Easing Into the AI Revolution
Kim: Hello, everyone. Welcome to today’s webinar, “Easing into the AI Revolution.” My name is Kim, and I’m going to moderate for you today. I’m just going to go over a couple things up front, give a quick overview of the agenda, introduce our speaker, and then we’ll get started. First of all, we are recording this, and you will receive the recorded version of it after the webinar. You can keep a look out for that. If you have questions at any time during the presentation, you’ll see a little chat messaging icon at the top of your screen, you can just send messages through there. I’ll collect them up and direct them to our speaker at the end of the session. Again, any time throughout the conversation, you have a question, send it in, and I’ll just collect them up.
Today, we are going to be talking about easing into the AI revolution. We’re going to start off with a quick overview of AI in the contact center, just to kind of set the landscape and give you an idea of what it actually means. Then, we’re going to go into a practical use case for implementing, kind of those first steps on how to get started with it, the three key business benefits, and then actual actions for you to get started. We’re going to close out with kind of a real-life case study of one of our clients, which is MGM, to talk about how they’ve implemented everything that we’re talking about today.
To introduce my speaker, we have Charlie Schrier. He’s the Director of Product Marketing here at SmartAction. He’s spends a lot of his time with industry analysts, with our customers and prospects, to valuate industry trends that impact the evolution of our customers, and as well as our solutions themselves. He specializes in content creation and thought leadership, and he represents us at a number of live events, as well as my virtual webinars. Being our key liaison to the analyst community, it’s given him a lot of great insight into the current state of AI-powered self-service, and as well as look into the future trends and expectations. This is a great area for him to speak. He’s very familiar with this space. And I am going to pass it over to him to get started. Charlie?
Charlie: Thank you, Kim. Excited to be talking to everybody. As Kim mentioned, having a lot of conversations around this topic with a lot of different types of people – analysts, salespeople, customers, end users – and it’s definitely a very hot trend. One precursor to the webinar is I’m fighting a cold today. If I do have to mute to cough really quick, don’t worry, I haven’t left you. I will be back. But that’s also why I sound a little nasally today. But still, very excited to have this conversion and looking forward to answering any questions you guys have.
As Kim mentioned, we’re going to start with an overview of contact center AI. And even in just the few years that I’ve been at SmartAction, these definitions have really evolved, and people’s understanding has really evolved. But everybody knows that AI is such a hot topic, both from a contact center, but also a business and a consumer technology function as well. And what we’re trying to do is help you boil it down to something that’s a little more consumable and digestible and practical for your actual world in the contact center. The way that we think about it is these two types of AI – there’s conversational AI and there’s process automation, or you may have heard about robotics process automation, RPA, focused a lot on analytics. I’m actually going to start there, and then we’re going to have a much deeper conversation about conversational AI, as that’s the key piece of this webinar and how to ease in.
Just to give you a quick overview of the process automation side. It’s working more in the background, working on back office processes that humans are typically required to do. But these processes are very repetitive. They’re very input/output-type situations. You might think about data entry or policy reviews, notes on policies, and that type of thing. And automating some of those using training sets and machine learning but helping to increase the efficiency of both your people and your processes, move a little bit more quickly and a little bit more accurately using technology. It’s more of an indirect relationship with the customer. The customer is certainly benefiting from this background automation, but they may not recognize it as much. They may just sort of feel a little bit more efficiency.
On the conversational side, this is really a communicator. This is on the front end, having conversations with actual customers and consumers. There’s obviously plenty of consumer examples out there, but that also benefits from machine learning, using training sets to help train the AI, train the speech recognition and the natural language. But it’s very channel-based. There are voice-driven platforms, there are text-driven platforms, chatbots, you name it, but we really think about it from the voice perspective. But it’s actually providing that customer service and support on the front end. Next slide?
Kind of a cool stat that we saw here. A prediction I guess of, as Gartner clients, we see a lot of this type of stuff, but by 2020, the average person will have more conversations with bots than with their spouse. Kind of interesting to think about, particularly if you’re in an airport, or you’re in the mall or walking around. People are always looking down at their devices, at their phones, their computers. And whether or not that’s a bad thing, I’m not sure. But we’re definitely much more likely to interact with bots moving forward than we ever have been before, and we may not even know it. And I think that that’s a great precursor to this kind of conversational AI discussion.
Let’s talk a little bit more about the characteristics of conversational AI, and we can get into a few more pieces of the puzzle there. Clearly, exceptional speech recognition and 100% natural language are going to be really important. Because “conversational,” that implies that you have to be able to understand what people are actually saying and typing. I think on the voice side of this, people kind of have these preconceived notions about, “Well, I’ve gone down the path of a natural language IVR before, and it did not work.” And they have some fears and are rightfully worried about what that might mean in terms of the training sets, the implementation times, etc., etc. And I think conversational AI is actually the next evolution of that natural language IVR. It’s actually pushing that technology forward and solving a lot of the problems that you may have faced if you had that preconceived notion around national language IVRs in the past.
But exceptional speech recognition, understanding what’s said, but also what’s meant by what’s said. We’ll talk a little bit more about that later in the webinar. A hundred percent natural language, that’s really the conversational piece of the AI puzzle. But a couple of other things. Ability to handle more complex exchanges. The way that we think about the world in the contact center, and we’ve been in the contact center space for almost 10 years now, and the way that we think about it is that, in the contact center, there are simple transactions and engagements that come through, there are complex, and then there are human-only. The simple ones are things that you can probably handle using your traditional IVR system or your legacy IVR system, very simple chatbots, that type of thing.
But more complex things are things that humans, up until very recently, had been handling, and those are the more complex, but they’re repetitive, very similar to that process automation stuff I was talking about. But it’s actually on the conversational front-end side of things. You’re handling these processes, agents are handling these processes that are repetitive, but they’re somewhat complex. They’re not able to be done with the traditional and legacy systems. Then, there’s those human-only. Those are the moments of truth for customers, tip of the sphere, most complex types of things, situations, things that require empathy, selling, that type of thing. What conversational AI is trying to do is push that line of what separate complex but repetitive and human-only, push that line farther and farther down.
We have some pretty good graphics around this that we’d be happy to share with you if you’re interested. But that’s the idea, that we’re automating more and more complex exchanges than we ever have before. And then, lastly, the ability to continue and learn. Machine learning is a topic that comes up constantly, and a lot of people think that that’s one of the only components of AI. It’s one of many components of AI. But machine learning is a part of it. It helps with the speech recognition and improving that natural language. It also helps with some of the intent capture and rules-based types of things that the AI is doing. Being able to pull from that data and learn from it is going to be important as well. Next, please?
Okay. Conversational AI, you know a little bit about it now. You’ve probably interacted with it. The value…I’m not going to [inaudible 00:09:29] this slide. I think those three topics across the screen are going to sound very familiar to you around any technology you want to implement, or even any process you want to tweak or change. Customer expectation. You know they’re really high, you know that they’ve evolved, and you know that people are more comfortable interacting with bots and machines than they ever have been before. The customer experience, that’s why we’re all here. We know how important it is to businesses and to what businesses are trying to achieve in terms of loyalty. And then, lastly, cost savings, delivering ROI. But not necessarily in reducing human staff or eliminating jobs, or anything like that. But more in optimizing the staff that you have, optimizing the processes that you have, and increasing the scalability of what you’re able to do from a customer service and experience perspective. Next, please?
Okay. How do you get conversational AI? And the title of the webinar speaks to it exactly – “Ease into the AI Revolution.” How do you ease into that revolution? One of the ways that we’ve found to be really, really valuable, and a way that you can kind of “dip your toes” into AI without this massive overhaul or massive undertaking of a project is through what we call an “Intelligent Front Door.” You’ve probably heard of the natural language, “How can I help you?” systems of the past. When you put the conversational AI on the back end that’s working in tandem with the natural language, that’s where you’re really able to see the benefits and the improvements, and that’s where the word “intelligent” comes from, around the actual better routing.
What is it? It’s basically, as opposed to offering menu-based routing, “Press 1, press 2, press 3,” or a directed dialogue-type of situation, you’re just asking, “How can I help you?” and using the natural language and the AI, capturing that intent. It’s important because it helps you achieve the value that we just talked about on the previous screen with conversational AI, conversational communication. This has been a way that we’ve found to be a really valuable way because it doesn’t require that massive overhaul, and we can talk more about that in a second. Next slide?
Great. Let’s talk about how it works. And even before we get there, what are some of the problems that we’re trying to solve here? You don’t want to just stick a solution where you don’t have a problem, but I think these problems may sound familiar, if you’re familiar with the contact center space, or if you’ve even contacted Customer Service in the past. Some of the problems that are faced when you face a menu that is limited in its options are, well, first of all, people don’t necessarily trust IVRs right off the gate. Oftentimes, you’ll get someone 0-ing out and trying just to get to the agent as fast as possible. What ends up happening is that you’re not capturing anything about that customer. And when the agent receives the call, they don’t know anything about them either, and they’re not able to help them. Likely, there are skills involved and the agents ends up having to transfer anyway.
There’s also issues where the menu can be very limited, in terms of what options are available. And people don’t totally understand what they’re looking for. They may have a specific question, but they’re not sure if it should go to Billing, or if it should go to Technical Support. And they end up in the wrong place, and now you’ve got some of the same issues, where you’re not exactly sure where that person should go. And then the transfers become an issue too. Maybe the agent doesn’t know exactly where to transfer, and that has to do with several other challenges. But at the top line, the problem is that people don’t totally know what they want and where they need to go. What ends up happening is you have these inefficiencies across the contact center due to the menu-based routing, and you’re not necessarily able to capture all the data around, “Well, this was actually a misroute, and then there was another misroute.” And that’s hard to follow and hard to capture.
How does an Intelligent Front Door help you solve that problem? Well, it greets every customer with a “How can I help you, today?” And using the AI and the natural language integrated together, it’s able to actually capture the intent at a much higher confidence level, with also a higher confidence level around what that person means and where they actually need to go. And we can use some examples, and we have the case study at the end, where we’ll talk a little bit about how we do that. But with access to data, you can actually make this conversation even faster than that, recognizing the caller at the output of the call, calling out maybe their most recent transactions that they’ve had. For example, “Hi, Jenny. We see that you recently placed an online order. Is that what you’re calling about today?”
These types of things, the more data that’s available to the conversational AI platform and the Intelligent Front Door solution, the more the greetings and the more the experience can be customized and personalized. In addition, if you’ve got the automation at the front end of the call, you can add in a lot more automation and self-service into the back ends of some of the calls for the actual processes and tasks that people are calling about. What’s kind of cool about the Intelligent Front Door, and then we’ll move on to the next slide, is that you’re able to start to capture information about why people are calling, and allow them to get acclimated to the automation at the same time that your business is getting acclimated to it as well. It’s sort of this dual-purpose, where both of you are getting to acclimate to it, and then they become more comfortable with the automation later in the calls. Next slide, please? Thanks.
Okay. What are the business benefits? We’ve talked a little bit about what the solution is, what it means. And, “What are the benefits?” is going to be the next piece of the conversation. Data, ROI, and it’s future-looking. The first benefit, data capture. Actually, I was just having a conversation with one of my colleagues this morning about a customer that they have who is getting like half a million calls a year, and they don’t have any piece of automation in there. They’re very opposed to it. But what they also don’t have is any data around why people are calling. You just sort of have the qualitative information from some of the agents. What’s this type of solution can really help you do, even more so than even having humans on the front end of this, is capturing all the data in the same place and understanding why people are calling.
Not only are you providing a good experience for customers who are comfortable talking to automation and comfortable moving through the system that way, but you’re also learning about and getting a better view of that Holy Grail, which is the 360-degree view of the customer journey. You’re starting to really get a look into, “Okay, it seems that people are calling 50% for this and only 10% for this. Why are we dedicating so much time here, as opposed to there?” You can also start to be more proactive about outreach to your customers, or proactive even in just answering the call, as I talked about before. And it allows you to also recognize places where you might be able to offer that additional self-service.
One of our customers is a retail customer. And it’s a subscription-type of service, so people are often calling to skip the month for their subscription or to check on an order, or maybe to return an order, and those seemed to be the biggest call drivers. The only way we were able to capture this is through an Intelligent Front Door. Now, we’re able to offer automation for those other types of tasks that are the main call drivers, and that’s opened up this whole new world for their agents, in terms of optimizing them, helping to provide upsells, helping to provide more value around that person’s actual retail experience. Providing those additional self-service experiences is very valuable as well. Next slide?
Okay. ROI. This one, I think everybody understands. But we’re big subscribers to Customer Effort Score, and we believe that customer effort is directly tied to loyalty. And when you have loyal customers, that’s ultimately tied to revenue. What an Intelligent Front Door can help do is reduce caller effort across the board. How does it do that? Well, it provides a more effortless experience. Because remember all those misroutes I was talking about in the Problems section of this webinar? A lot of those start to go away when customers are able to describe their problem themselves. Now, in the beginning in the first week, two weeks, three weeks, you’re learning about how your customers are describing they’re problems, and you’re getting better.
And that’s one of the benefits of conversational AI, is the improvements, the machine learning, the ability to make quick and valuable adjustments to the type of solutions you’re providing. But as you continue to grow and as you continue to learn and train, and teach the system, you’re starting to really optimize those processes. You’re really starting to optimize those agents, because they’re not dealing with a ton of misroutes. They’re not dealing with a bunch of customers that they have to authenticate and go through a series of questions just to figure out why they ended up on their headset today. Because what we’re doing at the front end of the call is validated and understanding exactly what’s going on, no matter how the customer describes it. Again, it’s that natural language integrated with that AI that makes that possible.
The final piece of that in there is the scalability of the solution. No matter how many calls per day, or maybe…Another example, in roadside assistance, you can imagine that when it snows, or I know that when it rains in L.A., everything stops, but the calls to roadside assistance tend to go up around bad-weather situations, which can often be unpredictable. Being able to answer that front portion of the call and then possibly, here’s where the real ROI comes in, now you’re starting to automate some of those additional self-service opportunities. That’s where the ROI can really compound itself if you’re able to automate that front end, understand the key call drivers, and then scale to whatever demand is necessary.
Okay. Lastly, it’s future-looking. If you look at that quote at the bottom of the slide there, it’s a little bit scary. Nearly 80% of contact centers don’t have the systems in place to meet future needs. Are you one of the 80%, or are you one of the 80%? Hard to know. But this is one of the ways that you can start to scale your contact center using AI, so getting AI in place in a very easy-to-implement and not an overhauling, rip-and-replace way, to actually start your company on that journey of introducing AI, starting to meet the future needs of not only your business, but also your customers. It’s not just a buzzword anymore. We have to make it practical. We have to make it work.
This is one of the ways that you can be future-looking. You can start to move your company forward into that AI revolution in a non-rip-and-replace, not difficult way. And what’s kind of cool about this too is that you can implement it on top of any of the systems you’ve already got in place. If you’ve got contact center infrastructure you’re happy with, if you even have an IVR that you’re happy with, right, you can sit this right on top of some of those legacy systems and start to capture the data, start to evaluate that. And then you can iterate and improve and add self-service, or start to handle other different processes, start to be more proactive and add that to the customer experience. There’s a lot of different things that you can do just by sitting this on top of the infrastructure you’ve already got. Next?
With that in mind, we wanted to show you a practical use case and hear from a real-life, in-production case study from one of our clients, MGM Resorts International. They’re a great client for us, and I want to talk a little bit about how they’re using the Intelligent Front Door, so you can get a better understanding of all of these concepts that we talked about today. MGM Resorts International, they own and operate about more than 25 properties across the world. And you’ve probably seen many of them in Las Vegas, if you’ve been there for either an event or for some fun. One of the problems that they were facing was that every inbound call to one of their properties was going to one specific call center. Those call center agents were handling everything from booking new reservations to, actually, towel requests or Room Service within the actual hotel itself. What was happening was these agents were really bogged down with a lot of different types of calls, and they weren’t sure exactly what type of call they were getting when it would get passed to them.
What it also meant was longer hold times for everyone that was calling, so sometimes up to five minutes that someone would be waiting on hold before they even got to say what it is they were calling about. These were some challenges that they faced. Ultimately, what would end up happening is someone who was calling about something within the hotel, if there was a department within the hotel, oftentimes that agent would get the call and then have to reroute it back to another place. These people are sitting on hold, only to then be rerouted again. Some challenges there, as well as the fact that the agents that work in that call center are also really highly trained. They have a lot of experience in providing upselling opportunities, all the way to offering and explaining luxury services that come along with some of the hotel amenities. They’re very well-educated, they’re very well-trained, and they do a lot of great things that really only humans can do, like sell and provide empathy, and kind of feel out what a customer is looking for in terms of their experience at one of the properties. And they just weren’t being utilized in the right ways.
What we did was we implemented, as kind of a proof of concept, an Intelligent Front Door for just one of their properties in Las Vegas. And capturing intent at the front end of the call, “How can I help you today?” we’re actually able to provide 95% success routing for all of these calls that are coming through. That includes things that are within the hotel to Room Service or to ask for more towels, as well as things like luxury services and upselling opportunities. What we’re able to do is actually find out, “Okay, what are people actually calling about?” And the beauty of it too, is that because we’re able to automate that first portion of the call now, and we know why people are calling, we can start to find other automation opportunities to provide self-service for customers throughout the rest of that call experience. Things like, “I need more towels,” or, “I want to order Room Service,” or, “I want to book tickets to a show,” these can all trigger either self-service or agent experiences. It’s saving many, many agent minutes per month, and it’s reducing time on hold because people that are calling this property aren’t having to wait on hold just to get the one thing it is that they need. They can get routed right away.
One other cool piece about this project, and then we’ll kind of move on and my sales pitch will be over. But they were so happy with this that we’re rolling this out to 11 other properties, with plans to continue that self-service journey, and plans to continue to automate more Front Doors for the rest of their properties as well. It can be really valuable, and there’s just a use case for you right there. I tried to litter a few other use cases throughout with retail and roadside service. And if there are questions around some other places that you might be able to provide a Front Door, we’d definitely be happy to take those questions. I’m going to pass it over to Kim, and she’s going to talk a little bit about follow-up and what some of the next steps are. But I really appreciate everybody’s time today. And hopefully, if there are any questions that come through, I’d be happy to answer them. And hopefully, you learned a little something about how to use conversational AI in your contact center. Kim?
Kim: Feel free to reach out to Charlie directly. His email address is on the screen right now, if you have questions. You can also reply to any of the emails that I’ve sent you leading up to the webinar. And in the follow-up email as well, with the recording, you can reply back to that, and I can get questions to Charlie. We actually currently have an infographic that is on the same topic, and it doesn’t go into as much detail as the webinar did, but it’ll give you a really nice overview of the process and the benefits in some really nice, beautiful graphics. The link on there, smartaction.ai/ifd will get you to that infographic. And then, for just more general information, you can reach us at smartaction.ai. Thank you, guys, for joining. You can expect emails from me at the end of the week with the recording. Thank you.