The New Digital Frontiers of Customer Service Excellence

The Gist

• Large language models revolutionize customer interactions. Large language models facilitate more natural and effective communication, streamlining processes and harnessing corporate knowledge.
• Contact centers must meet rising customer expectations. Modern contact centers need to anticipate customer journeys and provide seamless experiences.
• AI should prioritize service excellence over cost cutting. AI integration should focus on enhancing customer experience, leading to natural cost reductions through improved processes.

Episode Highlights

• Digital transformation accelerates post-pandemic. COVID-19 has expedited the shift to digital-first customer interactions, redefining the role of contact centers.
• IVR and CRM integration enhances personalization. Advanced IVR technology, integrated with CRM systems, enables personalized and efficient customer support.
• Improving self-service requires proactive solutions. Effective self-service systems should recognize customer intent and escalate to human agents when needed.
• AI enhances predictive analytics in customer service. AI-driven forecasting provides real-time insights, optimizing contact center operations and customer satisfaction.
• Leveraging AI for customer journey insights. AI tools can predict customer effort and satisfaction, allowing proactive management of customer experiences.
• Transformative potential of AI in contact centers. AI is reshaping contact centers by automating tasks, personalizing interactions, and improving resolution times.
• Adapting to digital-first customer expectations. Contact centers must evolve to handle the increased demand for digital interactions and provide seamless support.
• AI supports multilingual customer service capabilities.
Advanced AI models enable seamless communication in multiple languages, broadening customer service accessibility and effectiveness.
• AI-driven chatbots enhance initial customer engagement.
AI chatbots provide instant, accurate responses, improving initial customer interactions and reducing the need for human intervention.
• AI facilitates continuous improvement in contact centers. AI tools help identify areas for improvement in real-time, allowing contact centers to continuously enhance their service delivery and operational efficiency.

Editor's note: This podcast transcript was edited for clarity.

Dom: Hello, everyone, and welcome to another enlightening episode of “CX Decoded,” your go-to source for the latest trends, insights, and advice in the world of customer service and contact centers. I'm your host, Dom Nicastro, editor in chief at CMSWire, and today we have a special show lined up for you. We're going to dive deep into the transformative power of AI in customer service and how it's reshaping the way we interact with customers and streamline support processes.

We’ve pulled together some insightful takeaways from two interviews from our Beyond the Call show on CMSWire TV: Raj Krishnan, partner of technology at Microsoft (check out his TV show), and Sean Albertson, founder and CEO of CX4Rocks and former customer experience leader at Charles Schwab (check out his TV show). Both of these experts bring a wealth of experience and knowledge in the realm of customer service and AI, and they have a lot to share about the evolving landscape of contact centers in this digital age.

Throughout this episode, we'll be highlighting five key takeaways from my interviews with Raj and Sean, providing you with a comprehensive understanding of how AI and analytics are revolutionizing customer support.

The Power of Large Language Models in Customer Service 

Dom: Large language models are revolutionizing customer service interactions. Raj discusses how these models enable more natural communication with systems, allowing contact centers to harness corporate knowledge effectively and streamline processes for improved outcomes.

Raj Krishnan: So you know, one of the big opportunities that we have with the large language models, you know, we like I was just reading a report this morning, right? All these web self-service bots, were kind of picking up. I think what these new large language models have dramatically changed is the ease with which now people can actually interact with systems. 

And to me, that's the biggest benefit, right? So we're, for example, when we used to build a customer service bot, we had to use these natural language models to say what is the intent, what are the different utterances? How, like say I want to book a ticket and how in 10 different ways can I say so that I teach the language model to understand what we say. What has dramatically changed is that I just simply talk the way I talk, and the system understands it.

That's the power of the large language model, right?  However, the biggest limitation with that, along with it, comes these large language models are trained with a corpus of data, and to grounded to what is relevant to your data. like, for example, if I'm a customer service person with my own data, how do I leverage this ability of these language models to understand what I'm saying, but use my data to be able to respond?  So those are the areas where I see tremendous opportunity.

Why is it important? One of the big challenges is these contact center representatives, you know, they all come in with different varying skill sets. Somebody has deep skills about your domain, some of them are new. To me, what the AI is going to do is just set that level so that everybody can leverage corporate knowledge, and then interact with it very easily. I think a lot of the projects that I'm working on are more not about just using AI, but taking a business process and infusing AI everywhere possible to simplify and improve the outcome.

Related Article: AI's Transformative Role in Customer Support and Service

Meeting Customer Expectations: The Evolving Role of the Contact Center

Dom: Our second takeaway, Sean discusses how modern contact centers are grappling with rising customer expectations, where knowing a customer’s journey before they even reach out is no longer just a bonus, but a necessity.

Sean Albertson: Well, being that, you’re the second, third, or fourth, or fifth step that the client has taken, you know, ultimately in the contact center, we're already behind the eight ball in the consumer's mind, in the customer’s mind. You know, we of course don't know what they've tried necessarily and where they've been and how long they've already been working on it.

The reality is though, from that seamless experience the customer expects, they expect us to know in many ways what they've done and the effort they've already put into resolving their issue. So by the time they talk to someone, they're usually at a level now of, well, I've done everything I know to do in an easy way, and now I've got to talk to you. And so that puts extra pressure. Going back to your point, when the call center was the first stop, there wasn't that much expectation. 

But now there's kind of this expectation. You should know what I've already tried to do. And God forbid you should know if I tried to call you before, and this is my second or third call about a subject. And so we're already kind of starting off behind where the customer's expectations are in the contact center. And we have to think differently and ultimately act differently to really prepare ourselves for that new expectation.

Related Article: 4 Steps to Shore up Your CX and Meet Customer Expectations

Prioritizing Customer Service Excellence Over Cost Savings with AI Integration

Dom: Now we’re talking about the shift from cost-cutting to customer service excellence and leveraging AI to provide timely, accurate information that not only enhances the customer experience but also leads to natural cost savings by streamlining processes and reducing resolution times, and Raj drives that home here.

Raj Krishnan: To me, I think the cost savings should not be the goal, right? The goal should be, hey, the way that I serve the customer and want to serve them faster, I won't serve them with the deep knowledge of what I have in my doing this for over a year, over years. And then how do we use all that and provide the best customer service experience within a short period of time? This is automatically going to result in cost savings, right? It's not about how can we get rid of five people using AI. That's a very wrong approach to AI right?

The thing is looking for processes where AI can Oh, like one of the things that I mentioned, is this, right? like, I'm just working for the same company. And one of the things they do is they go and, like, create a knowledge repository by looking at your websites and create a, what we call a vector database for retrieval argument generation. 

So that when a question is asked, it goes and finds the answer, instead of somebody having to go copy and paste and then craft the response for you. To me, what it does is provide accurate results very quickly, and then you're addressing the customers’ issues and reducing the time. That's how we should look at it. That in essence, it'll automatically reduce costs, and improve customer satisfaction and provide the right information at the right time to the customer.

Accelerating Digital Transformation: The New Landscape of Customer Interactions

Dom: Sean next gets into the rapid acceleration of digital transformation, particularly spurred by the COVID-19 pandemic, and how it’s reshaped customer interactions. He looks at how companies have adapted to a digital-first approach and the evolving role of contact centers in this new landscape.

Sean Albertson: Well, we were already going to a digital environment prior to COVID. There was a lot of digital transformation exercises and a lot of companies were making those improvements. I mean, we've been doing that steadily in the early 2000s and through, but man, COVID just put it in high gear.

And all of a sudden everything, you know, if they weren't, if a company wasn't thinking about going digital first, they were during COVID because that was the only way obviously that we, that they as an organization could survive. And so it's just put that much more pressure on the environment. Now the value to the consumer is they've gotten a lot more options.

And now with, of course, the apps and the websites and chatbots, I mean, digital now has become the starting point for every interaction. And the challenge for most of us that have grown up in the contact center is now we're at, as you mentioned earlier, kind of that second seat. We come in after usually the initial work online has.

And it's been an interesting migration, but that acceleration of that digital transformation, digital first activity, it's really put us in a position where we have to think differently about our role.

Related Article: Transforming Your Digital Customer Experience With AI

Integrating IVR and CRM for Enhanced Customer Experience

Dom: Sean discusses advancements in IVR technology are revolutionizing customer service. He gets into how the seamless integration of phone systems with CRM allows companies to personalize interactions and streamline the customer journey from initial contact to human support.”

Sean Albertson: Well, yeah, I mean, think about the IVR, the phone tree and phone system, you know, from touch tone in the past and now talking to it. Well, at that point, you know, they know which number you've called in on, you know, using your Annie or your actual phone number to find you, assuming you're calling on the number that which nowadays everybody calls from their cell phone because nobody really has home phones anymore. So, yeah, they know who you are.

That integration then allows, you know, the IVR and that upfront system to kind of look into the CRM. And say, all right, I know this customer. I know their account. I know all sorts of good information about them. And that is being used and has been used then to present to the agent to say, hey, here's who's calling. But it's a tunnel, if you will, data upfront from an offline or a non-human interaction to then presenting to that human interaction within that integration. And that's the key behind it, for sure.

Enhancing Self-Service: Addressing Frustrations with Proactive Solutions

Dom: Now, we’re discussing the challenges and opportunities in self-service technology. We'll explore how improving self-service systems to recognize customer intent and seamlessly escalate to human agents when needed can transform customer interactions from frustrating to fulfilling.

Raj Krishnan: Yeah. So when we say self service, the biggest challenge we have had is that people like to start with our IVR, a bot, right, two or three interactions, you just want to shut that thing and just go talk to a human being because of the poor experience that we have. To me, this is where like we should be able to identify the things that people would emit. like, let's say, I want to get the status of an order, you know, I certainly don't want to talk to a person, I wish I could just go to the system and say, Hey, I'm the customer. And then it automatically uses my voice to verify who I am, and once it knows who I am.

And it goes and pulls my orders and tells me Hey, I know what you're calling about because there's an outstanding order, here is the status of that to meet that self service. Right? So service is not a customer asking other questions, getting frustrated, not getting what they want to be proactive and using that, to me, those are the candidates concerns of it.

However, there may be certain scenarios where I ask a couple of times the system doesn't respond, automatically escalate to a live agent, bring all the context and then tell the live agent, hey, this person is almost losing their patience. They've tried AI, but now let's let the human in. So kind of blending that based on the context is what we need.

Related Article: Customer Support: Definition, Importance & 5 Essential Strategies

AI and Forecasting: Revolutionizing Predictive Analytics in Customer Service

Dom: AI is transforming the landscape of predictive analytics in customer service. Raj’s explores the shift from traditional data modeling to advanced AI-driven forecasting, and how this technology is not only enhancing accuracy but also providing real-time insights to optimize contact center operations.

Raj Krishnan: I mean, so you know, the like, actually, that's a more of an established machine learning area, right? But surprisingly, I'm seeing more and more, the AI is almost taking over everything, right? I used to think that if I needed to do a forecast, I needed to get historical data, and then feed it and all that.

But now I just saw the other day, where you tell the thing, hey, can you just predict the, you know, sale of appliances for the next four years. And you're asking a ChatGPT, or this thing, this thing is able to actually come without me providing some of these things, looking at patterns and the existing data, whatever it has, and be able to actually create that Arriba model, the forecasting model, right.

So AI is almost creeping in into more of this. But I still believe that traditional forecasting, having the right input data, modeling it and knowing the variables, that the whole hyper parameter, the tuning the model, that is still an area that, you know, it requires a bit of a machine learning capabilities. To me, the contact center is one of the major, major things, right, in terms of predicting your load, how many agents should I have?

And what times and what like, you know, there was a project that I did I remember, we were just trying to predict the inflow of customers into a fast food restaurant, there are so many things that goes into it, one of the things we did was a real time camera, watching the inflow of vans, right to say, on an average, four people in the van, now I started five in the van coming and I'm seeing a trend go because there's an event.

Now that's a real time impact. So when you talk about particular call centers, right, something is happening, there is a, you get a lot of, so modeling and all that it will always remain a complex thing. So you're going to need some solid machine learning capabilities and scientists, data scientists to do those types of things.

Dom: Sean highlights the importance of integrating predictive analytics into customer service strategies:

Sean Albertson: Yeah, absolutely. And that is, you know, that's a key, right? And, you know, again, predictive analytics isn't new either, but fundamentally it's, it's getting easier because a lot of us are putting our data in the cloud. Our data is now more joined. We can start to use it in that way. And so, yeah, using as an example.

So what my rocks program, for instance, does, it basically looks at all of the relevant data associated with experience, you know, call center metrics, operational metrics, survey results and survey metrics. It looks at text analytics and journey analytics, the stitching together of the actual physical journey of the customer. And it uses AI to predict high effort, low effort across those scenarios. So think of it this way, you can look at, for instance, your call transcripts and tie that to this post-call survey, because again, we all know. customers don't leave, most customers don't leave comments and most of the comments we get are kind of useless.

But you merge together and use AI to study it, you can now look at the call transcript and say, what out of the calls is predicting high effort or lack of loyalty or sentiment or things of that nature? And then that predictive analytics, you train your models on the past data and now you can predict the same metrics, survey metrics if you wish on all the transactions that never took a survey. 

You literally can look across all your transactions and predict what their experience would have been or what their score would have been. And now you take proactive action to go after those clients and say, hey, we noticed this was occurring or even better, some of the tools and some of the vendors out there are getting even faster where you can almost do it in real time and you can interrupt the bad journey.

And try to redirect it in a positive way. And so those tools are becoming even more and more available using AI and using the power of data analytics, bringing those things together.  And certain activities and certain metrics will predict certain results better. That's part of the predictive analytics you run. I mentioned earlier, lack of resolution, the No. 1 predictor of high effort. Not the only, but the No. 1, the highest correlation between lack of resolution and high effort."

Dom: That wraps up our key takeaways for today. A big thank you to Raj Krishnan and Sean Albertson for sharing their invaluable insights on how AI is transforming customer service and the contact center landscape. Thank you all for tuning in to “CX Decoded.” Stay tuned for more discussions on the cutting edge of customer service. Until next time, I’m Dom Nicastro, signing off.

We encourage you to drop us a line at [email protected]. And if you have comments or suggestions for a future podcast, we'd love to hear from you. Additionally, if you like what you hear, please post a review on Apple podcasts or wherever you may be listening to this. And be sure to share CX Decoded and anyone that you think might benefit from these types of conversations. Finally, be sure to follow us on Twitter @CMSWire. Thank you again for joining us, and we'll see you next time.