Join Tom in this insightful podcast episode recorded at the HDI Philly event held at Princeton University.
Delve into groundbreaking strategies and cutting-edge technologies, including AI, to revolutionize your service desk operations. Drawing from his leadership at Expivia, where he expanded the team to over 600 professionals, and his innovative work at OttoQA, Tom shares advanced methods that enhance operational efficiency and elevate customer satisfaction.
Explore the latest trends, tools, and best practices reshaping service desks across industries.
Whether you aim to optimize your current operations or completely transform your service approach, this episode offers expert advice and actionable insights to propel you to the forefront of industry excellence.
If you are looking for USA outsourced customer service or sales support, we here at Expivia would really like to help you support your customers.
Please check us out at expiviausa.com, or email us at info@expivia.net!
Follow Tom: @tlaird_expivia
Join our Facebook Call Center Community: www.facebook.com/callcentergeek
Connect on LinkedIn: https://www.linkedin.com/in/tlairdexpivia/
Follow on TikTok: https://www.tiktok.com/@callcenter_geek
Linkedin Group: https://www.linkedin.com/groups/9041993/
Watch us: Advice from a Call Center Geek Youtube Channel
Join Tom in this insightful podcast episode recorded at the HDI Philly event held at Princeton University.
Delve into groundbreaking strategies and cutting-edge technologies, including AI, to revolutionize your service desk operations. Drawing from his leadership at Expivia, where he expanded the team to over 600 professionals, and his innovative work at OttoQA, Tom shares advanced methods that enhance operational efficiency and elevate customer satisfaction.
Explore the latest trends, tools, and best practices reshaping service desks across industries.
Whether you aim to optimize your current operations or completely transform your service approach, this episode offers expert advice and actionable insights to propel you to the forefront of industry excellence.
If you are looking for USA outsourced customer service or sales support, we here at Expivia would really like to help you support your customers.
Please check us out at expiviausa.com, or email us at info@expivia.net!
Follow Tom: @tlaird_expivia
Join our Facebook Call Center Community: www.facebook.com/callcentergeek
Connect on LinkedIn: https://www.linkedin.com/in/tlairdexpivia/
Follow on TikTok: https://www.tiktok.com/@callcenter_geek
Linkedin Group: https://www.linkedin.com/groups/9041993/
Watch us: Advice from a Call Center Geek Youtube Channel
This is advice from a call center geek a weekly podcast with a focus on all things call center. We'll cover it all, from call center operations, hiring, culture, technology and education. We're here to give you actionable items to improve the quality of yours and your customer's experience. This is an evolving industry with creative minds and ambitious people like this guy. Not only is his passion call center operations, but he's our host. He's the CEO of Expedia.
Speaker 3:Interaction Marketing Group and the call center geek himself, tom Leard. How's everybody doing? I know we got Phil here. Any Archbishop Wood graduates, parents, any? All right, I'm in a safe spot, then I can continue. So I coach high school girls basketball and Archbishop Wood just beat us in the state championship in Hershey last weekend for the third time. So I am, for you football fans, I'm the Marv Levy of high school girls basketball. That's my fourth loss in a state championship, but I will take it To lead into what I want to talk to you guys about today.
Speaker 3:This is going to be a little different. I think it's going to be a little unique. I hope that it resonates with some of you guys, because some of the topics could be a little bit I don't know just a little different from probably what you've heard here. Speaking with the basketball kind of theme, there's an awesome basketball coach and he's retired now named Hubie Brown, if you're into hoops, he was on TNT as a commentator and he was one of the best clinicians clinicians dude, who gives clinics right and I went to go see him about 10 years ago and he was supposed to be talking about how to attack, switching man-to-man defenses right, which is like a topic like I'm like, well, that's interesting for me. Well, he got up there and was like guys, I'm not talking about this, he goes. I just want to give you things that I think are really cool, that I think can help you with your basketball team. So he randomly talked about how to his favorite play against a two, three zone inounds plays, um, you know what what he does against a box in one, and that kind of resonated with me from all the things that I'm still continuing to use to that day.
Speaker 3:So, about two years later, I speak at the nice where I'm a big nice CX, one customer Um, and I speak at their user conference and I'm like I'm going to use this tactic, I'm going to do kind of the talk that I'm giving today. It's almost the same thing. Obviously it's different material, cause that'd be bad if I was giving you seven year old material, um, and so I'm all excited, right. I get my time slot, though, and it is the last day. On Wednesday it's the last slot, like one 20 in the afternoon after the customer party, so everybody's hung over from the day before Um, and then I go into the room and I'll never forget it was canary three. You know where the bird goes to die in the cave. I'm like they have no thoughts for this, um, but it resonated. The room was packed. It was really cool, and I'm not saying people still talk about it to this day, but it's something that you know.
Speaker 3:They've asked me to come back in every couple of years. I kind of re-up that. So when Jason and I were talking, I wanted to kind of bring that flavor to something that I think could resonate with you guys. So I'm going to talk about everything from IT stuff to analytics, to AI, to engaging employees. This is like if you hate my one slide, wait three minutes, because something else is coming that hopefully will resonate with you as well.
Speaker 3:Having said that, let's get this going All right. So that's me. If you want to connect with me, just take a quick picture. I post on LinkedIn every single day multiple times. If you guys have, if you get anything from this, please, please, please check out the podcast. I heard everybody has like an hour or 45 minute drive home, right, so it sounds like the perfect time to kind of hop that in. We have over 200 episodes of everything that you could imagine on that podcast and there's no sales on it. It's just how to help contact center people, service professionals, and I think I think that's how Jason and I, at least initially, kind of connected as well. Actually, I bought your book, bought my book. That's even better here.
Speaker 2:That's how I found you, and then everything else kind of came along your active LinkedIn presence and, you know, your thought leadership when it comes to, you know, to to your podcast, um uh I, it is a regular play with me.
Speaker 3:Thank you, Appreciate that. And then what I do to actually pay for the bills is we have a 600 seat BPO outsourcer in Erie, PA. Um, so flew down from Erie. So everything from tech support, financial services, retail, e-commerce. We have Xpevia Digital, which I'll talk about later, but I really want to focus not focus, but talk a little bit about AutoQA as well. It's a seven month old AI startup that we started focusing solely on fully automating quality assurance for contact center service, desk, ticketing, that kind of stuff. So, just again, I'm not here to sell that, but I want to show you, tell you, some of the cool things that we kind of learned. And again, I have there's really no agenda for this, Well, except for my agenda page here. So, again, just to give you some cool ideas to use in your, in your contact center, and I'm going to just dive into this right now. So, again, this goes kind of deep right from the door.
Speaker 3:So one of the things that we have found as a BPO is I need to kind of use every single tool that's out there and kind of like squeeze it to get every single bit of ROI and every single different unique use case for it to get every single bit of ROI and every single different unique use case for it. I know a lot of you guys struggle with SLAs, right? Depending on if you're just taking voice, if you're just taking tickets, if you're taking emails, right, how do we kind of manage that? And on the voice side, we have seen customers that come to us and say, hey, Tom, here's all my data, here's all my calls, here's my handle time. How many agents do I need? So we throw all that stuff in there and, hey, guys, to get an 80-30 SLA which everybody's still using, right, At least on the customer service side, you're going to need, let's say, 52 agents and I get all the time. I don't have the budget for that. And what can we do then? To say and be smart with how we reduce staffing. And I understand we can go chatbots and we can do some self-service stuff, but just from the voice side or actually answering interactions.
Speaker 3:And one of the things we found is to use analytics to start testing where customers get ticked off right From a time on hold or in queue. So, again, if you have like your higher ups that say to you hey, we need to lower our budget, you need to get rid of headcount, which is happening in a lot of different places. This is one of the ways that you can be. I guess you can really think it through from a logical standpoint. So we'll start at, you know, say a minute, and then we'll go to a minute 30, and then we'll go to a minute 40 and we'll just reduce headcount to where we're waiting and waiting a little bit longer throughout the day, and we will find and there's an absolute now we're using CX1 interaction analytics. But if you're using Observe AI, any type of AI or, I'm sorry, analytics tool, you can do this with it. So we have found you know what your customers do not get ticked off until two minutes and 24 seconds. So we can now, instead of saying 80-, 30, you know, maybe we can do 80. And my math is not going to be right, but you know three, 82, you know whatever that is. And we have found it directly correlates to CSAT, to NPS and obviously to sentiment as well. So again, just a unique way to use analytics, um to to help you kind of justify headcount or at least lower headcount in a smarter way.
Speaker 3:Yes, Love this. Yes, At the beginning, normally at the beginning. So you know, again, you can't just do it on one. So we'll take we'll take a hundred to 200 interactions and you will definitely see a a one of these, um, you know, at when we get to a certain point and we I wish well we'd done it a little bit with with some of our customers, but I've not done it like industry-wide. I'd love to benchmark, you know, uh, IT support versus retail, versus, because obviously you know, if you look at, you know when you call like Expedia, right, or you call the airline, you're, you're willing to wait longer. It's almost like you've been programmed for that right, when some like healthcare, you're not like you want to get to that person right away, so those everything has its kind of its nuance to it. But yeah, it's a. It's a pretty cool way that I think we've been able to kind of justify headcount to some of our customers in a very easy way to do it.
Speaker 2:Do you know MetricNet? Have you run into MetricNet yet?
Speaker 3:No.
Speaker 2:So they do that kind of benchmarking in the call center industry.
Speaker 3:Stole it from me, stole it from me. Kidding, kidding, they presented your one, yeah. So again, if there's any kind of tool, that's a really cool use case. Here's another one. And again I keep saying, like these are my favorites, but these are like my, my, my hits, my greatest hits, I guess.
Speaker 3:So have you defined, and this is I don't want to get hokey here because everybody, again, you guys are more IT guys that maybe you don't appreciate this as much, but you're running a 600 employee business. Right, culture is really important. So we define our culture down to how do we incent, how do we hire, how do we promote? So we say, and this kind of sounds mission statement-y, but a sunshine attitude with an entrepreneurial mindset which, if you think about it, is attitude and effort. So we've always been able to everybody can kind of incent effort. Right, you can sell more, you can do more tickets, you can, you can do more things. Right, you can figure kind of that out. Right, in my world it's it's a lot of cross sells and upsells. Right, did you retain a customer? But how do you incent attitude? How do you make sure that your customers or or or your agents are treating whoever they help, whether it's an internal help desk, whether it's external how do you make sure that they're treating them the proper way? And so, with analytics, again from a sentiment score, we incent all of our agents, especially on the customer service side, off of sentiment with analytics, so they're paid more. We post this every single day. So, again, me being an outsourcer, I can prove to my customers that you know, my agents are using the proper word choice, that they're saying the right things and we're actually this is one of my agents main thing that gets posted. Normally it was posted on a wall and now it's posted on Slack, but again, an unbelievable way to increase the way that your agents are talking. And also a great culture influencer as well, because you can really change kind of the dynamic of how they're speaking and then they're getting paid to actually do really cool things and treat people the right way.
Speaker 3:So, again, just another little thing. And then I have one more analytic thing and I'm moving on from that yeah, yeah, um. So, and that's kind of where, with the track and incentivize positive attitude. So that's kind of where we always struggled, right, we said we can, we can incent effort, but how do we incent attitude right. And so if you have every single day that you have your one agent or all your agents that are in the blue here and obviously I took a really good pic snapshot of a blue one, really good pic snapshot of a blue one You're actually, you can incent attitude like with your agents, right.
Speaker 3:Because for us, attitude means how do we treat the customer? Do we have more negative, do we have more positive sentiment? So from that standpoint, we just we feel like you know, if we can just show every single day that we're treating customers the right way, that we're using the right word choice, that we're maintaining our composure on calls, that's a way that we can incent attitude, positive attitude, in our contact center, and it has been again. It's a huge culture change and it's also a huge change on how your customers, for us, get treated. And it's a way for me to prove to my customers from an outsourcer that I am handling their customers the right way, a way for me to prove to my customers from an outsourcer that I am handling their customers the right way.
Speaker 3:All right, last thing, and I'm going to get, I got to. I have to talk a little bit about AI at the end. That's going to relate, I think, really well to you guys. But I'll blow through this one quick and I apologize for the wordiness of this, but it's a prompt. So one of the things, again from an analytics standpoint, that we also find is whether you're using level AI, interaction analytics, a CCAST internal analytics platform.
Speaker 3:You have to set up things right. It's cool to know the things we don't know, but we also want to set up things for what we want to track From a retail side. We want to know, you know, what percentage of customers used the phrase too expensive yesterday? Right? What product keeps breaking that customers keep calling in about? What is the number one or number two reasons why you know they need our help desk, our support, right? So building out for a customer of ours, building out the categories, subcategories, trending keywords, right to look for, was a pain in the rear, right, and we just have to listen to calls and listen to calls and talk to the customer. What do they need? And try to build the stuff.
Speaker 3:Well, we found that prompting works extremely well. Just in the regular desktop chat, gpt or cloud three or whatever it is right To be able to say, hey, what are the three things I want to do. Most of the time it's, you know, from a marketing in my world. It's from a marketing standpoint. It's what needs to be fixed because people keep calling in about the same stuff. And is there any type of price adjustments or any type of product adjustments that we can make so we can build out prompts, you know, from a retail customer to a financial services customer, to an IT help desk that can help you build out those categories, subcategories and keywords.
Speaker 3:One of the, I guess, really cool examples of this is we work for a hospital and the Trump family gives a ton of money to it and say what you want about the Trump family, that's nor here nor there. But we have to track right through analytics because customers will, or their hospital people will call it all the time hey, give the money back. Or Trump's they horrible, right, and they track that Right. And that's some of the things that we're. We can look and kind of track and kind of build categories and subcategories for, um, you know for, for specific things to look for. Okay, this is where I think, if this is low hanging fruit stuff, that fruit stuff that very few do, that is not very difficult to program to create.
Speaker 3:I really don't see us again in the BPO world and the contact center world. If you're just kind of queuing calls the old-fashioned way or queuing interactions the old-fashioned way, I think you're missing huge opportunities for customer experience, right? So data-driven call routing is basically just fully integrating your data sets within some type of telephony platform that you have, whether you're on-prem, whether in the cloud, right, and then utilizing the data to actually route calls. Right. So for there's multiple examples with this, but here's a couple examples.
Speaker 3:Number one we work for a company that has about a thousand to 1200 financial advisors under their umbrella. So when a customer calls in, we see that they're using Salesforce. Customer calls in, they see the Annie. The Annie then goes to what Salesforce record that is, finds out who the financial advisor is, and again, this is a pretty generic way of doing this, right? And then we let's say it's Susie James, financial Planner Agency. So then we answer the call Susie James, financial Planner. How can I help you? That's pretty easy. We've been doing that for a while.
Speaker 3:But what's happening now, too, is especially, I think, service desk wise is what happens if you're taking a voice call with someone external and they either get hung up on or they hang up or there's a mistake, right, and you have a 35 minute wait time right, there's no reason that that customer should have to wait anymore. Right, when we can call in, see that they just were there in the data, right, and then move them to the front of the line. If you have high value customers, if you have customers that have products that you need to service, if you have, maybe, the dean of the school, right when he calls in to move to the front of the line, right, utilizing data is extremely easy to do and something that I think is not done enough and it really doesn't make any sense to me. But with us, we can route calls, or you can route calls any way possible, as long as you have that type of data, um within your sets. This is another huge reason. This, this is is really important to integrate with, not just from this. But when you talk talking about AI, right, you're going to have to do it anyway, right, all of your data sets are going to have to be started to be integrated to your core, and I know a lot of you guys do that already, but I don't think you're utilizing the data to the maximum capacity that you can for very little little amounts of money.
Speaker 3:Questions on that, because I think that's crazy important and crazy cool. Yes, of course I'm giving, I'm getting right, but let's, let's just say you didn't do that Right. You have the phone number, I'm sure you have, but you kind of you can kind of see like or or maybe they're like, hey, let me go try that, and they, they hang up and they've got to go try something, right, to go fix something, and then they have to call back in again. Not everybody's, you know kind of how, how, how it's set up is proper, but I mean, your point's taken right. You probably would just call them back. Um, thanks, appreciate that. No, I got you All right.
Speaker 3:I want to talk about something totally off the board now, outside of technology, right, and some of the ways that we're engaging with our agents. Um, and some of the, I think, unique ways that and again, you can tell that I put this slide together so there's like no animations, it's just here. But number one is back. When we were on the office, we had 600 full agents. We had about 350 people per shift. It was awesome. I've been talking about it to a lot of you guys here. We have like eight people in the office now, right, so no one's there.
Speaker 3:So we've had to figure out some, some unique ways. One of the things when we had people in, we had a small like 12 by 12 office that you know, I had a glass dry erase board and any employee could walk in there and they could write down a suggestion. But it had to do with our culture. Right, how do you increase attitude, how to increase effort? Right, like, what are some of the things that you could do to do that? Now we use mondaycom now, where you can go in and kind of be anonymous and then once a month everybody votes on it, and whether it is something that we need to change with it from maybe a screen to maybe somebody wants to go to an Erie Otters game. You guys have the Trent Thunder, we have the Erie Seawolves right, go to a Seawolves game, those kind of things, which is, you know, kind of low hanging fruit. But every month we'll get a committee together and we'll go implement that.
Speaker 3:And I think the agents have found a lot of I don't know kind of goodwill that's come that we're actually listening to, kind of what they say the middle one, I think is really important. It started with the company that I came from. It was a larger BPO in the mid 2000s. We don't really have team leads, we have management trainees. So we rotate our management trainees in a six week kind of training program to where they're with everybody else's department. So they'll spend a week in HR. They'll go through how it looks like to interview somebody. They'll go through how it looks like to maybe let somebody go. They'll go with a client services talk with one of our client services managers maybe if they have an irate customer of ours, somebody that's not happy with something, and they get to see kind of all the things that they do on that level of the contact center or the service desk and how it affects all the other departments. And there's a lot of, again, feelings that kind of then come into play when they actually take their team over and they understand now why we're asking them to do specific things.
Speaker 3:You know, I think management and middle management training within the contact center is, if you do that well, I mean you're living the life really good, right? They're the ones that are actually managing the contact center managers. You guys really aren't to a certain extent, but they're the ones that have that every single day touch point with the agent. We do middle management summits. We used to do it quarterly with everybody at home. We're right now twice a year.
Speaker 3:The middle management guys, as I just said, they know everything. They know how to manipulate the platform. They know how to manipulate the platform. They know how to manipulate the system. They know what to say to customers, even if it's not in the script. So we kind of take those guys and we take them for a half day and we just have them present on some of the things that they understand. The middle management team knows, just knows, has so much information and especially in the BPO world, you know you have a financial services and retail and e-commerce and tech support and all these guys come together and kind of talk through some of those things that all of them are kind of dealing with. It's been a really cool again employee engagement thing and learning thing for all of us.
Speaker 3:And then the last thing is we do have our contact center leadership groups, so any type. Now this was a little bit different when we were all in house, but have you guys ever tried to order like 400 chairs for a contact center and not ask the agents if they like the chairs, right. If you haven't, it's a disaster. Um, so we learned the hard way a long time ago. With headsets, with any type of equipment. There's a full committee that kind of goes through that.
Speaker 3:Any type of onboarding of new agents. We have a welcome committee, right, that is kind of. We have a Slack channel they mentor those agents from. If there's any like this supervisor stinks, I can't stand her. Like they'll have those conversations and like, all right, well, listen, she's not that bad, right, at a different level than I think they can have with a management person. So that's again something that we have found to really be cool to engage with our employees. All right, this might be a little bit out there and, again, like I said, I'm going all over the board here. Let me ask you this how many of you guys have a CCaaS or are in the cloud? Right, one? Okay, great, this is going to be an awesome slide then.
Speaker 2:So I started with we probably all want to. No, totally, totally get it, but you know if you're working with 10, 15, 20, you know if you're just using some kind of legacy.
Speaker 3:A hundred percent, no, a hundred percent. So I started this or I learned about this back in the day on the Avaya DFINITY system, right. So I used Avaya and they had a thing called the business advocate, and the business advocate would basically keep reserve agents in tow. So when a call came into the IVR and the predicted wait time for a specific skill or cue was X and you needed it at Y, it would automatically move those reserve agents back and forth. Well, on the CCAS side, right, it's the most underutilized tool and, again, I think even on-prem guys have this tool. I know Avaya had it 15, 20 years ago and basically it allows that same thing. So, again, you can guarantee this is another really cool use case for, I guess, more being in the cloud, which is not resonating because only one of you are to be able to move agents in and out, to guarantee service levels, to guarantee specific handle times by agents, like we use this. For if you have an agent that is brand new and you have a six minute handle time right To work a ticket and let's say that they're at like seven minutes and 48 seconds, well, this can automatically lower their preferences and lower their priority, so they're not taking as many calls, so you're not getting absolutely hammered on your service level, right. So a lot of, I think, some cool things with that. It's just on everybody's platform. Nobody just uses it. I always kind of just call it out.
Speaker 3:All right, I want to talk about AI, and I want to talk about AI in a realm of what is real, what I've seen from the consulting side, from what I see of all the literally hundreds of demos that I've taken, what is not real and kind of where this thing is going and what you need to do to start. And what you need to do to start and I would say for most of you, it's to probably start to look at the cloud, which I know is something that probably is a little bit of a tough subject, but not to say you can't do this on-prem. But here's the stuff that you need to start thinking about this year. You don't need to be moving to AI, right, but you need to start to think about how do I integrate all of our platforms, all of our data sets, as many things as we possibly can, anything that you want AI to start to look at or start to think through in the future, right. Start to integrate that with whatever telephony platform you have.
Speaker 3:The second thing is your KMS needs to be on point, right. I think service desk is the low hanging fruit of and I keep using that word, but I think it makes a lot of sense of a channel that can move to AI, because you have so much material, you have so much data, you have so many things that an agent needs to know that a lot of times they can't know, so most of the time, the KMSs that I find in the service desk world is much better than, just, say, a financial services company, so that's extremely important. Think about your new policies and procedures. How does security work in this new world? What LLMs start to experiment with that have APIs that can connect, that you feel comfortable from a security standpoint. All of those things kind of start to be coming to play. And how do you QA the AI? Right? It's talking about like an LLM chatbot, right? How do you know it's saying the proper things at the proper time? You have to Q that. Qa that just like you would an agent, right? And there's things to think about through that. So if you don't buy one thing when it comes to AI this year, right, the AI this year right, start to think about this stuff. Start to think about how you implement this stuff and you're going to be, in a way, better place in 2025, 2026, when things really start to change.
Speaker 3:But this is the stuff that worries me for a lot of companies. I just I spoke in Miami about two months ago. There was a hundred people there. They're all CX1, nice CX1 customers, and I said who and it was just kind of on this, I stole this from that presentation said guys, who is even getting started with AI? Literally not one of them raised their hand. Right, so if you're not there yet, it's not crazy. You're not in a bad place, right, but start to think about how you start to get ready to go down that road, all right. And then again, all right, and then again, god, I use low hanging fruit a lot.
Speaker 3:So what are the big wins at little costs? Let's say your KMS is good. Let's say you're going to start to make some of these purchases, like, literally, what are the tools that you should look at? Right, not talking theory, right, but again, those legacy tools. If you don't have analytics, if you don't have WFM, if you don't have that kind of WFI that we kind of talked about, right? Those are the things I would look at first. Right, don't put the cart before the horse. A lot of these are going to feed your AI as well, right? So if you don't have analytics, I think that's something that needs to be looked at or thought through and start to think of how do we get that onboarded onto us.
Speaker 3:And then I think the next two things right, the I call them kind of the GPT tools, right? Auto summarization, auto QA. Looking at agent assist have anybody here implemented any type of agent assist? Okay, right, so this is kind of some of the stuff that a lot of the, even on the cloud side right, it's extremely easy to get every single RFP that I have found in the last 18 months to come into my BPO Now, granted, it's not all service desk stuff, but every single RFP I've had. Every single company has said that agent assist is mandatory, right, so it's not like something that's cool to have, it's something that I have to have, right. So, and again, these are some of the very inexpensive tools that are not crazy.
Speaker 3:And then we talk about chatbots, and you know, I don't I think you guys are all probably chatbotted out and again, I can have an LLM chatbot conversation all day long, but I mean, I wanted to just talk about again where everybody's at now what I think are the things that you should start to look at, compared to, like, going right to that first touch point, you know kind of LLM chat bot, where it can get a little crazy. I want to just throw this out If a lot of customers that I have talked to are very confused and I won't spend a lot of time on this slide, but it's just something that annoys me, right? So you have everything from Avaya to Genesis to Five9 to NiceCX1, to all of these platforms talking, coming out with some type of AI scoring, right From a quality scoring standpoint. Just remember that analyzing a hundred percent of calls when you hear that is not the same as scoring a hundred percent of calls right A, that is not the same as scoring 100% of calls. A lot of it is proprietary scoring. It's like, yeah, they did okay on that, I give it a 5 out of 10. But if you have any type of actual QA scoring where you're actually saying did they greet here, did they ask probing questions for the issue at hand, those types of things and giving points. Just be very careful with some of the language that people are using and I want to talk again about and this kind of goes to analytics, but I think it's the next step. So, when you can utilize and there's a ton of different auto QA platforms, right, that are really good, there's a company called Mia Rec that has one, there's Observeai has them that basically you're taking 100% of your interactions, of your calls, whether they come through email, whether they come through chat, whether they come through SMS, whether they come through voice, and you're scoring them.
Speaker 3:And when we initially started AutoQA, I thought that was what we were going to use this for scoring agent performance. And then what I found out is that so many customers were asking well, I have all these cancellation calls, right, we can't really get how many cancels and why. In analytics we can kind of, but so literally we're asking those type of questions. Now, that has nothing to do with the agent. This is actually a IT help desk that we are working with now that is trying to find the root cause analysis of why customers are calling them. What are some of the reasons that? What are the issues Higher up in the form is like the seven or eight ways that they think customers are calling and yes or no questions, and then we can take all that data, put it into a BI tool, give them an actual report that says 100% of your calls were done yesterday and 75% or 75 customers called because of this, 82 customers called because of this, just getting massive insights into the why and this is going to evolve into just like a what CX1 and a lot of these guys are doing with just you're going to see a Google search bar and you're going to be like what were the main reasons why people called and what issues did they have yesterday between the hours of 12 and five, and you're going to get a report right.
Speaker 3:That stuff's here almost now, but this is just a really easy way to do this as well, right? So anything that you can think of of trying to just get data from from a hundred percent of calls, we now have a way to use AI to get it done in literally seconds, all right, and then quickly, all right. And I won't spend a lot of time on this because I think everybody kind of understand this. Where we are today is with the legacy tools and we're deeper into the funnel with the agent tools and we're with the GPT tools. Like those are the mature tools of today. Like, if you went out and purchased this stuff, you don't have to be worried, right, as long as you're going with somebody reputable, that they're not going to do what they say and you're going to find some value in them. We can have a discussion on what is the proper tool for your exact use case, but these are the tools overall that I trust when I consult to say, hey guys, these are OK, go get them.
Speaker 3:When we get into the chatbots and the LLM stuff, I am not comfortable yet at all. Like you guys have seen multiple issues when we go off the rails on a chatbot from somebody that's not really shouldn't be really utilizing them in the way that they are. And the other thing is, you see, is everybody has, like my chatbot is greatai, that's everybody's company. Now, right, and you don't know what's real. And I think that that's that's a fair assessment. And until some of these CCAS guys come out with some of their stuff, um, that we know is going to be a little bit more secure, it's kind of the wild wild West when it comes to, to that stuff and this is another reason why I say don't freak out if you're not there, right, because this is going to be happening. This is the stuff that I, that is on my radar, things that I've talked to everybody this will be here within the next year is real-time language support, so I don't have to have bilingual agents anymore, whether that is typing text or voice. We have a company, we have a French Canadian company that we don't. I don't speak any French Canadian in Erie, pennsylvania. We struggle with that, so we use the language line right and the language line like. That's a business I would be super scared about, because real time translation is almost here, now it's coming. I'm kind of disappointed it's not here as of yet.
Speaker 3:Predictive analytics right. Being able to look not just at your data sets but going out to the internet, going out to social media, looking at your average handle time to see that you like somebody who talks a little bit quicker, because all of your calls are three minutes instead of the five minutes of our average handle time. So let's send them to Johnny over here, who speaks a little bit quicker as well. All of that data to predict the best experience for a customer. Dashboards are going away. You are going to start to quiz and ask, like you do to a Siri, like you do to a search bar, and you're just going to type in the information that you want. Can you give me the agents with the highest average handle time yesterday? Can you tell me what percentage of customers use the phrase too expensive on this specific product and what were the top five reasons that customers called in for service issues and instantly get a report on that?
Speaker 3:Ai for voice is coming. So currently, pretty much everything you need to do is a transcript, right? So if we want to do anything from an AI standpoint from quality to even agent assist it's real-time transcription. Voice is coming, which will then make analytics even better, because now we can really do not just what they're saying, but how they're saying it, not just in a transcript, but literally inflections and tone and voice. They're saying it not just in a transcript, but literally inflections and tone and voice.
Speaker 3:And then again, from a service desk standpoint, looking at different, literally pieces of data or refrigerator talking to a help desk on its own right to have things delivered. Some of that stuff is already here, right, but we're looking at a lot of of of AI. That's talking to AI, which kind of gets a little bit scary. But I think service desk is where you're going to see that stuff happen. First, right From a technology standpoint and you're probably starting to see some of that already and then again, video avatars. I think we get that All right last slide I got.
Speaker 3:So one of the things I think that is pretty cool is a lot of you guys from a service desk standpoint probably have, if you have, like a community or a customer area or even an internal place where there's a KMS where people can kind of go and maybe find their own issues. Maybe they can talk to some other people that had those issues. They can post things and say, hey, does anybody have these kind of issues? It's kind of like a Reddit, right. Well, communities are evolving, especially when it comes to IT support. This is something that was built out, that we built out for a 3D printer company. So they're trying to put a 3D printer into every single buddy's house, right From a retail side, not just a business side. So when they send out a new printer, they put a card in there that says, hey, here's the 1-800 number that comes to us and we get all the analytic data.
Speaker 3:But the other thing is they have a Discord right, and this doesn't have to be Discord, but Discord worked really well. So in the Discord we have fully integrated and you can fully integrate right, all of the connectivity. So when a customer says, hey, I'm onboarding, right, All of the connectivity. So when, when a customer says, hey, I'm onboarding um the one, there's other customers there to help them too. We have moderators that are that are sitting there. So when they hear I need help, if somebody goes into a specific channel, one of our agents start to help them through their technical issues. Um, we have listening bots again that are listening for any type of issues that may be happening in that discord or if a customer is having problems.
Speaker 3:But the cool thing is we can also integrate analytics. We can also integrate reporting. We have a full CRM for that customer as well. So we start to build their customer database without them having to go. You know, hey, can you? You know file and kind of sign up for our list, right, all of this stuff we're creating. And we can actually even go out to social media find out who influencers are. If there's anybody like we can get crazy with it as well, right, but I think building community for IT support is, it can evolve as well, depending on what your culture is with your organization, how you can change things. And again, even when it comes to looking at all that data, data is everything right in this AI world. So, whether it's Discord, whether you have your own communities, anywhere that a customer is giving you information, right, try to integrate with it. Right, so that you can pull that in, so that again, when you get to that AI world, right, all of this information is going to be at your fingertips. It's going to make all of your stuff even better, right, all of this information is going to be at your fingertips. It's going to make all of your stuff even better. Questions I think that's what I got.
Speaker 3:Any more low hanging fruit for you guys? Yes, yeah, so I mean from. Let me just say this first, we're using CX1, and they have a thing called Digital First Omnichannel, so we're fully integrated. We use Slack for our internal employees and we're utilizing this, and that's kind of how they go to Help Desk. And then the same thing for Discord. We're trying to do Reddit. Reddit has open APIs as well, yeah, and so what? I would say you don't even need to do that. I mean you would just connect to the Reddit channel through your DFO and I mean it would be kind of like that. But it'd be much simpler because those are already. I don't want to say they're prebuilt, but you don't need that robust where you're monitoring. But you can pull in some of that data in there as well, just like you would for Facebook, just like you would for any of those things. Anything else. Guys and girls, gals yes.
Speaker 2:I'm just curious because in the contact center, are they using something to look up that they're literally doing everything in? Because we have Genesis, fear Cloud and I still have BMC Remedy. I could see a world where I could get rid of the remedy and everybody's in Genesis because of the way it's the whole ecosystem, that you can do everything from there, but I feel like not everything's in there, so I would need to send the tickets to the areas that are not in Genesis.
Speaker 3:Yeah, I mean I wouldn't trust it yet, like I would have a third party tool. I mean I think they're good for what they do, like I would have a third party tool. I mean I think they're good for what they do, but they're not good for that. Or if you have a specific, again like they're not a CRM, they're not a ticketing system, but they're everything else right. So, even looking at all that data that even goes into your CRM and being able to pull that out, you know that stuff's just in this AI world. Data is everything. As I know that stuff's just in this AI world. Data is everything. I know. I've said it like four times.
Speaker 3:But to have all of that stuff fully integrated so that when the button gets pushed that you want to do some type of first touch point chatbot, some type of AI bot, it's got to know where all this stuff is coming from and it's got to be able to talk to all that stuff, so does your platform.
Speaker 3:As far as they want to be able to connect to the panel and then pull all of so with auto or with what CX1 does with analytics. So when it comes to a customer's KMS, we're basically uploading it. We're using CX1 for that, for their enlightened platform. Correct, now you can, right, you can like, so I can through APIs. I can connect, like, if you have a cloud based KMS, right, we can connect through APIs and we can, I mean, uploading it. I guess we're saying kind of the same thing, or not? Or if we actually have physical PDFs, right, we can actually upload that stuff too and we can connect through through connectivity through apis, to whatever their kms is, if they have a cloud based kms or anything like that I mean talked about customer sentiment.
Speaker 2:Yeah, um, how do you determine whether that interaction was a positive sentiment or negative sentiment, like what's? What's the underlying that gives you that insight?
Speaker 3:well on the. So again, this is. There's two answers to this. One's a cop-out answer and one's a better answer. The cop-out answer is I don't really know what interaction analytics does right, like their proprietary scoring for sentiment. They won't give away. I will tell you that it's pretty spot on.
Speaker 3:The thing that's happening now is real-time analytics, right. So you know like we can look at a screen and I can see agent, and I can get a ping on my screen that says there's a pissed off, ticked off customer, right, and they're negative right now and this call is going south. I can also get it for the same thing, that is, there's an agent on there that has now raised their voice. They've said a word that they shouldn't have used, right, so, and all that plays into their sentiment score, which is really on point Now when it comes to things like so for our AI platform, for AutoQA. Now, when it comes to things like so for our AI platform for AutoQA, it's more what they said and inflections of intent. So we have found the best prompt that we can possibly find and I think it's on point for scoring sentiment positive, negative or kind of nothing for that, and that's more of a prompting aspect where we're looking for tone.
Speaker 3:You can get inflections from a transcript, right. You can kind of get you know how that call is going based on what the customer is saying, and there's also very intent based analytics that kind of go into that too, when you see what's going on with the customer and the questions that then they start to ask. So again, I don't know if that's a great answer for it, other than there is, like I can even prompt. You can prompt for the unseen.
Speaker 3:I know in service desk it's very difficult to use things that are not in a transcript, like, let's say, you wanted to auto score calls, but this customer has your agent has to click on this and go to this screen and then download this and send that to them and then fill out an email, right. That stuff's a little bit difficult, but there's so many things that we can do from from seeing the unseen with, with what actually happens on those calls, right, that you can kind of repeat where an agent will be like all right, let me bring that up. Okay, good, let me. I know that sounds hokey, but when you do it over and over and over and you kind of see it and you kind of benchmark it. You can start to almost quote unquote, see the unseen, until we get a more cost effective way to kind of actually look at screens, which you can do.
Speaker 2:You can do now it's just not that cost effective and it's a little bit harder to deal with. Yes, I just have a question, kind of going off of that with customers and then, as you're, adding in more AI.
Speaker 3:How do you think that positive and kind of is there a regulation in place for that? You're not going to have real-time transcription, you're going to have real-time voice, so like with agent assist. Right now, if you go behind the scenes of agent assist, it's literally like agent talking Hi, how are you? I'm great, how are you? My name is Tammy, I'm from XYZ. Oh, thank you, tammy, I'm calling on. This is my issue.
Speaker 3:Right, you like read literally what's going on, right, and into play. It won't just be that, did they say the proper thing? But then we really get tone, we really get inflection, we really get sarcasm. Right, sarcasm is extremely difficult for AI to kind of get right now unless you prompt for it. Like, if the agent says this, then you know it's that and you can prompt for all that stuff that we found through through our platform. But you can literally prompt for almost anything. If you have a use case and you kind of just have some logic and you kind of go back and forth a lot of that stuff you can kind of figure out to auto score.
Speaker 3:But no, I mean, I think it gets better. Like, I'm an anti-CSAT, I'm an anti-NPS guy, like, and just because I think if you send a survey out, right, you're going to only get the people who are really ticked off or you're going to get the people who had a great experience Just because they were ticked off, it might mean that the agent did a policy that you guys approve of or that we approve of that the customer didn't like. Right, like I can't give you that refund or whatever that is. I can't send you something else out because that's the policy had nothing to do with the agent. But they get hammered on it. So I just like again, I'm totally fine with sentiment analyzing a hundred percent of the calls not scoring but analyzing and then really kind of using that as a marker of is the agent doing a good job servicing? And again then you get into QA and all that. But I think it's better than us hammering reps on CSAT and NPS scores. Me personally, Anything else.
Speaker 2:What's coming up in the podcast, anything intriguing that you want to share.
Speaker 3:So I'm getting so AI'd out too that I'm trying to go back to roots for the next at least quarter. Right To talk about more agent engagement stuff to talk about, we engagement stuff to talk about. Um, we've redone our management training program. We're going to do a full episode on on kind of that and some of the things that we're going through with with our middle management, those kind of things. And again, the ai thing, I think, is it just evolves so quickly that you can easily do an episode every week on what's evolved so quickly. But I think you know we're, we're let's take it in chunks now. Um, because we're seeing things just kind of go so fast, uh, but yeah, I mean I think, uh, the we're going to have on the CEO of CX, one Brock's going to come on, so that can be a kind of a cool episode where we can kind of drill down on on what, what those kinds of things are thinking.
Speaker 3:But other than that, it's just me talking to the world. Hopefully some people are listening. That's kind of how it works. You as a basketball coach, the girls, let's be analytical, right, no, but they mock my podcast Like how many times I've come back like the game is over, we won. And they're blaring the podcast in the locker room Like guys seriously back like the game. The game is over, we won. And they're blaring the podcast in the locker room like guys seriously like um, but no, I mean, we, we kind of, we joke a little bit about that kind of stuff. Yeah, yeah, yeah, for sure, for sure. Well, thank you guys appreciate it, thanks tom yep.
Speaker 2:Thank you.