of the world. Right. You know, like at some point we’re gonna be, every store is gonna have our face recognized and we’re gonna walk in, we’re gonna personalize advertising like minority report.

Right. But, but people that are resistant to that or have this like, like idea of like, oh, I don’t wanna surface too much information because I’m gonna upset somebody. This is still customizable, right? Like, this isn’t an all or nothing. You could be like, okay, we won’t tell the bot to surface all the information and still play dumb.

That’s an option for them, right?

Jim K Tennant: Correct. Absolutely. You talk about guardrails earlier. Mm-hmm. Guardrails could easily be Yeah, keep, keep the conversation specific to X, y, Z. Yeah. Yeah. You, you know, those things are available to it, but, uh, you know, code parameters can be written as such that, you know, keep, keep the conversation in line in your swim lane.

Talk about this with the end result being this. Absolutely.

Max Clark: Yeah. So let’s, let’s date this now, right? So as you [01:15:00] look, I mean, we talked, you talked about like roadmaps. AI is accelerating so fast, right? And just the pace of this is just outta control. As you look down the, the next six to 12 months, and we start talking about like, you know, near enough horizon, we have a relatively good expectation of it coming to fruition, right?

Not like what, not what happens in three years, but like, what happens over the next 12, you know, 12 months. What, um, what are you really excited about? Like, what do you see coming here changing? And I mean, that could be a technology question or that could be like an mm-hmm. Deployment operations use question, right?

Jim K Tennant: Mm-hmm. Mm-hmm. Um, well, we’ve had three product releases since we began the show today. Um, you know, you, you look at this and we’ve talked about, uh, the acceptance level of ai. Uh, we’ve talked about the, the resistance or hesitation to. The reality is, it, it’s, it’s, it’s lightning speed. It will continue to grow, it’ll continue to advance.

And by human [01:16:00] nature, we will constantly see new technologies develop advance, and companies will adopt it, whatever pace they wanna adopt. Um, with that in mind, I’m, I’m super excited about what’s, what’s happening in future. I, I foresee the ability to say, we’ll be able to do this podcast like we are today, max, with virtual agents that think, sound and write like we do.

So those, those times where you’re double booked those times where you really want to be two spots at one time, you’ll have the comfort knowing that your virtual agent will act lo, look, talk, and think like you, and give you notes and summarizations of things that you couldn’t do. Absolutely wild. Take that to a whole nother level.

What we’re talking about are LLMs talking to each other and IVAs talking to each other. That you’ve approved mm-hmm. Right. That you’ve written guardrails for, um, and that you’ve been able to summarize for. And if we’re on this call in a real [01:17:00] time way, your virtual agent calls you and says, you know what?

You really need to come to this podcast. There’s stuff going on here. You need to be, that’s, that’s the exciting part. It’s like, wow, you know what? I need to go because there’s another call coming in from my virtual agent that I really need. That’s, I mean, now. Are people willing to and wanna accept that?

We’ll see, we’ll see what the adoption rate is for that. You know, how many meetings can we be double, triple booked for?

Max Clark: Yeah. When you’re getting into conversations or your team is out, having conversations with new customers, with, you know, people looking, what is something that you wish more people knew about, observe AI and ai, you know, in the space in general, coming into the conversation that you think would make things better for them?

Jim K Tennant: I, I think, um, you know, we talked about point solutions. You brought up platforms. Um, I think, uh, clients get overwhelmed, uh, rightfully so, based on how much is out there. Um, I use that, uh, crawl, walk, [01:18:00] run methodology because the idea of being, let me, let me provide information to you in a very digestible way.

We wanna earn your credit. We wanna be able to validate our solution with you, um, based on our platform and the number of. Enrichments that we might be able to offer someone. Um, my hope is that we’ll continue to better understand what the client’s needs are. Um, either fix their, bleeding, their hemorrhaging, or have a better, more healthier operational efficiency discussion about what AI could do for their business and really understand and build that roadmap with them so that we’re not here to say, buy every enrichment, buy every SKU that we have.

We are here to say, why don’t we handle what you’re looking for today at the best of our ability to allow them the ability to grow the way they wanna grow and as fast as they wanna grow. And not, not try to say, you have to buy this license and utilize [01:19:00] 50% of the license. No, but buy a license and enrichment that works for you.

Choose the features that you need. Have the ability we’re gonna be having, I keep saying these enrichments, we’re gonna have the ability for clients to choose different enrichments within their scope of license and be able to change those enrichments based on what their needs are, based on their changing environment, as opposed to having a QBR and saying, you know, we’re paying x, y, Z for this license, but I’ll be damned if I’m using, you know, but 60% of it, there’s no value here that, that needs to go away, right?

Mm-hmm. So we’re recognizing the, the, the full value of what we can do. And as we continue to grow, um, you know, our, our sales team and operations teams work closely with clients to get that message out there.

Max Clark: Awesome. Jim, thank you very much. I, I always love the chat.

Jim K Tennant: Yeah. Great conversations. Good stuff.

Good stuff.