Connecting the dots: How can you best use the data you already have?

This is the TTEC Verint third episode of the three set webinar series that we have produced. If you've joined the first two, as you know, our first one, we, reviewed and discussed FedRAMP, stateRAMP, and how, the move to the cloud and the pieces around it. The second webinar, I joined as well and participated, and we discussed AI and CX, automation and how that is helping contact centers move into the future. Today, we will continue down the path of this, series and discuss connecting the dots, how you can best use the data you already have. I, myself, Scott Montgomery, is Verint's vice president of public sector. Been in this industry for over twenty years working in both public and private, focusing right now on all business related to North America, so US and Canada, and public sector. I will be joined today by Ken and Marcy, who will, introduce themselves here shortly. But before we go forward, I'd like to touch base real quick on, data itself to tee this off. So starting off, we talk about data access and utilization. And, again, as you listen to the last webinar, I spent a few moments talking about structured and unstructured data and why it's so critical. And and, again, when we start looking at data and we start thinking back in time, we recall that, you know, several years ago, pre COVID, etcetera, majority of our interactions, especially from an agency point of view, was voice or face to face. Yes. We were bleeding into self serve. We were talking through, email and chat, but majority of it was voice voice and face to face. Since then, the evolution has absolutely, moved forward quickly. We are all over in regards to not only, working with citizens via voice still, but we're obviously engaged via email, chat. We do, also IVR containment where we're having that communication with our citizens. We, still obviously have self serve as we mentioned earlier, which could be across social media as well. And all of this different unstructured data, we're looking at pulling together and making use of that. And that's where your first thing to think about, right, is using that data. The second thing I wanna quickly mention again, as we discussed in the last session, is is around why is that data critical. We talk about looking at that data because it has wealth of information, everything from how we're interacting, the good, the bad, you know, where we're being very successful, and where we can improve. That allows us to take that data and improve by it. You know, we could we could start thinking about, you know, the data is one of the key infrastructure points, right, or key elements, I should say, in making sure any AI or process automation is successful in organization. You know, why is that data critical? Again, when we think about success, we talk about, is it something we can measure? You know, is it simply as, hey. We've achieved the project under budget. Right? Again, we're looking at data to measure. You know, did we finish things on time? Yes. Again, data to measure. But now when we start talking about how can I apply AI and process automation into an organization, it's first understanding and analyzing that data to understand areas where you can improve? And then, again, once you apply that AI or process automation into or the CX automation into it, it's relooking at that data to see how you've actually changed or improved it. So, again, that's the start of this. What we're gonna do is hand it off to Ken next, let him start diving through the additional data, and move from there. Thanks a lot, Scott. Really appreciate the discussion on, citizen engagement. My name is Ken Kralik. As, Scott mentioned, I'm with Verint. I'm one of our directors in our public sector team. And, e o fourteen, zero fifty eight is the executive order for transforming federal government experience. As, we all know, it's relatively recent. It came out in twenty twenty one. And it's the attempt to really improve the customer experience by incorporating the voice of the customer and human centered design. So we can anchor a lot of our discussion on this executive order and a lot of the future projects that we're working with, various clients, come back to this. This understanding of behavioral science and the incorporation into our technology is really a fundamental way for us to, you know, improve that customer experience. I'll hand it over to Marcy now. Thank you so much, Ken, and thanks to all of you for joining us at today's webinar. My name is Marcy Reardon. I lead TTEC Digital's global analytics and insights practice. And, you know, thinking back to that executive order, when I read that, there's one part of it that to me was really the most relevant to this conversation. And it said systematically identifying and resolving the root causes of customer experience challenges. So when trying to do that, and that is a tall order, there's something important to keep in mind. And that is that leaders in customer experience and customer experience analytics specifically, they understand that analytics is is about more than data. It's about more than the technology, more than the data science. It's even more than AI, which is such a buzzword today. So we share this quote. And this quote, I know for me personally, is very much aligned with my philosophy about data and analytics and how to get the most out of it. So this is a quote from Brandon Purcell. He works at Forrester. He has published now for the last several years, a paper on it's a wave report on customer analytics service providers and who are the top in the industry and what are the trends right now in that space. And so what he says and, you know, we really couldn't agree more is customer analytics is about much more than models. In fact, the ability to create models, you know, that define customer analytics, it's really largely commoditized at this point. And so the organizations and businesses who really differentiate themselves in this market are those who invest in the capabilities to operationalize models and also to be very industry and domain specific while continually innovating. So think about that. It's about taking the data and being able to, as it says right here, the key to success is to embed the insights from your data and your models into an employee or a customer experience in a way that changes behavior. So let's talk now about that gap between collecting customer data, collecting feedback, and then really being able to leverage it to impact experiences. So, you know, the data exists, it needs to be integrated. It not only then needs to be integrated, it needs to be visualized in a way that's easily understandable. It needs to be analyzed, and then it needs to drive outcomes based on the insights. So what we recommend is starting be be grounded with something that's focused. Either focus on your key objectives or focus on your constituents' pain points, and try to develop some hypothesis driven use cases. Think about these use cases as typically centering around, you know, probably one of three KPIs. One is about efficiencies. So that can be like contact center efficiencies. So those are things like, how can I do more self-service? How can I deflect to lower cost channels? Or how can I improve my knowledge base to reduce handle time? That kind of thing. The next is focusing very directly on the experience. So you may have metrics around CSAT or NPS or something like that, but that's really about, you know, understanding and finding the friction points and then, identifying ways that those friction points can be improved either through interactions, you know, channel interactions, human interactions, or identifying upstream processes such as maybe there's a glitch on the website that needs to be fixed and that's causing a lot of customer frustration. And then the last is thinking about revenue and revenue increases, looking for, opportunities to, you know, retain relationships and extend relationships. So the point here is about understand the objectives, be able to scrutinize the data in a way that uncovers the insights. Great. Thanks, Marcy. I'll take, I'll take this next, part of the discussion around, drilling down into what is the data that we're talking about. All the integration that Marcy talked about, all the engagement, data and AI that Scott talked about, We gotta talk about where that data comes from. And in order to get more out of it, we need to make sure we're looking at all of the sources that we have access to. And we're not talking about, just web data. We're talking about experience and engagement data. We're talking about the data that's coming from all the channels, that citizens and other consumers, other customers, use to communicate with any government agency. So it could be, obviously, mobile phone with a mobile app, or or voice, could be email, could be a self-service portal, could be, via chat, social media, all the age all the, channels that we have listed across the top. Then across the bottom here on this slide, you'll see we note the digital channel is a big source, obviously, website, the contact center as well. So any of the, any of the discussions and calls and recordings, and surveys that are done through the contact center, provide that voice of the customer data. And then, then there's, that whole repository together that can bring not only the external web data and analytics, but also the internal organizational voice of the customer experience data that we're just talking about, bring that all together, and even combine it with other internal sources of data like your CRM data, bringing all of those data sources together, connecting the dots, our theme for today's webinar, really, is is the opportunity to map out that journey that that, that Marcy talked about, find those friction points, find out where the satisfaction is happening, where it's not happening, take a look at, where, tasks are not being completed, unsatisfactorily completed, and how the interaction was. How did it go sentiment wise? So deriving sentiment from the interaction is another important, insight we can get from combining all these data sources together just based on time spent and even looking at things like replays and clicks and, where, where there was frustration or, if there was a a preference that, you can also see the preferences, the channel preferences that customers and citizens are, are using. So that whole customer journey and that visibility, bringing all this data together brings us insights to the surface that we can act on the way Marcy had discussed, taking them back into the organization, help improve things like the web experience, help contain the, the interaction and answer questions more quickly before customers or citizens are moved over to the call center and and, spend more time and more cost both on on their side and on the provider side. So that's a quick walk through of the the the universe of data that we're, talking about in this discussion. I'll pass it back to Marcy, to talk more about the value curve. Thanks, Ken. And, you know, I do have one comment about these different sources of data. Sometimes if you're early in this process, sometimes it feels a bit overwhelming to, you know, think about all these data sources and how to stitch them together. Some have common match keys. Others, it's more challenging. Like, if you have an unidentified person who's, you know, doing research on the web and how do I really know that that's a customer or constituent of mine and, you know, there are techniques to do that. But what I would say is, the more data you can connect and the more holistic you can be with your view of the experience, it's typically the more is better. But you can also start with, maybe a few of these. And really, I still say go back to what your business objectives are, maybe what's more achievable in the short term, and start there and and then grow. It doesn't have to be all at once. So yeah. So I'll move on to the next slide, which is about the analytics value curve. And this is just a statement. This is kind of a widely referenced, you know, depiction of the full breadth of analytics and everything you can do. In some ways, it may seem that the left side of this curve is the less sophisticated, and then you move up in sophistication to the right side. While in some ways that that's true, the first piece of it talks about business intelligence and reporting and looking back historically on what's happening, in your organization. And, sometimes that can actually be quite tricky and quite challenging. So it's not to minimize even getting reporting right and looking at the right data and presenting it in a way that's super relevant to your business and clean and easy to understand. Sometimes it's a really, really important area to focus. But, you know, we start with kind of looking back historically, doing that descriptive analytics, what's happened in the past, and then moving over to kind of the why did it happen. And that's the, diagnostic analytics. And there are offerings around, let's say, conversation intelligence, which is really exploring your interactions or journey analytics and things like that. Moving from the what happened to the why it happened is is a really critical leap. From there, you have your predictive analytics, which again, it's all of this is kind of working with the same data. But the predictive analytics now says, I have enough data and enough experience with my customers to predict either what they're going to need in the future or what decisions they will make in the future and use that in a whole host of ways to be more targeted, to be more relevant, to be more efficient in your programs. There's basically limited use to, the way that you can apply predictive analytics. Predictive analytics does actually go to what I was saying a few moments ago about models. Right? So there are right now, there are many predictive models available. Some are out of the box and ready to go. But just having a model doesn't mean that, it will derive business impact. And that's where it goes to the very last step of the value chain, which is the prescriptive analytics. Now, how do I create strategies that will embed my learnings from the diagnostic, my predictive models, and actually create, you know, customer facing strategies that take all of that insight about the customer and what their needs and behaviors may be in the future and create optimized, programs that are all around kind of how do either if an interaction comes in, how do I best handle that? Where should that interaction go? Or things like, next best action, which actually anticipates what will this customer need and what type of communication or offer, should I present to them in the moment. So Ken, let me pass it back over to you for the next slide. Yeah. That's perfect, Marcy. Yeah. Let me build on that a little bit because I think, your point was really great about where to start. And, back to, again, our theme of connecting the dots and the data you have. So start with the data you have, the data you have access to, whether it's structured data or unstructured data or both, even better. However, any step that you can take to begin moving up that value curve, will start with listening. And when I say listening, it's kind of like listening to the experience, recording experience, capturing the experience, and bringing bringing, those experiences together and running them through one of the models that Marcy talked about, whether it's just the analytics at the earlier stages, whether it's through the diagnostic, diagnostics that come later further, the predictive ones, or even the prescriptive ones, whenever you can grab a couple of those sources of data, run them through your model, you're likely to, come up with two kinds of actions that come out of it to help you, again, improve the value that you're delivering to, to, to your end, end user and citizen. That that would be some short term automation, opportunities, and then also some longer term more strategic ones. So we kinda look at it in these two phases depending on what we can do on a short term, simple things like alerts. Alerts can be generated for action either into a case management system or into even an email queue, but, some some way to manage, some alerts and act on them immediately. And then longer term, you can look at things like where there are those friction points in the customer journey, where the, fail failures are happening in task, abandonment abandonment of trying to complete a task on a website, moving to the call center, making that connection, seeing where that extra cost and frustration is happening, and weaving that into your strategic plan for doing business process changes, for doing other system modifications. So we're looking at it kind of these two ways to, in the end, improve that experience, improve that engagement, have that, more conversational experience with your, end end user, and, hopefully, improving, improving, all the financial results and business outcomes that you're looking for. Marcy, anything to add on the, on the on the on the drivers there? Short term drivers? I think you covered it, Pam. Great job. I can actually move on to another point. We feel it's important to share this because there's a pretty critical and often overlooked aspect of, you know, effective data management and analytics, and that is data governance. You know, we often find that our clients, when they embark on a new project, it could be a reporting project, it could be a more advanced analytics project, oftentimes, the project will stall because there's not a data governance discipline in place. So let me give you an example. You know, there's a need for new and improved reporting. So a new set of reports gets developed. And that now all of a sudden the metrics in the new report are totally inconsistent with the metrics that had been reported before, or one report is inconsistent with the other. I mean, we hear this all the time. Or maybe, you know, the source data is changing and that requires a new process. And now there's, you know, the reporting is broken or the predictive model isn't working anymore. I mean, there's lots and lots of, kind of noise in data that really has to be addressed. Because if it isn't addressed, the entire analysis can fall into question. So, you know, a few thoughts about a data governance framework. It solves these issues. It helps with an understanding of the data domain and this coordination between the business needs and kind of what's going on on the data engineering side. So more specifically, we have, we recommend doing things like, cataloging your data. So that's all about where's the data, who owns it, how's it being used, how secure is it, and using this framework to organize those activity so that they're manageable and consumable. There's also an understanding of the data assets and the relationships to, to each other so that redundancies are eliminated. And we find that often, you know, there's multiple sources for data that, you know, kind of means the same thing. And that ends up in like duplication of business terms, confusion, etcetera. So, you know, that's another thing to really look out for. And then finally, you know, understanding data lineage. So this is about the data life cycle and how data moves through different environments in an organization. So, you know, questions about where a particular metric comes from, how is it curated for analysis, all of that, that's addressed through, data lineage. And I just have one more slide on data governance because it is so important. And this just talks about kind of data governance and practice. And here's what we recommend. Okay. So, take inventory of your assets, understand how they're related to each other, understand your data lineage, how the data flows through each of the different data environments, and there will be, you know, more than one. The next one is, having very well defined business terms so that, you know, users understand what's being measured, how it's measured. We often help with the development of data dictionaries, and those become so critical as there's questions, you know, when, either an analysis is interpreted or a report is reviewed, to create, like, kind of one golden standard of truth, that ensures consistency across the business. The next is around having really clean data quality so that the users feel, you know, confident that reports and analysis are reliable. And then, you know, don't forget about protecting your data. And there there's a concept here around, least privileged. And what that really is, is it actually comes from a computer science concept. And it's a process that allows users access only to the information and resources that are necessary for their specific business purpose. So I know I don't have to overemphasize, you know, data security and protection of data. It's, you know, it's more and more, critical to address in a very proactive way. And then, you know, finally, just have your subject matter experts, you know, think about people in the business being part of a data governance framework. It's really important we in an ideal state, there's a data governance team that's addressing all of these issues and just, you know, really ensuring that, we're we're focusing on every one of these, elements of the data governance to make the the output and the actionability of everything that we're talking about, you know, to to ensure that it does have the right value for the business. K. Fantastic. Fantastic. Marcy, just to pick up on the security, point you made in the middle there and take that a little further. For implementations that require a FedRAMP environment, Verint and TTEC have been working together now for years, to bring up, our voice of the customer solutions in a FedRAMP, certified environment. Verint, uses TTEC's Humanify Enterprise government portal, which, which, has, FedRAMP, moderate authorization as of August of twenty twenty one. So together with Verint and TTEC's technology, we can provide a a secure solution for those requiring, FedRAMP, FedRAMP environment. We usually do the contracting directly through Verint for, for our applications that are in the TTEC environment. Wanted to just bring that up, in case there are questions on that. We also have the ability to to go to, impact level four, I o four, if needed as well. So, that's if that's a discussion you'd like to have, please, let us know. You know, let's let's move away from the theory a little bit and tell tell you about some case studies. I have a case study that's actually not in public sector, but we thought we would share at least one because, you know, all of the best practices in customer experience are so, so relevant going all the way back to the executive order mandate we talked about at the very beginning of the webinar. So I'm gonna share a quick story of one of our clients and some of the successes that we, helped deliver, and then I'll hand it back to Ken for, a couple of examples in the public sector. So in this one, this is a payroll services provider, and the situation was as follows, big budget cuts, big goals to eliminate cost from the contact center. But, of course, when you do that, you don't wanna put the customer experience at risk. So they were really looking for, basically, all of the stuff that we've been talking about, looking for the data driven insights and opportunities to drive out the costs. What they had was lots and lots of disposition codes, that they had been collecting from their customer service agents, and they were a mess. There were literally, I think, something like three hundred different disposition codes. And those disposition codes were the only way that they could summarize why people were calling and where the friction points were. So we took the data, and rather than rely on the agent disposition codes, we ran it through conversation intelligence, which is taking, the transcribed recordings and then bucketing them into topics. And this is pretty granular topics, you know, what's going on in topics, complexity of those topics, and the caller sentiment of those topics. And we were able to identify some low hanging fruit almost immediately for there were really two angles. One was better training of the agents. And because we could see this information at the individual agent, we could do some very targeted training, but also identifying upstream processes that needed to be fixed. So this was a payroll provider, mostly business to business, And lots of people were calling because they were getting stuck at a point in the process where when they added a new employee to the payroll, they were being asked for the country code. And there was no sort of cheat sheet, available to tell them what the appropriate country code was. They would know what it was for the countries that they use often. But, if they didn't, they were actually calling in manually to get that number. So we were seeing a lot of issues coming through around that process, and that was one of those low hanging fruit I talked about. So they took it back to their, app development team, and they put a little widget in there that would allow that self-service. And that was one of several examples that drew many, many calls, away from the contact center. So you can see some of the results here. In year one, we drove almost a six million dollar five point seven million dollars The beauty of it was, that and other programs like that, coupled with a better agent training program, actually resulted in a fifteen point improvement in the NPS. So this is one of our absolute favorite, examples because it shows that you can in fact drive cost out while also creating a better customer experience. So with that, I'll hand it back over to you, Ken. Great. Thanks a lot, Marcy. Yeah. Great example. Nice to see the ROI on that one as well. Right? I mean, that's, that's I didn't even mention that. Yeah. Yeah. Yeah. Fantastic. So the next example I've got is, a little bit of a different, angle on it. It is with the, office office of personnel management. So, we don't have, as maybe complete of a use case as Marcy just pointed out, but I think it's still very, very powerful. So take a look at this one. We've got a little bit of the dashboard showing up here, and you can see where we, again, back to that executive order. We're taking the voice of the employee into the business process of of the employee experience. And with surveys of over two million federal, federal, employees every year. We are able to map out what that journey looks like, what the trending is in satisfaction, where the sticking points are. You can see, the various dashboards on the right hand side there going through the, the various, indices that are derived from the data that we have. And then we take those analytics, and we drive them back into the retention process and the recruitment process. So from there, that's been very powerful for the, for the HR agency and policy manager, team over at, over at OPM. So that's one quick example we wanted to share, a little bit of the the the dashboard and tooling that it takes for them to, to, get more out of it insights out of the data that they have. The other area I want to mention is a little more of a public domain where you can easily go to it, Amtrak dot com. If you take a look at the, the chatbot that's there, we call it an, intelligent virtual assistant, IVA. Amtrak has named their IVA Julie, and Julie enables automated booking and answers tons of questions. What's great about Julie is it's not just working off of one corpus of information. It's actually looking at the experiences that customers are having to improve the responses as well as pulling in the latest and greatest from their knowledge management system, integrating that, again, connecting the dots, connecting the data that they have available to answer more relevant questions and incorporate new published, new new published, articles and blogs and, posts that will provide better answers to that Julie can provide in the discussion. So the the business case and the ROI behind this one, is, again, improving customer satisfaction and reducing calls to the call center. So huge ROI on that one. Anytime you, you feel like taking a look at it, head over to am Amtrak dot com and, have a little chat with Julie, and you'll see what we're talking about. Yeah. I would say, Ken, we get a lot of questions. Oh, I was just gonna make a quick comment about that piece, which is we get a lot of questions once those types of IVAs are implemented, how they're performing. And so it's just another thing to think about. I mean, really from an analytics, the book ends around doing the analytics first to understand those use cases of, you know, where to where the IVA, is going to be best applied and how to develop all of the, you know, all of the chat responses and all of that. But then importantly, understanding where the containment is within that channel, what the customer experience is within that channel, why perhaps people are going out of that channel to other places. So just think about it as a continuum and, you know, back to all of the data. And, you know, it's it it is nice, again, to connect the dots across the journey through some of these cool solutions and just think of it as constant improvement. Fantastic. Great great commentary. No. No. That's fantastic, Marcy. I appreciate it. Hey. So on behalf of Scott Montgomery, myself and Marcy, we wanna thank you all for joining us for this webinar. If you had a chance to already attend the prior two, installments in the series, thank you very much. If you haven't, feel free to scan in one of these QR codes here and head over to see the initial ones, which were on the road to FedRAMP and SATERAM. The second one was around driving public sector efficiency. So we'd love to have you, join us, on those other webinars as well on the replays of those. Thanks again for joining us here on this one. I'd like to just again, on behalf of the team, all three of us, thank you very much. Feel free to reach out to us if you'd like to further the discussion or have some questions or some follow ups. We'd love to engage with you, further on the discussion.