Podcast: Data analytics, automation help FIs better understand customer sentiment

Whitney McDonald 0:03
Hello and welcome to The Buzz, a bank automation news podcast. My name is Whitney McDonald and I’m the editor of bank automation News. Today is September 26 2023. Joining me to discuss data collection to make performance driven decisions is Dmitry Binkevich of Qualtrics. Throughout his career, Dimitri spent time at banks, including Citi, and Barclays focused on business development and strategy. Please join me in welcoming Dimitri.Dmitry Binkevich 0:29
Very nice to be here Dmitry Binkevich, I lead the financial services industry practice here at Qualtrics. Globally, have been with the company for coming up on two years. Prior to that, I spent my entire career close to 20 years in financial services in a variety of roles both within financial services players, like Barclays and city and insurance companies, as well as as an advisor, as a consultant as an investment banker, serving the industry. So my approach to the industry as well as to experience management overall, generally begins with the business problems, right? What business problems can we help our customers solve? And at the end of the day, how can we make them either make more money, or spend less money. So generally, our objective is help our customers move their financial and operational outcomes, using experience and everything around it as a lever. Right, which is, which is a nice segue into into Qualtrics. And in general, the Qualtrics position in the financial services industry, our goal at Qualtrics is to make experience a little bit more human to make business a little bit more human. And so that’s what we help companies do. We help companies solicit feedback, which is, you know, your typical survey, right? When you think experience management, probably surveys, the first thing that comes to mind. But then we also help companies ensure that they’re listening into the conversations that are happening with and about the company, right? So whether it’s a phone call, or an email, or a chat or social media, right, there’s a variety of sources that customers can try to can can use to try to connect with a company or just opine about the company something like 85 to 90% of all customer feedback, data is what we call in, in the lingo unstructured, right. So it is not a survey data set, it is just a customer talking or posting or whatnot. And if the if our clients, the financial services, businesses are not listening to that, then they’re missing kind of, you know, nine tenths of all of the possible information. So Qualtrics serves the financial services industry top to bottom right, we cover all of the verticals, we serve over 1300 financial services clients, with, you know, probably 90 out of the top 100 financial institutions globally, right. So very, very rich data set, very rich client list, and they partner with us, because at the end of the day, we help them deliver business results, right? It definitely begins with customer satisfaction, right? But then we can help them deliver better business results, right customer satisfaction tends to result for example, in lower churn, higher cross sell higher revenue, other parts of our platform can help our customers lower cost, right lower cost of serve, whether it’s you know, increasing the amount of interactions that take place via, for example, an automated chatbot or reducing the number of calls into the contact center where a customer is actually doing things by themselves on digital. And finally, we help our customers manage their regulatory risk visa vie sort of complaints, obligations that are prevalent for banks in virtually every jurisdiction that we serve. In the US, you’ve got the CFPB as an example, in other countries, you’ve got the central banks, or security regulators. So we help our customers understand manage, and action, all of that. But our engagement and we’ll talk about you know, sort of the technology and the software later on in this conversation. We kind of it is our thesis that in order for our software to bring value, you don’t just need just sort of listen and under listen, understand, you need to act. And so when we work with customers, we partner with them to make sure that the entire organization is aligned on the value of what they’re doing that it is not just, you know, a CX team, out there in the corner, kind of doing their thing, you really do need the buy in of the entire organization in order to get somebody to do something differently. Our goal is to use the information and the insight that our platform brings to get our customers to do something differently to positively impact their business. So that’s a little bit about, you know, what we do and how we think about working with clients in the financial services industry. You

Whitney McDonald 5:44
know, I know you talked a little bit, you started getting into a little bit about the quantity of data that financial institutions have you talked through the ability to have that insight into that unstructured data in order to make those business decisions. Maybe you can set the scene here a little bit further about really the need for automating that approach to data to both increase the operation or enhance the operation side, improve the customer satisfaction ratings. Can you maybe talk us through a little bit about how Qualtrics plays a role in automating that data and the importance of having that type of solution to get into all of this robust information that fit is half?

Dmitry Binkevich 6:27
Yeah, absolutely. I mean, I think in order to do that, though, let’s think a little bit about how financial services experience and let’s begin on the consumer side, because that’s the easiest way to sort of frame it, how the Financial Services experience has changed. Right? If you go back, I mean, at this point, like 30, you know, even 30 years ago, right? Most of the Financial Services experiences that you had were in person, right? You went to a bank branch? Yes, you interacted with an ATM, but that’s a pretty, you know, inanimate object. But you talk to a teller, you talk to an insurance agent, right? If you needed something, you fax things, and you called right, so they were very big, they were person to person experiences, for the most part. If we fast forward to now, a lot of the experiences that we’ve got our, you know, person to person still exists. But I would venture to say that the majority of experiences in retail financial services are what I would call person to machine. Right. And so the person goes on the website, the majority of the transaction happens on the website. And so these journeys have fundamentally evolved and changed. And so has the expectations. So have the expectations of the consumers, right? consumer expectations are framed, but what by what they experienced in other sections of their lives, right. It’s the Google’s the Facebook’s, the Amazons, the Twitter’s, which is very personalized experiences, right? experiences that are not just I mean, it’s not even just personally, it’s almost like no me experience their predictive experiences, they know what I want, before I sort of realized that I want it, right. The gratification is instant, right? Because you know, you get the news, you click a button, et cetera. And it’s sort of very, very precise. And so for the financial institutions, to be able to deliver an experience like that, you really need a deep, deep understanding of your consumer desires, preferences, you know, thoughts and opinions. And in order to do that, you actually need a platform that listens in appropriate ways in every single interaction, where there’s person to person, person to machine and any kind of way, and not only listens, but sort of ties it all together, because the consumer thinks they’re interacting with the bank, while they might be interacting in reality, with a bank onboarding department, with the application department, with the service department, and then with the fulfillment department. But in order to succeed, and I would venture that every one of our clients is in the experience business, even though they think they are in banking, insurance and wealth management businesses. Right, in order to deliver those experiences, they need to understand consumer journeys, they may need to line up the listening posts in an appropriate way. For some it might be a survey, right? There’s always a place for solicited feedback. But if I just spent an hour on the phone, as an example, explaining in painful detail to my insurance company, what exactly happened in my accident? If somebody sends me a survey and says, Hey, how did that go? I’d be just like, well, I just spent an hour telling you exactly how that went. So please go ahead and listen to that. Right? Or if I’m on the website, and I’m frustrated, right. I sort of expect the company to be able to say, hey, looks like you’re frustrated. Maybe we didn’t do a great job, you know, building this page, how can we help do. And so what the Qualtrics platform does, it allows our clients to position listening posts along key journey nodes in the mode that is most appropriate for that journey for that node. And for that customer, structured, unstructured, Inferred behavioral, right. So everything from survey to call analytics to click analytics, right to session recording. And so, and on the back end, we ended up pulling all of that together and helping customers, our customers make sense of it. Because the important thing and experience management is not just the what, which is what I just described, it’s also the so what, right, as a, as a manager, as a leader in a financial services organization, if I’m just looking at like information or data, it’s overwhelming, right? What I really need is a needle in the haystack, so that I can figure out where to spend my limited resources to make sure that the results that I care about are moved. And that’s where the sort of the omni channel platform with a single back end, like Qualtrics, irrespective of sources really comes into its own.

Whitney McDonald 11:18
Now getting into the how I know that you said you’re linking into these different areas of the bank and making sure that you’re you have that tech in place, what does that look like? How do you really get into the the nitty gritty of the data on a tech on the tech side?

Dmitry Binkevich 11:37
Well, I mean, if you think about Reg, in any, if we take a typical bank, right, there’s a marketing tech stack, and like a marketing team, there’s an onboarding system and an onboarding team service system and a service team. And very often, these systems actually don’t Doctor each other, right? Banks are, and I’m going to use bank so as the most obvious example, but this applies to insurance and wealth managers and other customers that we serve. But companies typically don’t have these talking to each other very effectively. And so when we get into journey design, like you said, we really needed to figure out a way how do we plug into every single text, I can actually bring these things together. So Qualtrics is a SASS platform, right, from a technology perspective. And so the way we link into every single tech stack is via API’s in general, right, so the integration is generally quite easy. And we’ve got a series of over 150 pre built integrations with the most commonly used systems, you know, like a sales force or a dynamics on the CRM side, you know, Pegasystems, for example, you know, for actioning, you know, workday, for example, for ServiceNow, right for human resources, and ticketing. So, we’ve thought long and hard about how to make it as seamless as possible for Qualtrics, to be able to link into each individual ecosystem, not just to pull the data out right to be able to synthesize it, because we actually need the operational datasets to be able to contextualize the experiences, but also in order to help actioning. Right, if you think about it, not everybody at the enterprise needs Qualtrics on their desktop, right? The managers do, the leaders do. But if somebody’s working, for example, in Salesforce, and sort of, or in ServiceNow, in sort of processing tickets, we can ping our, we can trigger an alert or a ticket, for example, into ServiceNow, or Salesforce. So there’s no swivel chair for the frontline employees, right? They sit in the system that they’re in, they sort of are told what to do they go do it, they close out the ticket that goes back into the Qualtrics ecosystem for analysis. For management for leaders, we’ve got role based dashboards, right with the views that are specific to those roles and focused on the sowhat. Right, that, that those people need. But in general, we integrate via API’s. We have a deep, deep pre built set of integrations. And we’re always building more because we know that the ease of integration is one of the key hoops that we have to jump through if we’re gonna get our platform, you know, into our clients. tech stack.

Whitney McDonald 14:36
Yeah, thanks for talking through that integration. That’s really helpful. Another piece of the puzzle that you mentioned was the ability to predict right so you talked through Of course I’m I’m frustrated Didn’t you see throughout that transaction that I was frustrated? So talking through those predictive and analytics and I mean when you’re talking through anything, but especially bank to technology right now, you can’t really ignore AI. Where does artificial intelligence come in? Maybe you could talk to me through or talk through your use of AI here to benefit those financial institutions really get those predictive analytics into play?

Dmitry Binkevich 15:15
Sure, absolutely. The great thing is, is that Qualtrics has been on the AI or the machine learning bandwagon, you know, for the better part of the last decade and a half. Right? So many of our analytic capabilities have been enabled by AI, one of the, you know, specific ones, when we analyze unstructured data, for example, it’s a combination of sort of language models, but also AI, especially when it comes to what we call enrichments. Right? So if you think of the way that if we analyze a phone conversation, for example, or a phone conversation transcript, there’s a couple of layers of this analysis. First of all, what is that person actually saying in English? Right? So we have a natural language model that helps us or not an English actually, we’ve got, I think, over 20 languages that we sort of natively, natively ingest, but let’s say the conversation is in English. What is that person saying? in English? Right? What is the meaning of the words, including all of the nuances, right, when somebody says that, you know, the word sick, for example, like something is sick means very different, something very different from you know, I’m feeling sick, right, and you kind of need to catch those nuances. If you’re going to accurately understand what the person is saying, then you need to conceptualize it in context of the business, right. So if the person is going through banking, onboarding, there’s actually a very specific set of terms and banking, onboarding, right, that you need to understand in order to be able to deeply sort of get in order to get deep insight into why they’re having an issue. And finally, and this is really where the a lot of the AI investment comes in. We do emotion, intent and effort enrichments. So from the text, our AI platform is able to understand, how is this person feeling? Right? Are they angry? Are they confused? Right? Are they very happy? Are they very unhappy? Right? There’s a series of there’s a series of emotions that we’re able to ascribe using our AI engine, based on sort of the relative positioning of the words next to each other, and you know, et cetera. How hard was this to a person? Right? Like, as an example, if they say that your website is ugly? It’s definitely not a great statement. But it doesn’t indicate that they’re having a hard time. It just, you know, they find your website, aesthetically unpleasing. Right. And so, and then intent, what is this person trying to do? And when our clients see the output, it’s not just the understanding, right? Just the what, but also the overlay of how is this client feeling? What are they trying to do. And that is enormously helpful in creating the, what I call Nomi experiences. Because if I had an experience where I was really angry, in the contact center on one of the calls, or I typed in a very angry comment into a web survey, the next interaction that I have with this company, especially given the the single back end, what we call the customer ID, or customer directory, where every single experience gets written on to your customer record. So on my record, there would be, you know, what I said, how I felt, and a suggestion about what the person should do what the CSR should do about it, if I call next. So the next time I call, you know, the conversation doesn’t begin with, hey, please tell me your problem. It begins with, I see that you already spoke to us. And we’re very sorry, that we were not able to deliver the experience that you’ve expected, you know, I you know, haven’t evolved my management to be able to help you now, et cetera, et cetera. So which is as you can appreciate, is a world of difference in terms of how I feel about the brand, how likely I am to recommend the brand, how likely I am to buy from them again. Right? So that is just one small example of how we use AI inside of our platform, the other the other thing and I might be jumping ahead. There’s a lot of talk about AI and generative AI specifically to just sort of understand right understand and respond. Which to my earlier comment is really the what Leia, right, like, what is this person saying? How should I respond? The other way that we’re using AI is actually to try to get to the so what? Because in response to sort of this overwhelm of data, right, because every single conversation, every single thought is now sort of being analyzed, we’re investing in a couple of areas that will help the teams do their job better. And that is actually one big theme that we see in our application of AI, we’re not looking to replace teams, right? We’re looking to augment what these teams can do, right? Make them far more productive. So we’re looking to invest in summarization, right. So really be able to whether it’s video feedback, audio feedback, type, feedback, etc. Quick summary of what’s been said, Read the TLDR, so to speak, and tech speak. The second one is interactive analysis. And that is really cool. A lot of our dashboards right now are just like any dashboards, their data and they’re thoughtfully laid out, they will lead one to the conclusion of what’s important, what to do about it, etc. But we’re building capabilities that, and these are going to be released soon, that will enable you to basically type, Hey, what is the key theme in this data, right and have the AI on the background, do the analysis and give you sort of a thought of what you should pay attention to, right? If I care about customer churn, which parts of this data set, should I pay attention to right and have it. So it’s almost like having a very, very, very able assistant, that can help you with a lot of the drudgery. And then finally, semantic search, which is, and this is true for a lot of our research customers, people run project research projects, through the years and over multiple business lines. And often the left hand does not know what the right hand is doing. And so all of a sudden, you’re able to type in like, Hey, have we ever researched the propensity of, I don’t know, auto insurance customers to churn during price rises? And if the answer is yes, you will actually have that. Right. So imagine, like this, like having a magic library? It’s like, it’s almost like Hogwarts, right? Like you type in a query and sort of a magical answer comes out. So those are some of the forward looking AI applications that we’re working through.

Whitney McDonald 22:35
Yeah, that’s really exciting. And thanks so much for sharing what you guys are kind of looking through and having the works there. One thing I wanted to be sure to touch on was Qualtrics. In action, and example of a financial institution that you work with. That’s that’s benefiting from the technology and kind of talk me through where and how that’s all that’s all progressing?

Dmitry Binkevich 23:00
Yeah, no, absolutely. I would love to, I’ll talk you through with your permission. I’ll talk you through a couple one example. And there was a really interesting example of what we call cross exam, which is, you know, Qualtrics, obviously, does the customer experience employee experience, you know, brand experience experience across the entire 360 of the work. And for one of our customers for, for m&t Bank, we deployed both the CX, which is customer experience in E ex employee experience, and as they were going through the integration, so they bought people’s United Bank not that long ago. And bank integrations are fraught, in general, right, because they tend to lead to branch closures, they tend to lead to customer attrition, because it’s very difficult for customers to, you know, change, branches, interface, people, etc. And so what what m&t was able to do is, they were actually able to pull out drivers, I can speak to exactly what the drivers are, because that’s proprietary, but they were able to, to analyze e x and CX information jointly, right, and make sure that and what they found on some level intuitive, but that the satisfaction of the employees and the branches on how the employees felt about their job, their training, their environment, was very much related to how customers felt right about their experience with their new sort of owner with MMT. And so using that insight MMT was able to deliver, you know, targeted training targeted resources on both sides of that equation, right, both the employee side to make sure that they’re trained, enabled, rested, appreciated, etc. And on the customer side of that equation to make sure Have they had the information to make sure they have the extra help to make sure they had sort of an extra reach out to make them feel welcome when they were peoples United customers. So that was an amazing story of helping the bank really go through, I believe it may have been their biggest acquisition up to this point. And then another one we worked with, we worked with nationwide, a Nationwide Insurance Company to, to do sort of analytics of all of their data, including calls and what they were doing, it was super interesting. They were analyzing each call that came into the contact center using the platform that I just described. But not only that, they were actually scoring it on their bespoke rubric, right, they had a quality threshold that they sort of decided that every single interaction with nationwide should be of a certain quality. And so every call was analyzed and scored. Right, and based on the proprietary rubric, and what they did when the calls were not sort of up to par is fascinating. They call it proactive service recovery, they actually call the person back. And they say, Hey, we’re very sorry that you did not get the level of experience that you expect from nationwide, we’re committed to making it better. Let us work with you to make sure that your nationwide experience is outstanding. Right. So really, both from a from an experience perspective, right, you could think of an impact of that on something like an NPS on something like a renewal on something like churn. So those are two two really cool examples. I think that you know of how we work with customers and how we drive value.

Whitney McDonald 27:00
You’ve been listening to the buzz, a bank automation news podcast, please follow us on LinkedIn. And as a reminder, you can rate this podcast on your platform of choice. Thank you for your time, and be sure to visit us at Bank automation news.com For more automation news,