The CX Leader Podcast: "Predicting the Future": Audio automatically transcribed by Sonix

Download the “The CX Leader Podcast: "Predicting the Future" audio file directly.

The CX Leader Podcast: "Predicting the Future": this wav audio file was automatically transcribed by Sonix with the best speech-to-text algorithms. This transcript may contain errors.

Steve:
Wouldn't it be awesome if we had a crystal ball to look into the future and know what we need to do to be successful? Well, it's not magic, but we do have some pretty cool tools to do some, quote, fortune telling.

Troy:
When I talk about predictive analytics, it really is just anything that helps us to utilize current and historical information to estimate a probability that customers will either behave in a certain way or that they're going to hold a specific opinion about your company.

Steve:
Predictive Analysis, as one of the eight Essential themes driving the CX evolution on this episode of The CX Leader Podcast.

Announcer:
The CX Leader Podcast with Steve Walker is produced by Walker, an experience management firm that helps our clients accelerate their XM success. You can find out more at walkerinfo.com.

Steve:
Hello everyone. I'm Steve Walker, host of the C Leader podcast. And thank you for listening. As we like to say on this show, it's never been a better time to be a leader. And we explore the topics and themes to help leaders like you deliver amazing experiences for your customers. This is the third episode of our series "CX Now: Eight Essential Themes Driving CX Evolution," in which we take a close look at today's current topics and trends that CX pros need to embrace to be better leaders and take their programs to the next level. This series has been developed in collaboration with our partners at Qualtrics, which includes feedback from more than 50 CX leaders. So far, we've covered digital, interactive experience and personalization. If you miss those episodes, you can listen to them on our website, cxleaderpodcast.com. In this episode, we're going to focus on predictive analysis using artificial intelligence or AI, natural language processing and other tools to anticipate customers' needs and actions. And there's no better person that I can think of to help us understand this topic than our very own Dr. Troy Powell. He is our vice president for strategy and analytics here at Walker, and he's actually a legend on the podcast. This is his 12th appearance as a guest. He actually guest hosted for us once, and that makes Troy our all time leading guest on the podcast as I'm recording this one, it's episode 244. So Troy, you're like almost 5% of our all of our podcasts. And welcome back to The CX Leader Podcast for this special episode. So glad that you're willing to come on again.

Troy:
Thanks, Steve. It's always my pleasure and I'm always great to have that percentage going down as we continue to get great guests on the show that are talking about their perspectives and their successes and CX. So I'm okay decreasing that.

Steve:
Well, that's kind of you to say you're being modest. I think part of the reason why you've been on the podcast so much is that you are a great guest and that you really have unique perspectives. I've never met anybody that who's in this business that doesn't like a little time spent with Dr. P, So we're grateful that you're our all time leading guests. And on the off chance that somebody doesn't already know a little bit about you, let's let's just for context, give them a little bit of your background and experience and sort of what what you've spent the bulk of your career doing to to be qualified to talk about this today.

Troy:
Sure. Yeah. So I've been in CX for about sixteen years now, all of that at Walker, but getting exposure to all sorts of different contexts and trends, transformations that have gone on in this industry over that time and a predominantly focused in the space of data analysis. And then kind of the CX strategy of saying the core of what CX can do is collect this data that otherwise wouldn't be collected, Getting the perspective of the customer, but just having it isn't enough. How do we actually take that data, make use out of it, and to then drive decisions and actions forward? So that's been where I spend a lot of my time is how do we get the right types of data and then how do we make use of it?

Steve:
Yeah, you know, I should also say that you've been sort of the architect of this series too. I mean, you part of your job is to kind of look in the future and and say, where do we need to go? So maybe again, share a little bit about the the genesis of that and how we came up with this idea and sort of what you've been doing to to make this thing happen.

Troy:
Yeah, you know, we do always try to boil things down to essentials, to frameworks, to what are the kind of core pillars that we should be thinking about and focusing on. And and so we developed these eight things. And, you know, the reality is sometimes we create them real quickly and say, Oh, these eight sound great and move forward. And honestly, it's a little bit of what happened here is we've spent a lot of time in the space. We know some of the big themes, but also the fact that these eight things are very interrelated. And that was a piece we spent a lot of time on. Do we want that? Do we want to try and condense all of this? But end of the day is you know they should be interrelated. Like CX is a very interrelated topic. And so, you know, these things don't stand alone. They're not mutually exclusive. But we feel like they represent the real core of what CX leaders need to be thinking about and addressing and putting plans around because they're not going anywhere. But the way we address them is changing and is always changing. And so what we wanted to do with this is look forward to, what do we think needs to be doing to really address these themes as we move forward over the next six, seven years, moving towards 2030? So these themes are ones that we may end up changing some as they go along, but are the things that we want to keep as consistent focus points over the next 6 to 7 years of time within CX.

Steve:
Yeah, and this is awesome because this is the third in the series of eight, and for the first three at least, we've been able to do them in chronological order. And I already see in the second one I could refer back to some of the stuff we talked about in the first one and we will just continue to build on it. And actually what you're talking about is you're almost trying to create predictive analytics for the future of a CX pro, which I kind of like that, right?

Troy:
So yes, we are trying to do without analytics, which is always dangerous, right? There's a quote that I can't remember exactly, but basically humans are obsessed with predicting the future and we are absolutely horrible at it.

Steve:
Yeah.

Troy:
So, so that's why the value of analytics in that to help make up for the gap that we really think we know what's going to happen. We never do.

Steve:
Yeah, Yeah. Actually that's isn't that like I'm going to butcher this, but isn't life is what happens while you're making your plans.

Troy:
Yeah. Yeah. Similar. Yes, Similar concept. Exactly.

Steve:
All right. Again, you've been part of the architect of this, but, you know, in each one of these episodes, I start with our our expert in this case, you on predictive analysis. And just I'm going to allow you to define it for our purposes today. So, Dr. Troy Powell, how do you define predictive analysis as it applies to the future of customer experience?

Troy:
Yeah, well, and really, of all the episodes I've done this, this is going to be the easiest because it's just in the title. We need to analyze things to be able to predict like, that's it. So I think we're pretty much done. That's the definition. You know what to do, analyze data to predict, and that's all it is. I wish it were that easy. But, you know, predictive analytics is one of those things that's been talked about for a while. I actually look back the Gartner hype cycle they put out every year about these technology topics and where are they at in the cycle of being kind of the next savior of the world to down the trough of disillusionment, you know, and and predictive analytics in one form or another has been on there since the late 2000s, early 2010's, whether that was in data mining or other terminology. But ultimately the idea of, yeah, now it's AI machine learning, prescriptive analytics. So it's always a different terms, but at the root of it, it's kind of been the same thing, which is how do we make use of data which we are just awash in in some way to understand either and this is where it does get confused a lot.

Troy:
Everybody thinks it's just about the future predicting what's going to happen. Well, especially within CX, we also need to think about it as just predicting what we don't know about customers or about their experiences or about their perceptions. So when I talk about predictive analytics, especially in CX, it really is just anything that helps us to utilize current and historical information to estimate a probability that customers will either behave in a certain way or that they're going to hold a specific opinion about your company or have a certain perception. So it's really about using data we have to, in essence, create more data to create more useful data. And so, yes, some of that's future oriented, but not always. And the other element of that is it is it's customer level. We get a lot about forecasting trends, and that's important. But for CX and for predictive analytics within CX, I always want to focus on how does it help us understand a customer better and what they're going to do.

Steve:
Yeah, and you know, I have a saying, you probably heard me say it before, but you know, we dream forward. We plan forward, but we measure our progress backwards. And, you know, you mentioned you've been at this for 16 years. I've been at it a little longer than you, but we have actually seen progress in this. We have seen the evolution of being able to predict things. And, you know, just very simply, like we ask an intent to renew. Right. And if it's a subscription where you know, when the subscription is going to run out, you can actually measure that behavior and relate it back to what they said three or six months ago. So and then you can aggregate that. So that's just kind of a kind of a very simple example of how we can make it predictive. But what are some of the key points for a CX pro to think about as they start talking about being more predictive in their analysis?

Troy:
Yeah, I think it's really important to to not fall into the trap of thinking the magic of predictive analytics or of AI or any of these advanced things that it's kind of a push button magic solution because you've got to look at it as, Oh, throw this at some data and it's going to tell us what to do. That's just never the case. And that's why sometimes it has struggled. You know, it has gone through the the hype cycle of disillusionment because it was put out there, hey, this will solve our data problems. No, it's still hard. You know, I think one piece I really, really always talk about is having the right questions in mind for what you want predictive analytics to do for you. And whenever possible, try to get below the first thing you think about. So if you talk to CX leaders or business leaders, you'd be like, if you could predict one thing about what the customer is going to do or think, what would that be and be like, Well, are they going to leave us? Right? Are they going to churn or are they going to stop buying like that? And obviously, that's an important thing to pick. But. Well, I usually didn't try to say it was okay. What is it that affects that? We generally have an idea of the things that are precursors to churn.

Troy:
We don't have a perfect idea, and that's an analysis. Usually. We usually have some idea of, well, hey, if we see that they are using the product less three months before renewal, that's that's a bad sign. Or there's been a lot of operational analysis usually on things like churn or purchase of a new product. Right. And so really trying to focus in our efforts on some of those more specific aspects can be a better way to get a hold of it, because then especially in business to business, where these things are can be very complex. You can start now looking at, well, one of the precursors is that they have they're disliking the support they're receiving for the product. All right. Well, let's go now and look at well, how can we understand what's annoying people about support? How can we predict when that's happening and intervene versus waiting until we have to do some big save on an account when they're about ready to not renew? So trying to push upstream as much as possible is kind of one piece of it, and it's usually a little bit easier to get the data in those cases. Sometimes, too, when you talk about something really big, there's just a lot of data that's going to come into play with that that is sometimes hard to marshal.

Steve:
So it is if I'm listening to you right, it is sort of iterative. It's cross-functional within the organization. What are some of the other kind of key points you might look at in terms of trying to be more predictive?

Troy:
Yeah, Well, and so the other piece going in maybe a slightly different direction, but it's useful. One thing I found with predictive analytics because it is such a broad topic that there's all sorts of different kind of ways you can talk or think about it. So another way that I will often try and get people and clients to think about is. You know, when we talk about for CX leaders, particularly when we talk about what is it that you want to prove the impact of? It's usually at the top end. We want to show that the emotion, perception, etc., from the survey is able to predict or explain a behavior. And you mentioned it with the intent to renew, right. Or an NPS score or a loyalty score and say, okay, it's important to know this because it actually relates to behavior. And so proving that out is pretty important, and that's kind of a predictive little element to it. The reality is that's valuable Whenever we can do that, there's value. The issue is it's, as we all know, there's a small percentage of people who are going to respond to our surveys. So you're like, Hey, great, for those ten, 20% who responded, what's happening to everybody else? And so thinking about the ability of predictive and analytics to fill in the perception, the emotion, the intent that we get from experience data. And so that goes back to my point. If we think predictive analytics is always the future, but a lot of times it's filling in missing data. So say for the other 90% of people who we don't know what they think about their support experience or we don't know what they think about the product, how can we use data about the ten or 20% we do hear from to sort of project out to, well, okay, where is these groups that we maybe should be contacting or getting in touch with because they might have very negative perceptions about something. So thinking about it that way can be a really big boon for a CX program in a way to provide quite a bit of extra value.

Steve:
Yeah, I like that. So yeah, you got a set of customers where you are able to get them to respond. And you fill in some of the missing data. But now you've got the other 80 or 90% of customers who you didn't talk to. But we can profile and then sort of project those other missing data across them. And then that might inform some of our proactive business activities. Right. That's kind of the idea, right?

Troy:
Yeah. Yeah. And it's an aspect that helps a lot with personalization, which was a previous topic. One of the hardest things to do when you're trying to personalize services is if you don't know what people think about the services. If you're like, Oh, hey, we think they're going to love this because they're young. It was like, Well, not every young person is the same.

Steve:
Yeah.

Troy:
And so having some sense of, well, hey, it looks like, you know, these other factors come into play for people who really want a more digital experience versus a more human interaction or, you know, these pieces of things that can help us to actually do personalization in a more personal way by predicting or understanding that sentiment in a broader, broader way.

Steve:
Hey, my guest on the podcast this week is Dr. Troy Powell. Our our most frequent record setting guest here on the podcast. It's always a pleasure and Troy is enlightening us so and one of the eight essentials of CX Now it's our series on kind of the key trends that CX pros ought to be looking at. And today we're talking about predictive analysis. Yeah, So we we talked a little bit about the churn example. I think there's probably applications in customer support. You know, like you could use certain number of transaction surveys to maybe predict some others who we maybe don't know, but maybe they're having the same problem and then proactively go out to them. What are the other applications? Are there any applications that are kind of like more revenue generating?

Troy:
You know, one that I've seen work before, it's. Pretty interesting is within kind of a cross-sell upsell new product environments where the goal is to get either a new product that's come out and try and get people to utilize that, especially current customers, because you have, you know, and this is a piece that's also critically important for CX to think about the current customers. We have data on them that no competitor can touch. I mean, so that richness of data, if we're not utilizing it, is going to waste. And so with current customers saying either an upsell opportunity or new product opportunity, we I've seen a lot of times where sales ops teams will come up with great profiles of, Hey, here's the type of customer that's most likely to buy this new thing or buy this thing that's been out for a while and we're trying to expand its reach. The one element that can actually be really helpful there is the people who have bought it already, which ones have actually liked it, right? Because sometimes we force new stuff on client and half of them don't really like it. So wouldn't it be better to be able to understand, hey, these are people who are seeing great value in this and are telling us they love it and that it fits really well. Let's layer that in. And now we have the ability to say, okay, not only are these people likely to buy it, this group is likely to actually love it and keep it and grow with it. So that's the little angle, again, of predicting out this kind of sentiment or perception to try and get a better business outcome.

Steve:
Wow. That's that's fascinating. Again, for our CX pros listening to that, just think about that concept of taking that back to your sales leadership. You're really talking about an information advantage. The way you you frame that is like we're the only ones that have the data from our customers about our products and nobody else can access that. So what gold can we mined from that? And you started this whole conversation about collecting data that a lot of people think you can't get to. You know, I mean, it's like, well, we don't have that data. Well, let's you know, I think that's part of the real lesson here is let's let's go get it. You know, something's better. Nothing. And if we iterate on that a few times, we're going to really we are going to find that information advantage.

Troy:
You know, the other piece I think that's interesting is the data integration, which is a topic that we'll be discussing at some point in the series. But, you know, that's obviously the route to so much of this is getting the data together in a place that you can analyze it. But that's often really is a barrier to predictive analytics, but sometimes it's a barrier that we make because it just seems like it's so hard. And and what I've found is sometimes can help there is we don't need all the data. We don't need access to every metric in your service cloud instance necessarily to predict whether somebody is happy with the service of that. Just as an example, if we can get it, fine, great. That's useful. But. You generally, most companies, operational research groups have done a nice job of understanding what the top items are. And then you also we've seen through doing this a few times, you're like, Hey, what's some of the big things that keep coming up and predicting sentiment? So getting a group together to say, okay, well let's try and get these ten metrics or these 20 metrics together. To your point, it doesn't have to be perfect right away. That can be a way to sometimes get around the fact that, hey, nobody's got systems that talk to each other as much as we would like. That's just reality. So that's kind of a way to to think about data integration and not let it stand in the way of trying to get something done.

Steve:
Yeah, Yeah. I mean, it's a little bit of analysis paralysis. You know, you, you sit around and wait too long complaining about what you don't know, instead of really trying to get at some things that could could move the the rock a little bit up the hill. Hey, let's turn it a little more to the practical applications now. And this is kind of a thing I've been using in this series, but OK I'm a CX leader. I've got a pretty decent program, but I've not really gotten into predictive yet. And, you know, you're giving me all these ideas, so where do I start? Kind of what are some of the initial motions or actions I could take to start to get my company, my program on a path to being more predictive?

Troy:
Yeah, You know, I think the place to start, it's similar to if you go way back to our kind of X and O data discussions in the past, you know, how do we get experience data and operational data together and similar fact of talking to stakeholders. So say, all right, really, I think of it as like starting with the decision in mind. So what decisions are people in the organization making relative to how customers are being served that they feel like is something is missing within that decision? So the idea of, hey, we're deciding like our account reps are deciding on a weekly basis what customers who are getting ready for renewal do we need to talk to more? Do we need to that might be at risk? All of these things that's happening already, well, do they feel like that's good enough for them? Like that's a smart enough decision? Or do they feel like that decision could be something that is maybe smarter, could be made more accurate, could have additional data align to it, and it might be that they feel great about it. All right, wonderful. All right. Let's go talk to the next person or what's the other decisions your account teams are making? So you're continually having those discussions of where are people who are being asked to make continual decisions over and over you? Where are they feeling a little bit, you know, what the right word is, but not feeling great about that decision, not feeling as confident as maybe they would like to, and then seeing, hey, is there something that we can do as a CX team to kind of factor in or to help that out, create that data, create predictions, etc.? So I think that's always such an important place to start and something maybe we don't always do enough of.

Steve:
Yeah, I think you've given them, you know, this is a recurring theme on The CX Leader Podcast is you've got to go out and talk to the business leaders, the functional leaders, the BU presidents or whatever and got to speak their language. You have to convert some of our expertise into what they're trying to accomplish. So I think that's great. But I think this line of questioning, where are the gaps in their knowledge? Can we fill it in? Heck, some of it we might already have in our relationship survey that that exec is just never really accessed. You know, it could be already there sitting on the shelf. You know, one of the things that I'm taking from this, Troy, is, you know, in order to be good predictive analysis, CX pros, we've got to keep talking to the stakeholders and we've got to keep filling in the missing data. That's I think that's kind of a theme you here have here is to be predictive. We have to go find data that that maybe doesn't exist yet or maybe somebody doesn't know it exists. So you were taking us down kind of a path of how do you go talk to these users and unearth these gems of missing data that we can fulfill in our organization? And in doing that, we create great value. So talk to me a little bit more about how the pro might action this in terms of how they're reaching out to the other business leaders to find these places where we need data.

Troy:
Yeah, you know, I often talk about a series of questions that you can ask yourself or even better ask your colleagues. And, you know, and it starts with what I was just talking about, which is, you know, what decision do you want to improve? Think about the decisions that you or your people in your group are making on a regular basis that affect the customer. Which ones of those do you want to improve? Then asking, okay, well, how are those decisions currently managed? What systems are processes are in place? Because sometimes the decision you make, there's no real process to manage it, in which case, honestly, creating a better prediction or a better date is not really going to help if there's not a way to improve And the management process, that's the first thing you actually want to do. So that's a good thing to ask and understand. And then diving into, okay, what data is most relevant about that? And again, there's this expert opinion aspect that you can tap into that we often know we don't know everything. That's why analytics exist to find the new things. But we usually have a decent idea of the top items and then, okay, where does that data reside and how or can we get it right? And so those kind of questions help to really bring you down to, okay, this is a Canada this is something that we might be able to help with and fix. So that's a nice kind of cheat sheet, I would say, of how to go about it.

Steve:
That's awesome value. So and you know, it may be your take home value, but I haven't asked you that question that I ask every single guest and I've asked it of you at least 12 times before. But, you know, it's the concept here is we try to leave our listeners with kind of the one thing or, you know, a couple of key thoughts that they should take from your podcast and that they could apply to improve their program, you know, almost immediately. So, Dr. Troy Powell, what is your take home value from this session on predictive analysis?

Troy:
Yeah, I think I thought about it a lot. There's a lot of pieces of this, but to me it always comes back to start with the decision in mind. You know, it's not starting with what you want the prediction to be, it's what decision do you want to improve and make better? And really the value of predictive analytics to me is to make decisions faster, to make them more accurate, or to make them apply to more customers. Like those three objectives about decisions really where I think we need to be focused as the CX team relative to predictive analytics. So if you're so start with that decision. What decision do we want to improve? Where do we want to make it, a faster decision, a more accurate decision, or a decision that we can make for more customers because we have the right information now. So that would be my take home value. Start with a decision in mind and work backwards from there.

Steve:
My guest on the podcast this week has been Troy Powell. He is our vice president of Strategy and Analytics here at Walker, and he's a legend on The CX Leader Podcast. And I think you can tell why from this podcast, many of our listeners have already connected with you. But just in case, why don't you give him your Walker email and or your LinkedIn profile? So in case they would like to make that connection, they can continue the dialog.

Troy:
Yes, you can always reach me here at Walker at tpowell@walkerinfo.com or Troy Powell on LinkedIn. I think it might be Troy A. Powell to be on, but you look it up. Troy Powell working for Walker. I have a PhD. I'll be there.

Steve:
Yeah, yeah. Troy you will always have that. You always have that PhD that I will never have. And thank goodness for you bringing your talents to Walker. And thanks for being a guest on the podcast. If you want to know anything else about this podcast or about how Walker could help your customers experience, feel free to email me at podcast@walkerinfo.com. Remember to give The CX Leader Podcast a rating through your podcast service and give us a review. Your feedback will help us improve the show and deliver the best possible value to you, our listener. Check out our website cxleaderpodcast.com to subscribe to the show and find all our previous episodes, podcast series and contact information. You can let us know how we're doing or drop us a note for an idea on a future podcast. The CX Leader Podcast is a production of Walker. We're an experience management firm that helps companies accelerate their XM success. You can read more about us at Walkerinfo.com. Thank you for listening. And remember, it's a great time to be a CX leader, so keep doing what you're doing. Go out there and make your analyzes more predictive and we will see you again next time on The CX Leader Podcast.

Sonix is the world’s most advanced automated transcription, translation, and subtitling platform. Fast, accurate, and affordable.

Automatically convert your wav files to text (txt file), Microsoft Word (docx file), and SubRip Subtitle (srt file) in minutes.

Sonix has many features that you’d love including automated translation, upload many different filetypes, share transcripts, automatic transcription software, and easily transcribe your Zoom meetings. Try Sonix for free today.