Building Flexible Data Architectures for Enhanced Customer Engagement

[music]

Kevin Niparko: The problems that we see our customers running into that really feel intractable or are the ones more on the people and the process side of data. It's something that technology can help with. It's something that CDPs can play a role in, but I think we're also realistic that no tech or software is going to be the silver bullet, it's about different parts of the organization coming together and aligning on an overall data strategy that everybody will abide by.

[music]

Kailey Raymond: Hello, and welcome to Good Data, Better Marketing. I'm your host, Kailey Raymond, and today we're talking about the evolving landscape of data management and how to keep up with it. The key to staying ahead of the curve is having flexibility, extensibility and interoperability built into your data architecture. In fact, as you hear in today's episode, one of the most important requirements is that the requirements will change, this mindset shift is essential for modern enterprises to quickly and continuously adapt to the proliferation of tools and new technologies, a data strategy that can evolve with the needs of your business and serve your cross-functional stakeholders, and most importantly, your customers starts with an understanding that the only constant is change and that the ability to adapt is what will set you apart.

Kailey Raymond: Joining me is Kevin Niparko, Vice President of Product for Twilio Segment CDP. In this episode, we discuss future-proofing organizations to take advantage of AI breakthroughs, accelerating time to value in solving problems through data strategy alignment. But first, a word from our sponsors.

Producer: This podcast is brought to you by Twilio Segment. 92% of businesses today are using AI-driven personalization to drive growth. However, successful AI-driven engagement is only as good as your data. Be it targeting your top accounts with relevant ads or delighting customers with personalized experiences both online and in-store, Segment has helped thousands of companies like Intuit, Fox, IBM and Levis be AI ready by laying a foundation of data that they can trust. Want to get your data AI ready? Learn more at Segment.com.

Kailey Raymond: Hey, everyone. Today, I am joined by a special guest, Kevin Niparko, he is the VP of Product here at Twilio Segment. Kevin joined the team in 2015 to lead growth analytics, and now he is really leading Segments Product Management org. He's led a variety of different products over the years from connections, cloud sources, ETL, and profiles. Kevin, a wealth of knowledge. Welcome to the show.

Kevin Niparko: Kailey, great to be on. Thanks for having me on the pod.

Kailey Raymond: Well, first, I need to learn about this nearly a decade of work that you've done at Segment so far, you've really earned your startup badges, so tell me about what that journey has been like from when you started to now VP of Product.

Kevin Niparko: Yeah, absolutely. So, I started as Segments' first data analyst, so I was helping the early team figure out the business model, the product direction, helped define some of our go-to-market strategies through our own data and joined right around the segment series A. And so being the lone data analyst at a data company it means I was often playing this role of the internal customers or on the front lines testing the new products and features that we were rolling out and thinking about how data teams and marketing teams could work together to actually leverage this.

Kevin Niparko: You know, I remember on one of the first days on the job, our head of sales came over to my desk and had started asking me how familiar I was with SQL, how my D3 data visualization skills were, whether I'd never written an ETL job out of Salesforce and we laugh about it now, but that was definitely a deer in the headlights kind of moment. I knew absolutely none of these things and I was terrified, I had found myself in the wrong seat, but I really got to spend the next year learning the hard skills, the tech and the soft skills that are required to build a data organization in a rapidly growing start-up, so it was definitely a ton of fun and a lot of learning.

Kevin Niparko: I think a lot of the questions that were being posed to me were the ones that data teams are still trying to answer, trying to better understand our customers, the ways that we can serve them and the ways in which our business model can evolve to provide even more value for them. And so what I realized is that Segment was really at the front lines of helping us understand what we needed to do as a business, how to grow and how to scale, and there was so much opportunity for data teams to actually use customer data in more advanced ways. This is right around the time when data warehouses were just taking off and a lot of people, ourselves included, we're figuring out how they would fit into our stack, and I think that's a story that we will continue to explore and talk about. From there I helped start the product team with Peter, our CEO at the time, and a few others, and have been building out the product and team ever since.

Kailey Raymond: That's a very relatable story of somebody like high up coming up to you and you know deer in the headlights, absolute panic moment, but you've learned a ton and built a ton since you've been here. That's so awesome to hear, especially the dog footing, drinking your own champagne from the very get-go and really feeling the acute problems of our customers is like one of the people that's actually leveraging the product. Very cool.

Kailey Raymond: As a VP of product, you're obviously building that vision to really meet the evolving landscape and the needs of what's happening in the list landscape of customer engagement. Given this front line really view as to what's happening, I wanna just dive into some of these overarching technology trends that are shaping customer experience and engagement. You mentioned it already, but one of the things that we've seen in the past few years here is this ever-growing presence of these data warehouses. Right. So can you just walk me through what your thoughts on are on the rise of data warehouses and how that's impacted folks in terms of their data management?

Kevin Niparko: Yeah, absolutely. So there are two big long-term trends that started well before I started at Segment that I think are really behind the rise of the modern data warehouse. The first and this is an obvious one is that the price of cloud data storage has been coming down consistently year over year. AWS storage costs have come down by close to 7X since they first introduced Amazon S3, their cloud storage service. And so the cost of storing massive amounts of data has come down significantly. I think the other piece of the modern data warehouse is really this concept of separating compute from storage, which means that you only really need to pay for the on-demand queries you wanna run when you wanna access your data, and so this gives you highly scalable cost efficient and performance architecture around your data. And so when you combine these two trends, the wave that every data team has been able to ride over the past few years is that data warehousing and the ecosystem around it has become bigger, it's become cheaper and it's become overall better.

Kevin Niparko: But one of the hard lessons that I think anybody can relate to, if you've ever moved into a bigger house or a bigger apartment, when you move into a bigger place, it's not like you keep the same furniture, you get more furniture, you get more stuff, you accumulate more things, you have more surfaces that you need to clean, and so in that same way where houses are opening up all of these new and emerging use cases, but storing this type of data presents new problems. Right? You now have to get the process right around governing this data, there are new tools that need to be plugged in on top of this data. And so I think it's really about thinking, really from first principles about what is this data warehousing technology mean in the context of your overall business and your overall strategy towards data.

Kailey Raymond: I love that bigger house analogy. Extremely relatable. Make perfect sense. And I'm glad you're bringing up this idea of flexibility, which I think is becoming ever-increasing in this landscape for enterprises, is they're really trying to just continuously adapt this proliferation of tools and these new technologies emerging. They want that strong foundation, but they wanna be able to test and learn and adapt to everything that's really happening in the market around them. I'm wondering if you can walk us through some of these components related to that trend of a flexible enterprise.

Kevin Niparko: Yeah. Over the last decade, I've probably seen thousands of different implementations of Segment and CDP across thousands of different enterprises and data architectures, and I think the one thing that I take away from seeing all of those is that there really is no one-size-fits-all data architecture, right? There may be some modern data platonic ideal of what the best architecture is, but I can guarantee you that that does not actually exist anywhere in the real world. And I've never talked to a data team or a marketing team that didn't think that they were approached to data where their architecture could be improved on. And so why is that like why after hundreds of millions of dollars of venture funding and billions being spent on data infrastructure every year, and literally the best data technology that has ever existed in the history of civilization. Why is every data and marketing team still feeling unfulfilled in their approach to data?

Kevin Niparko: And I think the reality is, is that the needs of a business are a moving target, and therefore your data architecture is also a moving target. The sales team had Oracle CRM yesterday and today they're migrating to Salesforce, a support team is opening a call center with Flex and the growth team wants to add live chat to the marketing site, right? And so each part of the organization is moving independently, it's evolving in healthy ways that are pushing the business forward, but data teams are responsible for bringing all of that together and providing insights that can then fuel the customer experience. And so that's a really hard challenge, and I think it's really the reason why we take extensibility and flexibility as a first class design principle as we think about data platforms.

Kevin Niparko: One of the most important requirements that we advise our customers to think about is that your requirements for data will change over time, it's a consistently moving target, and you need a flexible approach, you need a platform that can be extended in various ways as your business evolves, as teams evolve, as the organization moves forward, so really consider flexibility and interoperability as key design principles in your data architecture.

Kailey Raymond: I just wanna underscore that for one second, you said something that's just so important and profound, and I just wanna say it again, which is just one of the most important requirements is that your requirements will change, that is... Yes, that's exactly right. It's like... The only constant is change. Well, it's true, right? So it's one of those things you have to keep in mind. I'm gonna keep that one for the show, I'm gonna put into my pocket and then I'm gonna re-use it on the rest of the guests, so thank you for teaching [0:12:09.2] ____.

Kevin Niparko: You're welcome to it.

Kailey Raymond: As a marketer I've obviously seen a lot of the methods and technologies completely changed since I started in my career, so I think this principal holds true for us as well, something really familiar to a lot of marketers, I think it's probably a list upload. Right? That's something that I've been doing for a really long time and that time decay, obviously, it kind of kills some of the momentum of deals, taking that data from those IRL events that are happening and then translating that into an actual usable insight for rep is super important, but I've actually had an experience where sometimes you do it too fast, where...

Kailey Raymond: I have this memory of about five years ago, we're running Splash for check-in, and one of the reps actually called one of the people on site when they were still at the event because Splash was automating in real time and it queued up for the SDR to actually call it 'cause they [0:13:07.9] ____. So I'm just bringing this up because I think that when it comes to flexibility, I think about real-time data really often, and I realized that not every campaign is built the same and not every campaign necessarily requires real-time data, which is one of these examples that I just gave. So I'm wondering about this kind of tension related to what we're talking about with data and the architecture, rounded of the necessity of real-time. Is it all or nothing? What does that look like for you?

Kevin Niparko: Yeah, I think that's a great example. And I think what we find is that customers and data teams come at this question from really two different extremes, some of it depends on the business model and the product that you are serving to your customers, but I think about an app experience like Robin Hood or DoorDash where the experience is predicated on being in the moment. A lot of those teams are saying real-time is the only way to approach data, and then a lot of B2B businesses or more traditional enterprises are used to doing everything in batch, and so a 24 to 48 update window to their CRM may be totally standard and acceptable for their sales team. But I think what we find is that it's not on either end of those extremes. The most mature data teams are really thinking about the diversity of use cases that the organization has, both internally and externally in their products, and are designing their data infrastructure in a way to serve the different performance characteristics of those use cases. And it ultimately boils down to creating this hybrid approach to real-time data.

Kevin Niparko: This often means you have highly performing set of services that are focused on real-time marketing use cases, and then a set of services that are more focused on offline batch workloads that can serve the latency insensitive use cases. So on either end of these extremes, you can see the on-site personalization where a new insight or audience needs to be generated for a user in the blink of an eye before the next page loads, whereas there are email campaigns that are constantly going out where it doesn't necessarily matter if the email is sent at 8:00 AM or 8:02 AM, and so really thinking about an architecture that allows you to service those different needs across the business and those different latency requirements is really key.

Kevin Niparko: I will say that one of the interesting historical points of Segment is that we built for the real-time streaming use cases first, so our tracking API can handle 2.5 million events per second, just massive scale and performance of our systems. And it wasn't until a few years later where we really started expanding the platform beyond on real-time use cases, and a lot of legacy vendors in the space, we're going the other direction, trying to add real-time to a system that was originally designed for batch workloads.

Kailey Raymond: I feel like starting with real-time and go in the other way, it's probably a whole lot easier than the reverse.

Kevin Niparko: Both have their challenges for sure.

Kailey Raymond: Yeah. But I like that, a hybrid approach sounds right to me, and I mean, obviously, in a lot of circumstances you do want that real-time, but it's not needed for everything. I think about ad suppression is one of those, where it's a great example where real-time can save brands real money on ad spend, events. Maybe the SDR didn't need to call that person when they were still sitting in the seat, then that lag time might be appropriate, especially for more B2B use cases. Kevin, I don't know how we've got even 10 minutes into this conversation without saying the buzzword, I'm gonna say it. Okay.

Kevin Niparko: Go for it, let's do it.

Kailey Raymond: AI.

Kevin Niparko: There it is.

Kailey Raymond: We need to address it, I think it's here, it's the time in the show. We've been talking about data management, we're talking about warehouses and how businesses are demanding this flexibility, I'm wondering how all of those trends together are intersecting with this really just undeniable force of AI.

Kevin Niparko: Yeah. I think one of the things that our teams have been reflecting on is that the AI models that we're working with today are the dumbest and least capable that they will ever be, and that's kind of a hard thing to internalize and believe because the latest models feel very capable and very smart. And so broadly, I think businesses and ourselves included in this are really prioritizing laying a foundation for incorporating AI into our products, into our processes, into our organization, and setting up an operating model that will allow us to be flexible and agile as the landscape changes, as these models get better.

Kevin Niparko: I think the limitations that we're running into today around AI are likely to go away in the next year, so really question is, how do your future-proof your organization, your data platform and your products to be able to take advantage of new breakthroughs in the AI landscape. It's not just about what you're doing for GPT4, it's how are you setting yourself up for GPT5 and GPT6 and all of the additional capabilities and intelligence and reasoning that are likely to come from the evolution of these models.

Kevin Niparko: And so the other thing that we're seeing is this Cambrian explosion of both models and the toolkits around the models, and so while we're in the early innings here it's great to see a lot of experimentation, but we also recognize that a lot of these paths may turn out to be dead ends. And so we're thinking about ways in which we can experiment quickly, see what's working, throw the things out that don't seem super promising, and not to get hooked on any individual implementation because it's likely to evolve over the course of the next six to 12 months.

Kailey Raymond: I think that's really smart. I mean, we're in that really test-and-learn kind of mode as it relates to AI and nothing can be held sacred. I think the thing to remember though is like, you got to keep in mind the backbone of test-and-learn is good, trusted, complete data so you actually know what outcome you're steering towards and whether or not there's validity to it. So we'll get back to that, I'm sure. But how are we using at Twilio Segment AI in our product development strategies?

Kevin Niparko: Yeah. I think we are approaching it through a few different lenses. The first is really thinking about accelerating time to value with copilots. And how can we get more businesses up and running with good data faster? Historically, there's been a lot of upfront planning and hands-on keys implementation to get data right. And so the idea behind copilots is that we can automate away some of the repetitive tasks required to maintain and to govern good data at scale and that's really what we're thinking about with our copilot strategy. A good example of this, we have this copilot that can automatically, automatically map data into a new integration, and in some cases, even write a entirely new custom integration from scratch. And so you can take what could have been hours or days of an engineer or data person's time and turn that into a quick review of the work that the copilot is doing and I think this is so exciting.

Kevin Niparko: A second area that we're thinking about is really advancing the modeling and prediction techniques on top of the customer data that Segment is helping businesses collect. And so we're introducing out-of-the-box customer AI predictions. These can generate recommended audiences and predictions about what users are likely to do in the future. Most marketing segmentation today is based on what users have done previously, right? So generate an audience of users that have purchased in the last seven days. But the question is like, who is likely to purchase in the next seven days? And who should you be marketing to? That's predictive audiences and predictive journeys. We're seeing customers with some really exciting, impactful early results here. So one customer saw a reduction in 26% customer acquisition costs through paid channels with predictive audiences versus their standard audiences. So that can go a really long way for marketing teams who may be facing increasing pressure for profitability and preserving return on ad spend.

Kevin Niparko: And then the third area, last area that I'll touch on is this concept of a trust layer for AI. As businesses are looking to rapidly adopt and incorporate AI, we're talking about flexibility, there needs to be the right checks and controls in place to ensure that the data that is being used and input into these models is trusted and compliant and accurate. And this is really where consent management, centralized privacy center, and data governance capabilities can play a role in making sure that the inputs into the models are both accurate as well as compliant.

Kailey Raymond: As a marketer, that idea of pre-built predictive audiences is just like, it has me absolutely salivating. I mean, I've been in roles where I've had to sit weeks or months waiting for an audience to be built of your right. Things that were behaviors that have already happened. So very, very cool technology and having that out-of-the-box ability to build those audiences is a game changer for efficiency, obviously, but you're also clearly seeing the results with ROI and some of that as well which is cool.

Kailey Raymond: And to your point on trust, I think that that's something that, especially on this show, we always talk about this, it always goes hand in hand with the AI conversation, which makes perfect sense because organizations likely are never gonna be comfortable with adopting AI until they actually crack that code on trusting the data that it's built on. So it's kind of like you got to start somewhere and it's usually gonna be with that data infrastructure to make sure that you feel comfortable with even testing it out. What do you think the future of AI marketing and customer experience looks like?

Kevin Niparko: Yeah, well, I remember coming across this conversation with Danny Meyer, who runs a bunch of high-end restaurants in New York, and he was talking about what makes the Michelin star customer experience. And I was amazed at the level of research and diligence that his staff goes into for the folks eating at his restaurants, right? It's also that they can greet their guests by name. They can create this really magical dining experience for somebody who's celebrating an engagement or a birthday or a major life milestone. And Danny has this nightly ritual where he reviews the names of the guests that are visiting the restaurant and the off chance that he runs into them on the street, he can ask them about their experience and what they thought of the food and really engage with them on a personal level.

Kevin Niparko: And so I feel like those types of luxury, the high-end experiences, the moments that feel magical and spontaneous, but it's really a lot of upfront work and preparation that goes into making those magical moments, that can become more of the norm. That doesn't matter whether it's a visitor on a website or a returning shopper to a store or someone coming in to order a slice of pizza from the local pizza restaurant. You can imagine that if you were able to sit alongside all of your customers and advise them on how to best use your products, the problems in their lives that you could potentially help them out with, they would have a much different experience than they do today. And it feels like AI is going to make those experiences more prevalent and more abundant.

Kailey Raymond: It's so personal and it's so human. And that's what I love is like, it's kind of allowing for us to return to humanity and really get to know each other a little bit more deeply. I mean, I love your Danny Meyer reference. You're talking to a girl who lives in Brooklyn so shout out to my New Yorkers. Appreciate that, Kevin. [laughter] So a lot of what we've been talking about is obviously not easy for enterprise businesses to solve. You talk to our customers all the time. What do you think are some of the biggest challenges as it relates to building this data foundation to drive customer engagement?

Kevin Niparko: Yeah, you know, I do think that there are some hard, unsolved technical problems that data and marketing teams are really wrestling with right now. I think advanced identity resolution that can incorporate signals from a variety of different sources and resolve identity to user profiles in near real-time, still a hard challenge. Operating real-time infrastructure at scale, I think continues to be a challenge, especially as the volume of data increases and the uses for real-time data are also proliferating. I also think that connecting to tens or hundreds of different systems, each with their own API continues to be a lot of toil and drag for engineering teams that they want to just focus on the actual product and customer experience. So no doubt these are hard and unsolved technical problems today.

Kevin Niparko: But more often than not, the problems that we see our customers running into that really feel intractable are the ones more on the people and the process side of data, right. It's getting the sign-off for the ways in which data can be used in a compliant way because you don't have the right consent controls in place. It's marketing teams waiting for data teams to produce that CSV that you mentioned earlier that they need to run the campaign. These are the types of real-world challenges that ultimately slow the enterprise down and can really be a bottleneck on the use of data and really putting the customer first in the experience. And it's something that technology can help with. It's something that CDPs can play a role in. But I think we're also realistic that no tech or software is going to be the silver bullet. It's about different parts of the organization coming together and aligning on an overall data strategy that everybody will abide by.

Kailey Raymond: I really like the practicality behind that, 'cause I do think that oftentimes folks can fall into the trap of believing that tech is this panacea. But the reality is that you're bringing together an org and then designing processes to support that vision. Like, that's sometimes the hard work, and that's often the place where folks can stumble, and most or a lot of tools might become shelf-ware because of that. And so, you know, this concept on the marketing side that we've been playing around with is, that's related to this, we're calling it the customer engagement stack, which I think kind of speaks to this a little bit, which is like, you know, a tech stack, for instance, might be owned and operated and optimized by the marketing team, and then it kind of further cements those silos that are causing some of these foundational issues because it's oriented specifically around one team not necessarily an outcome. So this idea that you can build like a cross-functional stack that really helps bridge together the data stack, the MarTech stack, the whatever you name it stack and the service of the goal of customer engagement is a reflection of that strategic alignment and kind of this issue that you're pointing towards which is really a people issue, not a technology issue.

Kevin Niparko: Yeah, I think that's really well said.

Kailey Raymond: Thanks, appreciate that.

[music]

Producer: This podcast is brought to you by Twilio Segment. According to Twilio Segment's state of personalization report, 69% of businesses are increasing their investment in personalization. And with the world going cookieless, first-party data is the most reliable, trustworthy, and compliant way to create tailored experiences. Imagine suppressing consumers that just purchased that pair of shoes from your ads and decreasing acquisition costs by 5X or sending the perfect text promoting your other beauty lines and converting 50% of customers. That's the power of Twilio Segment. Learn what's possible at segment.com.

Kailey Raymond: Now, we're in the moment of the show where we get to the namesake, the classic question of good data, better marketing, which is, Kevin, how would you define good data?

Kevin Niparko: I describe it as good data is anything that helps really clarify your understanding of the world around you. For marketers, that often means a better understanding of the user journey, the wants and needs of the customer, and the ways that they can really help the customer solve the problems that are in front of them. I think a key ingredient to good data is it also needs to be usable data, which means that it's accurate, it's complete, it's trusted, end users have consented to how it will be used. And I think all of those are really important to get right from the beginning, because you don't wanna find yourself with a data set that you either don't trust, feels inaccurate, or can't be used in the ways that are required by the business.

Kevin Niparko: And so you know, more and more, I also think as we look ahead and we touched on some of the ways in which AI will change the customer journey, but I do think it will also change our definition of good data, right? So it's evolving from something where it's really helping humans make better decisions to an input into personalization models and automated decisioning. And it's interesting to think through the ways in which that might change how we define good data and how we actually collect and govern it because the use cases around it are likely to change.

Kailey Raymond: You just blew my mind a little bit with that one. Yeah, when the robots are taking over the definitions and the yardsticks and the benchmarks, it's just a completely different way to think about the world. Wow. Okay. I'm going to think about that one for a little bit, but I really enjoyed that. That was really astute. I think the usability part of it, you're right, it really speaks to me directly to the heart. It also like bringing it back, it means you trust it. The themes, Kevin, they always just, they really connect. I'm wondering if you have any examples of how people are using good data, any stories or highlights from our customers maybe?

Kevin Niparko: Yeah, a company that we've gotten a front-row seat to is MongoDB. So MongoDB is a modern database and B2B platform. They have this really complex sales cycle with different buyer personas and buying groups that do a lot of upfront diligence on the product before they make a purchasing decision. And hey have all of these different disparate touchpoints, across live chat on the site, sales calls, support tickets, in-app, and long-doubt behavior and so this became a pretty gnarly data challenge for them to tackle. They were really looking to get a better sense of their customer records, their account records, and make those more usable across all of those different touchpoints.

Kevin Niparko: We've been able to pair with them for the past few years and it's been amazing to see what they've been able to accomplish. They now have all of these trusted golden records for their users and their accounts. They sync those into their data warehouse via Segment and the capability we call profile sync, but they also feed data back out of their data warehouse so it's a bi-directional connection with reverse ETL capabilities. And so they're able to leverage all of the breadth and data catalogs that they've built in their data warehouse alongside their real-time user and account streams, which is super cool. With their user profiles in their data warehouse, they're actually joining in 181 other tables and identifiers and dimensions to really round out their understanding of who their customers are and where an account is in the cycle. And this has allowed them to get the most out of their data warehouse, but also service new use cases around real time live chat and really accelerating the time to value and pipeline for their sales team.

Kailey Raymond: Wow. Yeah. And in turn they're also reaping the benefits of customer satisfaction. So those open rates, click-through, all those things that me as a marketer, I look to everyday, event registrations. I love that MongoDB story, mainly because I think it's an awesome example of what we were just talking about which is this collaboration between teams. The engineers and the technical worker is being done in the background and often might go unseen because the actual activations are being done by the marketers and the benefits are really being reaped by them, but it's a collaborative joint effort of really making sure that two teams are stirring in the same direction and that direction is customer satisfaction. So super cool example of pairing those two teams together. Any companies that you think are doing this whole thing right that you admire?

Kevin Niparko: Yeah, so, you know, I wanted to share the story of the only direct physical mail ad that has ever converted me and this is a recent example, so this isn't like way back when. I'm the proud dog dad to a pup named Abby, and she's definitely a little bit pampered. We definitely take good care of her. We got this flyer in the mail, I think it was for a free visit or a coupon off the first visit to get started with a service called Modern Animal. They're a modern approach to the veterinarian clinic, and they were opening up a new location near us. So, scanned the QR code on this piece of paper, signed up for the app, and it turned out the overall vet experience was just 10x better than any other vet we've been to.

Kevin Niparko: Abbey now has this fully digital medical record. We're automating scheduling and reminders for different visits and vaccines. You know, it took this pretty antiquated offline vet experience and turned it into something that was super easy for dog owners and our family. And my friends are probably now tired of hearing me rave about how awesome our vet is, but I can't stop talking about it. And so I think there are sort of two takeaways from the story. The first is that, direct mail can still work. [laughter] And then the second is that the marketing and product experience go hand in hand. And if you can deliver a 10x better customer experience, people will be raving about it and sharing your product through word of mouth because it is just so much better than the competition.

Kailey Raymond: I love that. Yeah, I mean, vet offices for sure have a long way to go. So they are crushing the competition with all of their automations for sure. [laughter]

Kevin Niparko: Absolutely.

Kailey Raymond: And multi-channel marketing, way to go. They're killing it. That's amazing. I love it. Last question for you before I let you go today. What steps or recommendations would you have for somebody that's looking to up-level their customer engagement strategies?

Kevin Niparko: Yeah, I think the most important thing is that you think through your data strategy and really be willing to play the long game. I think customers can see through short-term hacks and transactional relationships. It's not where the world is going. It's not what anyone wants. People wanna connect. They wanna to understand the people behind your brand, and they wanna feel like there is somebody on the other side who is really looking out for their best interests. And data can play a really key role, especially in digital interactions, in helping you understand those needs and better build products and services to serve them.

Kailey Raymond: When it comes to data, playing the long game. Kevin, thank you so much for being here. I learned a ton and it was just a real pleasure.

Kevin Niparko: It's been a ton of fun, Kailey. Thanks so much.

[music]