Indico Data releases industry-first large language model benchmark for document understanding tasks
Learn More
  Everest Group IDP
             PEAK Matrix® 2022  
Indico Named as Major Contender and Star Performer in Everest Group's PEAK Matrix® for Intelligent Document Processing (IDP)
Access the Report

Unstructured Unlocked episode 17 – How automation enables insurance companies to increase CX metrics

Watch Christopher M. Wells, Ph. D., Indico VP of Research and Development, and Michelle Gouveia, VP at Sandbox Insurtech Ventures, in episode 17 of Unstructured Unlocked. Listen in to learn more about how automation is enabling insurance companies to increase various CX metrics.

Listen to the full podcast here: Unstructured Unlocked episode 17 – How automation enables insurance companies to increase CX metrics


Christopher Wells: Here we go. Hi, and welcome to another episode of Unstructured Unlocked. I am co-host Chris Wells, joined today by co-host Michelle Covea. How’s your day going?

Michelle Gouveia: Good so far. How about you?

CW: Same. Yeah, just one thing to the next, but never bored. I’m thankful for that. Today we will discuss how automation enables insurance companies to increase customer satisfaction. So things like CSAT and net promoter score. And I have to confess. I don’t, I don’t know a lot about this, so

MG: I will right along with you that my experience in this, this arena from my days in the insurance carriers was, was limited. So I’m sure it’ll be more hypothetical and theoretical today than <laugh> maybe we have in the past.

CW: So, if you’re a listener and you have customer experience experience, in the insurance industry, please let us know. We’d love to have you on the pod. Okay. So you, you have some limited experience. So let’s start there, and then I’ll ask many naive questions, and we’ll see. We’ll see where we get to.

MG: So I witnessed or, or was adjacent to during my time working within the insurance carrier space. This was around, so I’d say around 20, 14, 20, 15, at least at the carriers I was at, that’s where that’s when there became a big push, hmm. To improve the customer experience, right? Whole organizations were being stood up, focused on how to better engage with the customer, how to improve, you know, the net promoter scores or, you know, survey the customers more efficiently or more effectively. Yeah. to then take those insights and, you know, help with claims workflows or help maybe even with product development, depending on what you were looking for. There, there were a lot of different flavors to it.

It was also around when the carrier stood up these. I’ll call them self-service portals. So this is, you know, you can buy your insurance online, and that’s a brand new place to interact with the customer. And there was a lot of emphasis on building an omnichannel experience, right? How does that feel if you’ve reached a certain point here and then you fall out of the system or you have to make a phone call? Like one clean, yeah. Transaction or process that you’re not missing key steps or having to restart. So that’s, from what I, around that time, that’s what I heard a lot about. And so the work being done, and at least initially around the areas where I worked a lot of process engineering about understanding the workflows and where these touchpoints with customers were happening.

And so the customer can be broadly defined, right? The carrier’s customer could be the broker, or the agency they’re working with. In some events, the carrier could be a customer, so that the reinsurer might be doing this. And so there’s a lot of work to be done to understand how and how each of those interactions happens. And that process engineering work was being done to say, where do we interact with the customer today? How what does that interaction look like? What is the information that we’re capturing? What information will we like to capture that we’re not? And, you know, doing that along the whole process, there’s also probably touchpoints that, I guess you always want to talk to your customer whenever you can, but probably some that felt redundant or felt like not helpful and the information wasn’t being captured in a useful way. So, a lot of mapping out where and where those things were happening. So, that’s kind of where I look at it from when I think about this net promoter score customer experience side, is where along those types of efforts would, would something like automation maybe be helpful?

CW: That’s, that’s interesting. I was sort of under the assumption that these things were most important or maybe solely important in the personal lines. Still, it sounds like you’re saying the same sort of thing, just a very different framework for, like, the commercial lines and even reinsurance, it sounds like you’re saying.

MG: Yeah. But even outside of, of, I mean, definitely for personal lines, but, you know, you have to think about customer experience from the and the big obvious one right, is claims handling, right? And you get claims for all lines of business, so you’re inter you’re still interacting with somebody. So what is, what is their experience during the claims process? But there’s also, you know, we, the underwriting process, and what is it like to get all the information in for a submission? Is it easy? Is it a lot of back and forth? Is it super manual? Do you have a lot of pre-fill already enabled? How often do you have to ta talk to that person? So, I think it hits more than just personalized, but for sure, as insurance customers, right? As consumers where, where we’re buying, you and I are buying our insurance. We’re doing it on the personal line side. So for sure, that’s probably how we think most about it.

CW: Yeah, it’s interesting around that same time, sort of that 2015, maybe through 2018 timeframe both in the insurance and asset management space. Like, there were a lot of initiatives going on too, like, let’s build the chatbot for hmm. To interact with. So like, you know, whether it’s someone, you know, someone wants to rebalance their 401k, 401k, like, can we do that without a human involved? Down to buying your insurance, everyone was trying to build a chatbot. And I’m sure that will be revived today with G P T and friends. Hmm. Because I, most of those initiatives from what I had seen, again, sort of third party stalled out cause it’s a, it’s a pretty hard problem to build a chatbot that people want to interact with.

MG: Yeah. I think from, so what, what we’ve seen in, in the space just from, you know, in the, in the VC world is, you know, a lot of this, this, these chatbot technologies and now AI behind the scenes trying to improve it, where, where the end state, right, is ultimately the chatbot can answer the inquiries as well as a human where you don’t have to connect with, with a human, right eventually? Like, you just want everything. You want your questions answered quickly. You just want to know how, how do I do this? Or how do I obtain this information? And that it’s, it’s pretty, pretty smart to do that. But, you know, you can deploy the chatbot technology in many of those places, right? And I’ll talk a little bit about some of the companies we’ve seen generally in that there’s the chatbot that you think about when you go on when you log into a portal or a websit,e and it’s there, and it’s like, Hey, I see you’re searching this, or how can I help you?

Right? And you start interacting, and eventually, you’re like, I need to talk to somebody, right? Like that. Exactly. There’s, there’s, there’s like the, the, the first level of triage or maybe even two or three levels of triage that these chatbots can do, and those are the, the outward facing chatbots, right? In many ways, this is the result of some of this. This process mapping work to say like, where do we interact with a customer? What, what’s easy, you know, quick wins that we can do where we don’t have to have a huge workforce behind the scenes to answer those, those questions that are simple from, from our perspective. But then there’s also what, and I’ll talk a little bit about, you know, we, we’ve mentioned before that there’s, there are two schools of thought on this, but that in some ways, a lot of people think that eventually small commercial insurance will be sold the way personal lines insurance is now, where it’s, it’s mostly through, through online.

 And I think you had an initial swath of companies trying to act as brokers and help small business owners understand what types of insurance they needed. And so behind the scenes there, right? There’s either a chatbot that’s, that’s saying, what kind of business do you have? You know, where are you located? Things like that. And then trying to offer up insurance recommendations. Obviously, with an engine behind the scenes, or a lot of this, you know, again, customer interaction is providing more information. So a lot of those, you know, those question marks in a little circle next to a field on a website that you’re filling in, right? Like some that, that to me is also a mechanism of customer engagement and what’s the experience you’re providing to them, and do they feel like it’s easy to work with you? Do they feel that it’s an educational experience tand hat they’re not getting duped in any way? Right? and so this is pretty broad definition, I guess, of, of where you want your customer experience to to come from. 

CW: Yeah, you touched on something there, which is, you know, if people are confused about how they should answer the questions, part of doing these automation right, is, like, make your process transparent, right? So mm-hmm. <Affirmative> mm-hmm. <Affirmative>, you know, on the that’s on the sort of underwriting side, like, can we get the information cleanly and accurately? And then, on the claim side, part of making customer satisfaction go up on the claim side is just making your process not take a week, right? Make it easier for you and your people to do their jobs.

MG: I think many people would appreciate it if it only took a week. The claims professionals are going to come at me for that. But, you know, <laugh>, it’s complicated, right? Like yeah. To, to, to, when the claim comes in the door, all of the associated workflows that come with that, and then if there’s, you know, multiple parties involved, you know, behind the scenes, everything has to get, get all the information’s gotta come in, it has to be exchanged, it has to be connected. So there are real reasons why things aren’t just like this. But to your point, how transparent can you make that process so that ultimately at the end of the day, the customer feels that, you know, they’ve been heard and seen, by the insurer and that they’re getting the best possible outcome they can.

CW: Yeah, a big part of it is like, don’t, if don’t come back to me with a question about what I’ve submitted as part of the claim, like way down the road, right? Like, you already have all this stuff. Figure out if it’s complete or not right away. Right. I’ve seen that several times. Yeah. That leads to a ton of frustration.

MG: Yeah. And some, and some of it comes, you know, this is getting a little outside of the, the net promoter score, like customer experience conversation, but in, into the claims workflow is because a lot of claims look different. There’s some kind of standard of like what documentation that do you need that, that comes in, but depending on there, there’s no standard for like industries outside of insurance that are working with insurance to, to get you that information, right? So, so you might be working with a law firm that emails you everything, right? You might get a claim directly from the insured as opposed to them going to, you know, to their attorney first, and the attorney is doing something on their behalf. So there, it’s never just a straightforward interaction. And part of the challenge might be even recognizing that you’ve gotten everything, like what, what are, and I not the expert in this, but what are the checks and balances that you have to say, you know, we have received all of misinformation.

And if not like, where is the automation that you can say we, we’ve hit up against our, you know, list of requirements, we’re missing this what, is there an automated message that can go out either to the claim adjuster or if you’re interacting through, you know, an a carrier app or, or a portal, like how does that information get pushed to you so that you can, you can get to that and to your point, not go log in three weeks later and there’s a big, you know, red exclamation point that says, this can’t be done because you’re missing, you’re missing three pieces of information that are key to this.

CW: Yeah, absolutely.

MG: Yeah.

CW: I think, I think there’s an opportunity, well, actually, let me phrase it this way. I’ve seen a lot of processes automated that involve large teams, and I’m sure underwriting and claims are no different from that. And one of the problems I see all the time is that the process depends on the person running the process. Hmm. And so that I don’t, just as a consumer, when I call, you know who, who, who knows Verizon, and I talk to one person, and then I call back later and I talk to someone, and it’s a different experience. Like, that’s extremely frustrating.

MG: Yeah. That, that, that’s an excellent point. And that gets back to, that’s, that’s probably a whole separate episode on just like talent and training. And but it does, it does get to, to the customer experience side of you know, in, in a lot of, in a lot of ways you know, everyone thinks about these carriers as these, these huge, massive companies, but the, the underwriters and the, the claims teams are really the main interactions with, with the end consumer, right? You can, I mean, these, these companies have millions of dollars worth in like marketing spend and, you know, all that’s great, but at the end of the day, you’re calling a number and you’re talking to someone that is representing the brand, but also you know, could be having a bad day. You could be having a bad day, and you’re making a claim. 

CW:  You probably are having a bad day.

CW: Yeah. I’d love to know systematically if like CSAT on the claim side is just lower than CSAT on the underwriting side, just inherently that would be my, it’s my hypothesis. Someone proven and

MG: So anyone, anyone dials in, just let us know. <Laugh> I think, but to, so we’re talking a lot about the individual Yeah. And how, how they connect with, with the, the car or, you know yeah, whoever they’re talking to in the carriers side, but there’s also the carrier and its interactions with its network of agencies and, and the brokers as well, right? So how easy is the carrier to do business with, right. Do, is it it’s that same question of transparency, right? If the broker, the agent sends through a quote, right? How transparent is it to know? Where is that in, in the process with the underwriter? What, how, what is that response time from, from quote to bind or, or quote to decision not, not to bind. What is the communication there? What’s the channel by which that communication is happening, right?

 And so that, that was getting to my point of a lot of this customer experience process engineering work was being done across all these various channels to say, how do we do it today? What’s the, you know, the, the end goal state? And then how, how far are we from, from doing that? And this was the, this was during a time where a lot of these, the, at least the larger carriers you know, were going through this, this huge digital transformation effort and getting new core systems in or new capabilities in. And so how does, how did those things change and open up the conversation for, for better interactions with, with however they were defining the customer at that time? 

CW: Yeah, this is probably naive and silly, but it seems like, it seems like these platforms for insurance especially in the more complicated flavors of insurance could draw inspiration from like the DoorDash app. Like Mm. We’re, we’re cooking it, it’s on its way. There was a problem. You gotta talk to the deliverer, right? Like, that’s a pretty nice user experience. And it, it feels that I mean, at the end of the day, there’s some recipe for the insurance policy that you’re binding, you’re buying and then binding. Yeah. Why not treat it that way? Like logistics? Well,

MG: Well, well, you hit on the DoorDash app example is, is a great example. The, the one that I think you hear a lot, or that I heard a lot around that time was the Amazon experience. Yeah. Right? And that even got into, you know, someone’s buying an insurance policy and it’s, you know, the bottom’s, like other shoppers have also searched for this or have purchased this or like that, that image at the bottom of like, what you’re looking to buy plus two other items, right? And then look at the total package. There was a lot of conversation about how do we try to not try, like how do we show what else? Like who else is like you, right? Who else like you as an insurer that matches, has a similar profile? What are they buying? And then should you consider that too a little, again, a little bit of that trying to be educational and how to help you know, the, the consumer buy the insurance product.

So, but yeah, a lot, a lot of those non-industry examples is what people were referring to when they were trying to say, we, we as an industry want to improve customer experience. Oh, one question I had, I had for you to get your thoughts on was we hear a lot about the challenge of, and we’ve talked about unstructured data and everything else, but the challenge of, of extracting information and insights from what I’ll just broadly call notes, right? Yeah. Or, or text, right? You, you hear about a lot in claims is, oh, a lot of the, the adjuster notes or you know, conversation is logged in, in an open, you know, free text open format which is the, just to your point, pretty much like an email, it’s as unstructured as you can get, right? Yeah.

 and I have to imagine that when you’re ser, depending on how you do it, even when you’re surveying customers, there’s some of that as well, because yes, there’s like the simple, like right, your experience one through 10, but there’s also, I, I’ve been on you know, calls where at the end of it they say, will you take a, a brief survey? And then there’s an option to essentially leave a voicemail or leave a message to give any additional feedback or color. And so just how, how do you capture that information? I think that, like where can automation help? Where there, there, I think technological gaps there. I haven’t heard of anyone able to really do it, but you’re more in the space than I am. So,

CW: So I, I don’t actually think of that as an automation problem. And I’ve, I’ve had this conversation with a number of large companies and different verticals, including insurance, where it’s like, you know, we have all these chat transcripts, we have the emails with the same account, right? We want to know what’s in them and can your, can your technology do that? And my answer, my response is always at least first well, what do you actually want to know? Like, what are the questions you’re trying to answer? And <laugh> the response I often get back is, well, we just wanna know what’s in there. It’s like, well, okay, then you don’t, you don’t have an automation or a technology problem. You have an, you have an analysis problem. Like you need to get all of this data in one place and then be able to slice and dice it, right?

So you know, are you, are you looking for correlations between tone, like sentiment, for example, and outcome? Are you looking for particular demographics or particular you know initial questions that are asked in the chat that lead to certain outcomes? And that’s the kind of thing where there’s no tool that’s going to be able to tell you exactly what those questions are. You have to figure them out. And that that problem, that problem is solvable with, you know, a language modeling mm-hmm. <Affirmative>, you know, plus search technologies plus you know, the ability to write good prompts for large, you know, large language models can be very helpful for this. Of course, I, I don’t get paid to mention them by the way, but <laugh>, they’re worth mentioning this context, but even, you know, even without the fire power of open ai, you, you could build these systems.

The problem is folks don’t want to go to the trouble of building out the system. They just, they actually just want the genie that will come in and answer their questions about all of their data. The, the, the other issue is it’s a massive data problem, right? So to get at some of those answers, you really want to know, like, here was the chat, here’s everything else we know about the account. Here was a call transcript. And by the way, we don’t just want the transcript, we actually want the audio so we can figure out mood and mm, sort of inflection of the, of the words. So

MG: It’s a massive, and that’s all, sorry, sorry cuz you’re talking all about the interactions that are happening within someone sitting behind the desk that’s getting all this too. But, you know, where, where you’ve got someone out in the field Yeah, a field adjuster that’s sent that’s, you know, got their own form, maybe that’s still manual, maybe it’s, you know, I’ll call it a fillable P D F or, or some kind of structured document that is sending back in, but that’s still getting attached to the claim. That’s a whole different party that’s involved. Right? And so back to your, your point I, it kind of goes back probably to like the legacy systems and the, the infrastructure within these systems just wasn’t set up to capture it that way. I, I can’t speak to now kind of what, what they look like. I’m sure it varies by line of business cons and even within the carrier, right? Cuz you’re probably doing some kind of customization of what are the fields that are important to you. Yeah. 

CW: But, and then there’s always gonna be like a blank paragraph at the bottom that you can fill in with whatever, and everyone treats that differently.

MG: Yeah. Yep.

CW: Yeah. I, I think it’s, it’s tough for, it’s tough for insurance because these companies were not, if a, if an e-commerce company decided to become an insurer, I think they would be amazing at this because in e-commerce you have to get this right. And my, you know, my mental model for what makes CSAT go up in insurance is to your point earlier about education, it’s did it, did everything happen fast according to whatever measure of fast I have in my head? And also a little bit of a delayed effect. Did I buy the right thing? Like, did you help me find the right thing? So Amazon saying like, your peers bought this really, really helpful. And so with that e-commerce mindset, those tools already exist. And also the data flows already exist. You know exactly what a customer profile is and how it can be meaningful and you know, how to pipe everything into a common repository so that you can, that you can query it and ask meaningful questions. I just don’t think insurers are set up for that. But again, I’d love to be proved wrong.

MG: Yeah, I I, I would have to agree with you on that. And it’s, and a lot of, a lot of times too there, if you think about it from it, it’s not necessarily the carrier that’s having that interaction either, right? Yeah. So it’s, it’s the, the, the insurance agent that’s, you know, able to sell, but they’re also probably appointed with a number of carriers, right? And so you’ve got them working with you to help you understand what insurance you need, but there’s, there’s also, you know, things going on in the background of who do they have the strongest relationship from an insurance carrier perspective. You know, there’s a question of the commission structure and may maybe, maybe that that particular agent prefers to work with the carrier because they know they will get faster, you know, quote to bind or, or something like, like there’s, there’s a ton of factors that event that you have to pull full on to, to, to get to the heart of what do we really need to change and improve to improve the end, the experience for the end customer and every customer along the way.

So it’s, it’s very <laugh>, I don’t want to call it siloed, but it’s over it, it’s not a straight clear process. There’s just, it’s, you know, there’s a lot of, I’ll call it ch links in the chain of like B2B to B to B2C kind of interactions. <Laugh>, there may be an extra B in there, I don’t know. But to, to your point from a few episodes ago, applying automation to something, to a process that’s right, that’s not, that’s broken or you don’t have clear insights or clear direction of what you’re actually trying to get out of that process it just doesn’t make sense to, to automate it. So I think, I think there’s probably a lot of work that still care people wanna do and that needs to get done to, to apply automation to, to improve customer experience.

CW: Yeah. I, I though I gave you the pessimistic answer, my optimistic version of answering that question is, if I were building this from scratch for an insurer today mm-hmm. <Affirmative>, I would start with what is all of the information? And, and you named a lot of it, what is all the information that I would need to map out? What is a customer and what has that customer’s journey been to date? Mm-Hmm. <affirmative>, that’s where I would start. And then I would figure out like, okay, where’s it all gonna go? That’s, that’s an infrastructure question and very answerable. And I would also figure out of course, where are the gaps? So we see the customer’s journey up to this point, and then they disappear somewhere and we don’t hear from them for a long time. What’s going on in there and how could we learn more about that?

 And then they come back and why did they come back? Was it a claim? Was it a policy expired? So you build that holistic picture and now you’ve got this, you know, not to get too deep into the weeds, but you’ve got this data structure, but that is like, here is the one file that is everything the customer does and how they’re thinking about things. And now we can ask questions. Now we can do that exploratory analysis to say, when this happens at this point in the journey, just like Amazon does with their delivery trucks, they’re doing this in that sort of middle mile of the delivery process. When the truck goes on this route, what happens to the delivery time? And then the csat likewise in the, the, the insurance journey you know, what happens when they reach this roadblock and their documentation wasn’t correct, or they bought the wrong thing and they had to, they had to edit it. Then that that broad question of what’s in the data becomes possible to answer. So that’s the optimistic. Yeah.

MG: You, you made me think of something. So we’ve talked a lot about just the experience from, from when there’s an inbound from, from the insured, right? So something comes in to an underwriter that comes in from an agent or a broker from a portal from the insured, or there’s a claim that comes in from the insured, but there’s also the what information does the carrier have or do they need to capture to try and be proactive? And there’s a lot of, it’s really interesting of on how you identify customers and how you keep up. Because to your point, you’re only having certain interactions, one, once that that policy is bound, right? You’re probably having minimal interactions unless something like a claim comes up, which just inherently they’re having a bad day. Yeah. So it’s always gonna be an associated, like bad quote unquote bad experience.

So how do you proactively try to engage with the customer? And, and also in my insurance days I did a lot of work on projects on what types of services can we provide our insureds? That one is another touchpoint with them, but two you know, it improves their experience with us. So this could be and a lot of carers are, are are working on this and have solutions for this, but you live in, you know let’s, in an area that’s, that’s prone to hurricanes, right? And there’s, there’s a hurricane that’s, you know, forecasted to hit, what do you get? The insurance care can, can reach out to you and say, looks like a hurricane is gonna hit, you know, in three days time, make sure you do da da da da da da da at, at your home to prepare for it.

Or you know, where, where are there places that sh you know, trying to be helpful? Provide that type of information. The challenge with that from a information that the carrier has perspective is because you’re so i’ll, I’ll call it infrequently, that’s probably not, not fair, but infrequently engaging with, with the ins with the insured. Yeah. The email may be out of date, you might not have a phone number. Yeah. It may not be the right phone number. You know, are you doing this via email by mail, by text message? So there’s a lot of ways that they can try and reach out to you. It’s never a guarantee it’s gonna be there. So there’s a lot of that data quality question around email and phone number and a lot of debate going on of, of which is the right, if you, if you could only collect one Yeah. What is the one that you should be collecting? And there’s ongoing discussion, debate and, and challenges associated with, with those. That I won’t, I won’t get into <laugh> here. But I think that’s a question too of how do you improve, you know, CSAT and NPSs trying to be proactive and not just on the interactions that come in that are just by necessity already starting kind of off on the back foot because they’re not, they’re not great, great starts <laugh>, so why you’re reaching out to the carrier?

CW: No, absolutely. It, it’s a massive data management problem. And again, there are technologies for that, but I don’t know that there are technologies for figuring out, like, does this customer prefer email over text? Although mm-hmm. <Affirmative>, maybe demographics are helpful. Like, please don’t ever email me anything. I’m not, yeah. I’m just, I can’t have an unopened email in my inbox, so it’s going away.

MG: Yeah. or I do not have a landline in my house, so <laugh> Yeah. You know, exactly. Like you know, things like that. I, for sure. But yeah, I think there’s, there, there’s a, this is an interesting area just because there’s, there’s a lot of different ways that you can attack it. And I think again, the, it could have been happening much, much sooner for, I’m sure customer experience has been at the heart of <laugh> Oh yeah. Of insurance for a long time. But I, from my experience, the crux that really hit and, you know, the 20 14, 20 15 timeframe, that’s when I was, was kind of around it and seeing all of these different roles and efforts kind of stand being stood up to, to try and prove it. And I think it was because of all of the new digital channels and ways people were shopping for insurance, and even that, to your point, the new demographic coming up to buy insurance for the first time. Right. yeah. So all, all of that,

CW: My, my daughter’s gonna want to buy her insurance by exchanging memes with someone, I’m pretty sure. So <laugh>, gosh, gotta gotta stay ahead of the curve.

MG: We didn’t even touch on this, but you know, we’ve talked to some companies whose strategy is to use social media influencers Yeah. To get to, to get their product out there in the market. And that, you know, I am a millennial, but I’m the millennial that, like, I’m, I’m the older millennial, and so I didn’t spend my entire life, you know, exchanging on, on the, the, the web. And that to me is like brand new to me. But I, I, I recognize that now it’s, it’s an actual career path to me, <laugh>, you know, on, on social media and, and to have these types of engagements where you’re representing a ton of brands and, and selling that. So yeah, that’s a whole other set of questions on how, how the customer experience and those interactions and how do you select people that are representing your brand to a whole new nonprofit.

CW: Yeah. Who’s gonna make the best unboxing video for your marine vehicle insurance you just purchased? 

MG: <Laugh>.

CW: I so marketing wanted us to talk about how automation enables insurance companies to increase these metrics. I, I actually wonder if automation should be used to increase these metrics. Mm-Hmm. <affirmative>, I think there’s a limit in the sense that, you know, to, to your point earlier about the chatbots, eventually you’re just gonna be spamming, like, get me to a representative. And I think there’s a question of like, how far you should go with this and how much mm-hmm. <Affirmative> should be involved. Cuz as much as I don’t like calling up a vendor for help or whatever, if, you know, someday you’ll have to, and at that point, I don’t, I don’t want to have to figure this out for myself. I want someone to, I wanna feel like I have a partner that’s helping me through it. Right? 

MG: Yeah. But a partner that has read your exchange with said Chata and knows where you are in the process and then can just pick up where you left off as opposed to restarting

CW: Yeah. Yeah. The, the old handoff to three or four different departments and it’s like, okay, here’s what we’ve been through, here’s why I’m now fix this. Yeah. Yeah. Good. I want to, we’ve been all over the map here. I want to try to Yeah. Summarize for everybody before we, before we roll this up. Okay. So there was a, there was a period where this was a huge focus for insurance in the recent past, and now it’s, it’s sort of coming up again, especially as we’ve been talking about intake on the underwriting and claim side. There are obviously chatbots have been deployed to varied measures of success. We talked about that. We also talked about how there is a strong belief that analytics on all of this multi-channel data that we have is could be really powerful, but it’s a struggle to formulate the right questions to ask mm-hmm. <Affirmative> largely because of gaps in the data and understanding what, you know, what a customer wants and needs and what state they’re in, in their journey. And then I think finally we talked about optimism that we can get there. You just have, it’s, I think it’s a, the insurance company needs a different mindset, a different analogy for that customer journey than what they’re, than what they’re thinking about today. Did I miss anything?

MG: No, I think you’re spot on. I think the, the one thing I would add is that it’s probably being worked on differently depending on what, what segment that you focus on or what channel you focus on, or whether it’s underwriting your claims. And so, again, I’ll just put out the call again, if there’s anyone listening that does this day-to-day and is focused on this, we’d love to, to, to chat with you about about what you’re, what you’re seeing, what problems you’re trying to solve, and your take on, on automation and how it helps. I think we, we touched on a lot. So this has been another episode of Unstructured Unlock. Thanks for listening.

CW: Bye everybody. 

Check out the full Unstructured Unlocked podcast on your favorite platform, including:

Subscribe to our LinkedIn newsletter.

Get started with Indico

1-1 Demo



Gain insights from experts in automation, data, machine learning, and digital transformation.

Unstructured Unlocked

Enterprise leaders discuss how to unlock value from unstructured data.

YouTube Channel

Check out our YouTube channel to see clips from our podcast and more.