Customer Journey Optimization Through Understanding Customer Intent With Alex Levin

Let me ask, why do so many companies lose their customers before they even know they are slipping away? Why are we still chasing clicks and impressions instead of connecting with real, ready-to-buy customers? In a market flooded with noise, how do you actually cut through and reach people at the exact moment they're ready to act? And what if the old way of waiting for customers to find you is no longer enough? Could understanding customer intent be the key for your next growth breakthrough? Welcome back to Mastering Tech Growth, where we interview industry leaders to bring you actionable insights and inspiring stories. I'm your host, Mike, and today we're diving into intent-driven marketing and proactive engagement. Our guest is Alex Levin, co-founder and CEO of Regal.io, a leader in customer journey optimization through understanding customer intent. Alex has guided companies to go beyond reactive strategies and meet customers exactly when they are ready to engage. Today, he'll share how anticipating and responding to customer intent can unlock untapped growth. Alex, we're thrilled to have you on the show. Thank you for joining us today.

Thank you for having me.

Of course. It's gonna be amazing, Alex. So we have this question before we kick off into topic-driven questions, which we ask every single guest. When I say mastering tech growth, what comes to mind?

So, you know, if you're gonna go the path of raising money from outside investors, you know, there's gonna be very high expectations on growth. So I think the first thing, you know, I often talk with people about is like, is that really what you're signing up for? Like, do you wanna really go and raise money from outside investors? I know it sounds glorious, but it does bring higher standards. So if you've made that decision, then yes, you're gonna have to start figuring out how to do it. The money that you're raising presumably is to go and do things you couldn't have done without it. With the growth specifically, I'd say two things I think about always. One is I think people misunderstand how growth and marketing works. They believe if they just go to some you meet a customer once and they say, oh, here's this great widget or product or whatever that I bought, service, you know, buy it, you know, they'll just say yes. And the reality is that doesn't happen. Even if you have a great product, you know, it takes 5, 6, 7 times, you know, for you to sort of be showing that product to the customer in some way before they get interested. So I don't even pick a very big company like an Airbnb. It's not like the first time you heard about Airbnb, you go, yeah, let me go and live in somebody else's house. You know, you had to have a friend who did it. You have to see an article in the New York Times. You saw an ad, you got a discount, and finally maybe you tried it. So first, there's a lot of touches that go into it. And second, especially in sort of growth, whatever is working today is probably not going to work tomorrow. So it's constantly this, uh, sort of role where the, you know, or sort of function where reinvention is necessary. So, you know, you're going to come up with some new idea and you're going to go out and do it, and for a while it's really going to be great. And then the reality is, if it's a good idea, other people are going to copy it and the payback is going to revert to the mean. Or even worse, it's going to result in something that's not successful. So you're you're going to then have to be smart again. So it's kind of a job that's thankless in some ways because you're going to constantly have to come up with new ideas for growth.

Okay, so there's a, there's a lot to unpack there. Um, can I rephrase this, Alex? What if I ask, you can only choose one factor from all the factors possible which influences the growth, what in your opinion is like the factor for growth?

I mean, I suppose like if you're really gonna get like, you know, depends how you wanna think about it in terms of math or sentences, but you know, the rate at which, you know, one customer can turn into more than one customer or, you know, $1 can turn to more than $1 is ultimately, and people call it different things, but that factor is what leads to growth. So, you know, people are trying all kinds of different things to do that. In the end, you know, it's some version of like, can you really satisfy that customer with what you're doing? If you can, right, you know, one customer starts spending more or they recommend to other people or they stay on for a long time. If you can't, they churn, they leave. And that, you know, one customer turns into zero customers and your business is really dead because you're constantly having to go find new customers. So I think if you can somehow turn $1 from a customer or one customer into more than that, like that's magic.

Okay. So we're talking about network effect over there where you get one customer, that customer recommends more customers, these more customers recommends more, and it's just And that's one, you know, you're asking me for one thing that I think about.

So in different businesses, it's different things, right? Some businesses it's referral, right? It's referring other people. Some businesses it's getting them to pay more. Some businesses it's how its usage goes up. Some businesses, you know, just, you know, they, they stay forever and, you know, they never leave. So depends on the business.

Okay, I get you. All right. Let's go back to the main topic then. So what does the term customer intent mean in the context of modern marketing? As in, when I say modern marketing, I mean like today's marketing with all the social media available, with all running ads and SEOs, with everything that you can do today. What does the customer intent really mean?

Yeah, I mean, so take a little bit of a step back. Call it 30 years ago almost exactly is when, you know, the big retailers started coming online and, you know, we started having a different way to look at customers because we weren't just having them walk into a store. They actually like were leaving these breadcrumbs all over the internet. And very quickly people realized that we could use those breadcrumbs to start to understand a little bit better what they were looking for. And so when I say intent, that's what I'm talking about is, you know, what product do they want to buy? What value propositions matter to them? You know, What, what things do we know about them that are going to help us understand what is the right next step for that customer? And, you know, by and large, particularly in the United States, customers have decided that trade-off of, you know, giving companies intent data or sort of online data in exchange for better quality services is fine. In Europe, the trade-off is a little different, and people are a little more nervous about giving their data in exchange for better experiences. But in the States, that's become sort of the standard. Marketers, you know, very quickly sort of grasped that there was valuable data there and they started improving the marketing for customers by offering personalization, A/B testing, different pathing, different segmentation based on what they can know about you. So you come in and you're clicking on SUVs, well, the marketing should be about SUVs. You're clicking on sedans, it should be about sedans, so on and so forth. I think now at this point, there's quite a lot of sophistication in marketing around what is considered intent-driven or event-driven marketing. The shocking piece, actually, the place where we work more is that contact centers have not gone that way. So if you think about the full experience you have with the brand online, you have a self-serve experience, you have some marketing you engage with, and you engage with the contact center, right? You have human beings on the other side that you're engaging with. For whatever reason, in the contact center over the last 30 years, retailers have seen it as a call center and have said, "Oh, we're not gonna invest there. We're just gonna," lower the amount you talk with our contact center. So whereas in marketing they invested in using that data to give you a better experience, in contact center they didn't. And so the result is today, you know, uh, customer satisfaction with contact centers, it was customer service is an all-time low. People hate working with the customer service of these big brands. Um, luckily there's some new brands coming online now, and there are more complicated industries coming online now, like, you know, healthcare is coming online and insurance and lending, where that sort of old low-touch model in customer service service does not work. And they're starting to take a lot of the playbook from marketing of intent-driven marketing into the contact center and say, hey, the same ideas we have in marketing we need to bring to the full customer experience. So, uh, we have to figure out, you know, which customer should we talk to, when, with what message. When it's inbound, what do we do with them? When it's— when we're reaching out to them, you know, what do we do and what channel and when and why? And I think brands that are doing that are seeing huge ROI because, again, it's a new thing. So what we were talking about before, if you're the first one doing something, you have a strategic advantage. And so the brands that are starting to do this in the contact center are having to massively overhaul their technology to do it. So there's real investment, but are seeing huge results as a side effect of making these changes.

Okay, so when we say contact center, I mean it's people making you a phone call, right?

Phone call, email, SMS, AI agent, you know, all these things. Basically the contact center is responsible for the business critical communication with that customer. And it sits completely separately from marketing on most orgs, even though these are probably the more critical experiences, right? So the marketing is sending you something when you don't care, but when you do care and something goes wrong, guess who you're talking with? The contact center. So in some extent, those critical moments are higher profile moments, but the brand is not paying attention to them and not figuring out how do I treat this person like an individual in that moment? So I can't tell you how many times I go work with a contact center And they go, "Oh, we're so good at treating everybody the same. No matter who comes in, we're gonna treat 'em exactly the same and we're so efficient at it." And I go, "Guys, you shouldn't be proud of that. You shouldn't be saying that." Instead, you should be thinking about based on exactly everything we know about this customer when they come in, how do we give 'em an experience where they go, "Wow, like how did that brand know that about me and treat me in a way that I really feel like an individual?" Because now, you know, I'm a customer for life.

Can this become scary like very quickly where, you know, if I'm reaching out to your brand, I've been doing business with and I go like, "Hey, I need this help with," and they go like, oh, you know, we know yesterday you were driving Tesla. How about, how about, you know? Yeah.

So this is why, like I said, in Europe, people are less comfortable with the trade-off of using personal data to drive personalization. In the States, by and large, people are comfortable with that trade-off. You know, you go to some website for a car and it shows you the same car you were looking at yesterday. Or maybe, you know, you leave that website and you go somewhere else and it shows you an ad for the same car you were looking at yesterday. People aren't stressed by that. This is the same concept in a contact center. Right? You know, the contact center should also know, hey, you want a red Tesla that has whatever. Now, there are very concrete rules, like, you know, you can't use data from other places. Like, you can use first-party data that you have the right to use. So there are concrete rules, but it's not about being scary. I think about it as, you know, imagine 50 years ago, you might walk into a restaurant or a private club and they go, oh, you know, Mike, it's so great to see you. We're going to put on the music you like, give you the table you like, bring you the drink you like, you know. And people appreciated that level of high touch, right? So to some extent, a lot of what we sort of work with brands on is how do you bring back that very high touch experience and use modern technology to do it at a very low cost?

Okay.

So for the last 30 years, contact centers have been seen as a cost center and they've said, oh, well we have to lower costs. And we lower costs, customer experience is worse, but that's, it has to be that way. No, it doesn't. You can both keep it very cost efficient and give customers an incredible experience with your contact center. Okay. And brands that realize that have a huge advantage.

So it's just about making the customer feel VIP in some way, shape, or form.

That's a baseline for sure. Like, you know, if you were to go to some company today and like you put a product, I don't know, you put a bunch of products in your basket and you got an email from them saying, "Hey Mike, we don't know anything about you." Like, you know, we don't know, you know, you'd be surprised at how bad they were at marketing. Like you've come to expect that brands know what you put in your basket and they know a little bit of what you're doing. I don't know, why don't we have the same expectation in the contact center? Like that should be baseline, just what you're calling VIP in my mind should be baseline. VIP should then be the next step ahead, which is, okay, now that we have this baseline where we're treating everybody the same, how do we start using these contact center interactions? So proactively engaging customers to get to outcomes. So take a healthcare example. One of our customers is Roman or Ro, which is a big healthcare company in the States. you know, and they're trying to decide, you know, who is right for Ozempic, who is right for different hair loss medication, you know, who maybe has a drug interaction that's wrong. So instead of just sitting back and waiting for the customer to contact them, their ability to reach out to somebody and say, well, actually, uh, your doctor just prescribed something that has a drug interaction with the drug you're taking, you shouldn't do that. Or based on everything that, you know, we're seeing, we think that this new drug will be better than the old one. Or there's a new opportunity for you. So the ability to now proactively get into the person's life in a way that is valuable to them is really incredible. Think about a school. So, you know, through some of the technology companies we work with, uh, they're pushing students into the big universities like Harvard using Regal. So as they're doing admission and they're convincing students to do it, there's a lot of opportunity in that to help students feel like they're, you know, really cared about by Harvard and really somebody that is important to Harvard, so they want to go there, right? Or you could just step back and say, oh no, we don't care about you. So, you know, you've got to decide as a brand, are you going to actually do the things that show you care about customers?

Okay, I get that. If I'm small business, Alex, so how— you mentioned basically data, you know, we need to have that information about the customer. So If I am already with the service and I already have the customers, I could be accumulating their data, you know, how they use my service and just trying to understand them and, you know, get that personalization injected into the support, you know, or call centers. If I'm a new business, I mean, I do know my customers, you know, I'm trying to acquire them, so I'm trying to get that market share. Is there a way to somehow you know, get the faster going situation by buying customer data from somewhere?

Look, I, I'm not in favor of buying customer data by and large. Like, I don't recommend that. I mean, certainly there are cases where there are reasons to do it. I'm not saying never do it, and I'll give you an example where I think it is valuable. But by and large, I'm not suggesting going and buying data. Like, let's say I start a new insurance company, just as an example, uh, you know, and I'm advertising in a specific state, you know, in the state of Colorado, you know, buy my new insurance and we're better for these for all sorts of reasons. As somebody comes in, even just from their IP address, you know where they live, the neighborhood they're in, maybe as they sort of enter some information, you know, do they own, are they buying a house, how big is their house, do they have a family? So I'm not saying collect random information, but if you're gonna be able to give them a better experience by collecting just a bit of information, either from like an IP address or some basic questions, just do that even for new customers, right? So I'm not suggesting you bought all this information in advance, but still, like, that's valuable. In this insurance example, there— it is an example where I think it can be valuable to have data in advance. As an example, somebody comes in and you're trying to sell them home insurance, and you— and to do that correctly, you need to ask them 50 questions about what is your home made out of, how close is the fire station, you know, when was the last time there was a fire in the house, blah blah blah. The homeowner may not even know that, and that makes it very hard for them to onboard. So it's an example where if you as a business can go out and buy that information, it's not scary because you're not buying information on a person, you're buying information about home. So that when the person types in their address, you go, okay, I know everything about the home. We can very easily price insurance for you. You've actually improved the experience. So where you can do something that is not scary to the person and not violating their privacy, but in fact improving their experience. Sure, I'm all for using data and buying that data.

Okay. From all the intents, actually, let me rephrase this. So I do have a service, how do I collect that data? It's just as simple as I imagine, you know, a simple installment of a software, I add the script on my website, so I start collecting all the data, and I imagine different intents are coming into the system, maybe they click something, maybe they liked, maybe they watched a video in full, a little bit, maybe they read an article. And all these different things which effectively gives me some business analytics to understand my customer. But from your point of view, what is a high intent signals where the businesses should just never avoid if they actually have an ability to look at different intent data?

Sure, so in every business is different. So it's easier than ever to instrument your website and email and all these things to get all the data. But now you have a million different data points and you go, what in the world matters? So some of the famous examples, you know, in e-commerce online, very early they figured out if the customer goes and puts something in their cart, but they don't buy it, that's high intent, right? Like why would the customer put it in their cart if they weren't very interested in buying it? So e-commerce, like cart abandonment became a very sort of critical moment. Facebook as just a very different kind of business was famous for identifying that if they could get you to put, you know, if you put 7 friends in your Facebook, whatever group, or you invited 7 people, like they knew something special was happening. So they really look for how do they get you to 7 friends. And if you got 7 friends, like they knew that was another step. LinkedIn has similar sort of mechanics. Most online sort of B2B businesses also sort of understand what are good indications of intent, like somebody going to a pricing page or things like that. But every business is different, it's not necessarily the same. So you have to figure out for your business what is the data that matters? Not every piece of data matters. And how are you gonna then use that to give that customer a better experience? Today, consumers don't, really sensitive to when brands do a bad job. So often the way this comes out is a brand tries to do something with a consumer, send them an email, send them a call, and they see the consumer, say spam to the email or not answer the call or something. And they go, okay, so customers must not wanna talk to us. So brands have taken away from this that customers don't wanna talk to them. But when you, we go and interview consumers and say, hey, do you wanna talk to the brand? Like, how do you wanna talk to the brand? They all say, yeah, of course I wanna talk to the brand, you know, and I want phone calls and I want emails, but they have to be relevant to me. So what brands, you know, a lot of brands do a terrible job of is they just send everybody the same gut. So even I'm a big fan of Patagonia. I think it's a fantastic brand, but I get a million different emails from them. And like, honestly, like I'm at the point where I'm gonna unsubscribe just because I'm not getting the stuff that's relevant to me. If Patagonia paid a bit of attention to the products I buy, you know, what I do with them, what sports I do, then they tailored their emails to me. Like I would never leave that because I do think there would be value to me. So, I think it's all about how the brand actually considers what they know about the customer to then personalize it, you know, instead of just doing what Patagonia is doing now, which is basically sending me every email that they have and not ones that are relevant to me.

Okay. So effectively it's a homework you need to do as a business to just sort of sit down and understand what behavior patterns you consider high intent with your customers, put it in a system, start capturing some data. And then effectively, I mean, yep.

Yeah, for sure. A customer like me, just take Patagonia's example, you know, they should be sending me stuff about rock climbing, about skiing, about fly fishing. Like those are clearly the things I buy stuff for. In certain times of the year, they probably should be calling me and helping me and saying, well, okay, we have 5 different products in this thing, but this one's the best one. And here's why we think we should have, you know, they could go much deeper and continue to earn a bigger share of wallet from me. You know, otherwise they're gonna be the ones when I go buy from other places instead of Patagonia, Patagonia because I don't understand that Patagonia has the right product. I don't know that they're really the right one to buy for something. So I think, you know, today they don't do that. They automate everything and they're sort of— they're very hands-off, which I think is the wrong approach with sort of customers.

Would you— would they lose you as a customer if, if a competing brand somehow would get your information and they would make you that phone call during the season?

Yeah, as an example, backcountry.com does a lot of work to to have— they call them gearheads— to have individuals that connect with customers that are spending more. They are— Backcountry.com is very clear what categories I spend money in, you know, when I buy different things. And so, you know, Backcountry.com says, look, you know, I know you usually buy a new jacket every 4 years for skiing. You've always bought Patagonia, but, you know, this one from Arc'teryx is better. Well, I'm going to switch because they've done the work to go and understand. I like to ski. I switch every 4 years. Here's the time where I'm going to buy a new jacket. So I do think like there's, you know, there's another company called Curated, which has a similar customer experience, even though they've done it a different way, where again, they're investing in the human experience to make sure that that customer feels like there's somebody really helping them.

Okay, that's a very good example, which I feel answered the question I wanted to ask about proactive engagement when it comes to the company. So, you know, once they set the behavior patterns, you know, they start capturing the data, they start understanding these high-intent signals, then it basically leads to what we talked in the beginning, is that proactive engagement. Just don't sit on that data. Make sure something is happening with the data, and sending an email is a distance. Sending SMS, WhatsApps feels like it's closing that distance, but really putting that phone up and going like, "Hey, what's up with Alex today?" Yeah, it depends how certain you are.

So as an example, in the last business we were in, we were the largest home services provider in the world. So we own Angie's List, HomeAdvisor, a lot of brands in Europe also. Uh, and think about moving. Moving is a service where if you tell us you're moving on Tuesday, you know, July 1st, guess what? Every year, that's the year when your lease comes up. Leases are annual, you know, if you're renting. And we know if it's a building with rental or not based on the data that we have. You know, if you own, maybe you're going to be there a little longer. But if you're renting, then it's not to say you're going to move every year, but we know every year there's a potential. So there's a high likelihood that we should be using that information right at the next year and saying, hey, if you're, you know, did you sign a new lease? Because if you did, we'll leave you alone. But if you didn't and you're looking about moving, here's the 3 things we can do with you to make it better for you. Well, we already know all the stuff you had last time, so you don't have to reenter all that. We literally have the same guys who came last time. We did a great job, ready to come to you again. In fact, because we know the date we're moving, we've gotten a truck that is only working in your area and we can give you a better price because it's going to be easy for them and for you. All of a sudden the customer goes, wow, like, you know, these guys are really looking out for me, instead of the company going, well, I don't even care about this customer or whatever. So it's an example where without any subterfuge, you're getting information that can help you make sure you're serving that customer better.

Okay.

Same with cars, right? But you have a car lease, everybody knows when the lease is coming up, you know, and what can you do as that car lease comes up to make sure that customer's experience is great? You know, there was a dealer that was offering me like to come to my house, check out the car, make sure it didn't have any scratches, you know, get me a loaner if I needed a new— like, I'm going to go work with that dealer that does that, not the one that goes, oh, I don't even care what you're doing.

Okay. So while you were giving this example, Alex, I thought about something popped in my head and it was at which point or size of a business you actually are capable, financially speaking, to allow yourself a service like that. Because, you know, I was thinking about all the infrastructure, all the data capture, or potentially, you know, buying into a service just like yours, uh, you know, to get that data set up and then, you know, do those actions, you know, get— getting people to actually make the phone calls. That, that all felt like, um, like sort of like a, you know, maturity level for certain businesses. So like From the clients you guys having, like what's the average of business size who actually utilizes strategies like that?

Yeah, so we can talk about it in a couple different axes. So first, you know, what is the value of the customer that is required to be worth doing this stuff, right? So if you're selling pencils and you know, you sell one pencil at a time and a customer is worth 20 cents to you, you're not gonna be able to do very much because there's not a good payback. You know, if you're selling life insurance and a customer is worth a few thousand dollars to you, well you could do a lot, right? So there's definitely a range. Mostly today we focus on working with customers that have at least a $300, $200 one-year customer lifetime value because that's sort of a customer you can do quite a lot. If they're below that, it's not that you can't do anything, but it's less interesting, you know, the things you can do. I think second, we think about, you know, as a, you know, as a size sort of indication, you know, do you have enough customers to do, to go and hire somebody who's going to work on this stuff. So, uh, you know, if you're a company that ultimately has 10 people, you probably don't have one person that's focused on technology. It's not— you don't have the time to really put effort into setting up all this technology, forget the cost. So if you're 100 people, you probably do have one person that works on technology and can set this stuff up. So are you a big enough business that you can have a full-time person working on technology? In the end, the cost of the technology is nothing compared to the cost of the human beings that are actually setting it up. So the human cost that matters more. The third one that we start thinking about is do you have enough users that are coming to you to start doing A/B testing and doing things in an intelligent way? As an example, if you're getting, you know, 1,000 customers a month, very hard to do A/B testing on 1,000 customers because even if you split it into two groups, 500 and 500, there's probably not enough data to know if the thing you're doing is significant. Now, if you have 100,000 customers a month, Very easy, right? 'Cause 50,000 in each of the two variants, you're very quickly gonna get to significant. So I'd say, you know, anybody with less than, call it 20,000 new customers a month, it's pretty hard to do real good A/B testing. You could do a lot of the strategies we're suggesting, but you're gonna have to fly more by the seat of your pants and what you think than doing sort of very numbers-based testing and A/B testing. So, you know, where do we focus? We focus on those sort of mid-market and enterprise customers where those things are true, right? They have somebody in the company that can make sure they're running the technology. There's enough users, they can do good A/B testing. You know, it is worth it for them. You know, the dollar value is high enough. That said, there are, you know, SMB products that start to approximate this stuff. They just do it differently. So as an SMB, maybe you don't use Regal, you go and use, I don't know, what's an example? We have a company we invested in called Slang in the restaurant space. Instead of like you setting everything up, what Slang does is it has an AI agent that responds to inbound sort of calls as a restaurant and helps either give person information or schedule bookings, you know, schedule a reservation. And it's, you know, it sort of is it a little bit of a black box or it's a little preset because the— they know the SMB doesn't have time to set everything up. They do it kind of for the SMB versus if you were to— if you were a larger restaurant group and you work with Regal, we would do everything custom with you. So there are SMB versions, you know, small and medium business versions of the products that we sell.

Okay, so let's say you come in to a new client, and for the sake of argument, let's just think of like e-commerce business. You know, there's a website, you know, they're selling things, just all industries, you know, from shoes to paints or whatever. And, you know, they have a size, they have customers, you know, they have all the numbers. So they do have some, some type of business and analytics. They probably have some tracking going on. It's potentially not very sophisticated, but you know, they do have some data analysis happening, people within looking at the numbers, you know, potentially coming up with different marketing strategies. So you come in and how does your approach look like when it comes to the customer journey optimization? It's like, you know, maybe not specific to Regal as a product, but like in general, like how people listening right now going like, okay, well that sounds great, I would like to try it. Like, what is the first step?

What do I do?

Yeah, yeah.

So a couple different thoughts on it. So one, when we come in, you know, we try to understand what data people are collecting on their customers, making sure that that is the right data and that we can get it, you know, into our system. Two, you know, what are they currently doing today? Sometimes people like to start by just replicating their current user journeys in a new system before they go and do anything new. That's fine. And then 3 becomes what are best practices. So we have very good data on in each industry what they should be doing to drive, you know, outcomes. And it's not— a lot of it is not brain science. It's not like it's crazy ideas. So I'll give you like a random example. Let's say like you get lots of people that come in and you're trying to call them back to say, you know, do you want to buy a house? Let's just use that house example. If you put them all on the same list, you call them all 3 times a day at times that work for you, and if they don't answer, you just keep calling. That's kind of not great. But if, you know, you take each person, you call them right when the lead comes in, right? And if they don't answer, you text them. And then when they pick up, you say, hi Bob, I see you're interested in this, you know, it's personalized. Well, the outcome is going to be much better. So it's not brain science to say the second one is better than the first. Largely, it's a problem with technology. Companies don't have the technology they need to be able to do the second, and so they stuck in that first, which leads to worse and worse results. So yeah, for each industry we have similar playbooks that can sort of help them move more in the direction of a sort of better customer experience and better customer outcomes.

Okay, so that personal optimization of that journey effectively then potentially need to reshuffle the tools to make sure you're capturing the right data, which we talked about. Just understanding the behavior.

And you need tools that are capable of using that data to do something, right? So like, if you have a phone system where all it can do is take a segment and call everyone the same, well, you're going to call everyone the same. If you have a system like Regal where each individual user can have a different experience, well, now that's possible, right? So, okay, it's very different. And then at some scale, once you sort of set up best practices, it becomes about A/B testing. So it's a constant exercise in trying new things. Hey, should you know When there's a, you know, somebody who's interested in a home and we have a new home that comes on the market that looks like one that they want, what should we do? When should we text? When should we call? What should the message be? How should we reach out to them? You know, so on and so forth. And it's A/B testing to find what's best for you.

Okay. So it does seem that there's a lot which depends on having that quality data and understanding and actually having that personalization level of data. Is there any like real challenges when it comes to collecting that data, which people should be aware and potentially how to overcome that?

Yeah, I'd say that historically companies were not very good at collecting this stuff in real time. So 30 years ago, when people started wanting to do personalized marketing, they didn't have in real time, oh, where did the person click on the website? What segment? That was like something that maybe once a day you got that data. Now most companies have what I would call an eventing infrastructure, meaning they're either using CDP, a customer data platform like Segment or, you know, mParticle, or they've built their own eventing infrastructure to be able to in real time pass the things that are happening in email, things that are happening in the website into marketing tools. Most companies have that. A lot of companies also take all that data and every couple hours update Snowflake or some kind of data warehouse with that information if it's something that doesn't need to be real time.

Okay.

That's the starting point, right? If you don't have some kind of real-time eventing infrastructure, you can't do it in real time. It's not possible. Yeah. Um, what we do very well is we make sure in Regal we have the capability of matching your data scheme. That means making sure that in real time we can identify what kind of data you have. Uh, means we have custom objects so that if you have sort of unusual, uh, custom objects for your data, we can handle that. So there's all kinds of things we've done to mirror people's data structure so that no matter who you are, we can ingest the data and make sure we can use it.

That's quite cool. But backend, what type of database are you guys using?

So I mean, I'd say the important pieces, if you think about it, are that we've created sort of the ability for real-time data streaming and then the ability for sort of this data in real time to be sort of entered into our system. So in CRMs historically, the way it worked is you had to predefine fields. You said there's first name, last name, date of birth, you know, free text field. You predefined it and you sent data And Regally, you don't have to do that. You can just send us data. We automatically identify that it's a date, create that field, and start creating the data in our system, which allows you to be much more flexible. As long as the piece of data you're sending us has some identifier, you know, email address, user ID, phone number, we automatically append it to that customer's profile, meaning in real time then we can then use that data to do something. So it makes it much faster to be able to use that data to to, you know, do something for the customer.

Okay, so it's just like API-driven integration where people just make a call, pass the data, you have some mappings, data comes in, and then your system is just capable to be worked on.

Yeah, but that's a big step forward. So even, you know, a lot of advanced companies, you know, for a long time didn't have real-time data streaming, or if they did, they only had it for the 3 fields they identified, not for everything. So yeah, this sort of new world, this explosion of data is a relatively modern data stack.

Okay, so it is, it is definitely the maturity piece to all of that. You know, you need to be a certain size, have certain people. It's not something like you can just wish for.

You can do versions. Again, there are products that are like SMB versions of all of this. Are they as sophisticated? No, but if you're an SMB, you probably shouldn't be using it. As much sophistication. So in home services, as an example, there's a great product called Hatch. I think it's like Hatch app, which is like, if you're a, you know, if you're a 2 or 3 person, you know, local services provider and you want something that gets a bit of customer information and lets you text them and call them, and you want to— it tells you kind of, do you want to do it on day 1, day 2, day 3? It kind of does that. Now, if you're a bigger home services provider doing hundreds of millions of dollars a year, that's insane to be using that SMB product. Yeah. You're leaving so much money on the table. And so they use a system like Regal, which allows them to figure out, well, we wanna call at 9:00 AM when the person gets to work, and then we wanna text with exactly this text based on what we said. Like, there's just more sophistication that leads to better outcomes.

Okay.

And at the bigger volumes, a 1% improvement is worth real dollars. You know, if you're a 2-person shop, a 1% improvement doesn't matter.

Yep. So what's the, like, if we share some real, use cases, real success, real inspiring stories when it comes to return on investment of the practice of such that Regal delivers as a product. But, you know, if we just think about the practice itself, what can people expect really to go from, I don't have this, but, you know, I think I'm ready, I'm gonna get this done. Is there a way of thinking of how can you calculate the, you know, the return on investment of the whole operation? Or just maybe if you could share some success stories of happened, you know, we had the client, they came in and, you know, after a year or so, that's what happened.

Yeah, of course. So first of all, it's much faster than that. Our average implementation is anywhere from 4 to 8 weeks, which is quite fast in this industry. You know, we replace the whole, what's called CCaaS software, contact center software. So all the contact center moves on to the new system, you know, integrate with the marketing tools, we make sure we're getting all the data, we set up all these new journeys, we call them. I'd say on average, and, you know, across industries, then I can give specific examples, we lift, you know, revenue or customer outcomes or 25%, 30%. But obviously there are cases that are wild and much better than that. So, you know, we have a, a good relationship with AAA, which is a big, you know, roadside assistance provider in the United States. It's effectively an insurance business, and they care, they care deeply about their members, right? And so One of the things we've been able to do with them is lift their answer rates very significantly. So when they're trying to reach customers, you know, instead of the customers like not answering, now they answer, you know, at a much higher rate. Or, you know, we have a customer, um, you know, in the education space that serves, again, Harvard and Georgetown, all these schools, and they've been able to massively improve the admissions process for those schools and have much higher conversion on the phone when they're actually trying to get students to you know, commit to these schools. Or, you know, we have a healthcare customer who does a lot of work to make sure that people are basically not getting diabetes. And what they do is they pass all of that person's health information to Regal, including their latest, you know, blood test. And depending on what the blood test is, it has very different engagement models. So basically, if you're okay, you don't need anything. If something is a little off, like, hey, you get a text that, you know, maybe some food. If something is really off, literally a nurse is calling you, getting you an ambulance to the hospital. And so like in real time, it's using medical data to make sure that you have a much better patient outcome. That's not about revenue, but it's about improved patient outcomes, for instance.

Okay. Revenue follows, I guess.

Well, in that business, if you have good patient outcomes, yes, the insurance companies pay you, but they're certainly not, not optimizing for revenue in that business, right? They're optimizing for that person didn't get their foot cut off because we made sure we got ahead of it and got them in the hospital.

Yep, true to that. It's a very broad topic. So we started sort of, you know, just sort of scratching the surface here, but I'm like, we could probably talk for hours here, Alex. So the next question would be like, if people want to just learn more in general about, you know, customer intent, personalization, optimizing these journeys, you know, maybe they need a little bit more education before they start buying in the tools or thinking about strategically putting that into their technology strategies, you know, as CTOs to actually get this done. It's like, do you have your own go-to resources? Maybe books, maybe, you know, people would you recommend? It's like, if people just want to learn more generically about the topic, like, where—

what would—

what would be your recommendations?

Yeah, so there's a lot written about marketing related to this stuff. So that's always where I'd start, is starting to read sort of marketing books about how to use personalization marketing. There's really very little written about how to do this stuff in contact centers. It's related to all the marketing content, but, you know, if you're a, you know, call it a company with, you know, a sort of middle, mid-market or enterprise size in a sort of higher consideration product that where you have a dollar value lifetime customer value of over a few hundred dollars, we're always happy to chat. Like I used to run a few thousand person contact center like in the home service industry. So I love this area. I know a lot about it and we're always happy to talk with customers and help. You know, they don't need to buy today. Like we're happy to be just a trusted advisor.

Okay. So if people want to just learn more about Regal or potentially connect with you, Alex, and just, you know, pop some questions, understand the context and and applicability of it, how would they get in touch with you?

Oh, you can always go to regal.io or email me at hello@regal.io.

We'll have the links, so we'll add the resources in the show notes. So before we sort of, you know, finish the show, we always ask people to, you know, leave them with last wise words which connects with the topic. So from your perspective, Alex, what would that be?

So the big thing now is, I'd say, the shift to AI agents in the contact center. And two themes I'd leave you with there. One is, I would bet 5 years from now, 10 years from now, the majority of interactions in the contact center will be done autonomously by AI agents. And that experience will actually be better for customers and it'll be a lower cost. So that shift is gonna happen really fast, you know, get ready for it, you know, it's happening. Two is, there's an interesting side effect of that, which is actually that you'll be able to do even more personalization than you could do with humans. In our experience, there's a lot of examples where we know a lot about the customer and could do more personalization, but it's very hard for the human agent to ingest everything we know about the customer in real time, say, hey, hi Mike, I said, you know, blah blah blah, you know, whatever, because they're doing all these calls. An AI agent doesn't have that problem, right? An AI agent can use everything that you throw at them in that moment to have the right opening, the right accent, the right, you know, hey, you like the Yankees or the Red Sox? Like, all of that can be right. Oh, we know last time you spoke with Bob about this specific thing and you're concerned about that, without any delay. So there's actually going to be an— not only are AI agents very high quality now, but they're literally going to know more than humans and be available at the right time and actually provide a better experience for the customer. So I'd say that's the big transition we're making with brands now. Most of our customers, I'd say, 10 or 20% of their calls are now AI agents, the calls that are easier and more formulaic. But over the next 5 or 10 years, that's going to become, like I say, the majority of them.

Do you think there's a psychological adoption from customer's perspective to understand that you're actually not talking with a human being and accept that as a— this is where the world is right now?

I think there's an important sort of barrier to overcome, which is, is this good enough, right? I call— I'll pick on Amtrak. I call Amtrak and they have some— I don't know if it's AI or like just whatever automated voice or some kind of bot. It's terrible. Like, it doesn't know who I am. I say something, it doesn't really listen, it sends me in the wrong direction. I never want to interact with that. On the other hand, if, you know, I call some of our customers, so you know, we use our agents. I mean, it is shockingly good. We have— Regal has gotten very, very good at this. Now, I'm not saying we can handle everything, but, you know, we can handle the qualification, the scheduling of a next call, the routing of the right team, all these things really natively at the time the customer wants with a lot more personalization. So I don't know, as a customer, that's great, right? If I get what I want out of it and I get it faster, I'm much happier. Um, so it's really about like, do you meet that threshold that it's good enough? And if it is, yeah, customers are gonna be very happy. So right now I'd say talking to a bot, a deterministic bot, is considered a worse experience than a human. Very quickly, talking to a generative AI agent is going to be considered a better experience than a human. So it'll be fascinating to see actually people going, oh, well, why don't you use AI agents? Like, that would be better for me than your current agents who don't know anything about what's going on.

Okay, so effectively what we're saying is that there's a threshold of the value which the customer gets, because currently the value is not as good as interacting with a real person. People tend to shy away and say, you know, I hate the bots, I hate the help support, which is AI agent driven. But once that value actually overpasses what you get from human, it's going to be just faster, they'll know more and be able to help better. People will just accept that that's actually a better outcome for them.

Yeah, I mean, not just accept, they'll demand it, right? So like, you know, in the same way today people go on Twitter and they go, I'm so pissed off at X Airline because they sent me to a chatbot instead of a human when I was trying to do something that the chatbot couldn't help with. It's going to be the opposite. They're going to go, I'm so pissed off that X Airline sent me to a human instead of an AI agent because the AI agent would have gotten to my issue faster. It's gonna flip really fast.

Okay, so we're saying that that flip is coming in like 5, 8 years. I think that that's why you mentioned it.

What's gonna happen is, you know, in mid-market it's gonna happen much faster 'cause they have less regulation and less sort of overhead. And so they're gonna, in the next 2 or 3 years, you know, significantly switch to using AI agents. Yeah. In enterprise, it'll take longer because they have more infrastructure and more things doing the old way.

Yeah.

So actually I think it'll be an interesting period where mid-market companies have an advantage over enterprise because they're going to adopt this technology faster.

Okay, so effectively the smaller the business, you know, the more easiness they have to be early adopters of this, and as a result they just play with that and get to the point where, you know, it's a good feature. Let's call it like this.

And even if I look back at, you know, building websites, or if I look back at when people started using apps on phones, you know, So yes, sure, there were a couple of big brands that tried getting into it early, but really the innovation wasn't happening in the big brands. It was largely happening in these smaller players. And you saw some huge success stories because they got it right. The first small companies to get websites right, right? The first small companies to get, you know, like Amazon was a small company, right, with a new website. You know, the first companies to get apps right and build games and things that really were native into, you know, the app. Really became huge success stories. So we're gonna see the same thing in AI where there are some mid-market companies that get AI agents right and their business is just gonna take off much faster because of that.

When you say that, Alex, do you have an industry in mind? Are we saying like, yeah, like you in industry you operating in, which is effectively, you know, the support, the call centers. Yeah. How would we call this? Like a helping industry?

Like, Well, yeah, so I think it's, you know, depends how you look at this. So a lot of AI, you know, there's a lot of AI agents going into, or AI assist, or, you know, not just AI agents going into like engineering. That's unlikely to wholesale change the industry, but it could change the cost structure, which means somebody could price differently and potentially win on price. Versus like in what I'm talking about, in how the contact centers engage with customers, it could wholesale change the way you interact with a brand, right? So, you know, imagine an insurance company where everything was the best possible experience as opposed to being the worst possible experience. Or imagine, I'll pick on Comcast. Comcast is in the United States a big internet provider, right, which is historically seen as like a very bad customer experience. They've invested a lot in deterministic bots. Maybe they'll make the shift to AI agents and that'll completely change people's perception of them as a brand, leading to much better outcomes for customers and much better customer retention. So I think, you know, we're gonna have to see how it works. The other thing people are betting on is completely different interfaces for customers. So in the same way, like a website was a different interface and a mobile app was a different interface, you know, people are now looking at chatting with an AI agent, but it may also be voice with an AI agent.

Yeah.

As a different interface to actually be introduced to a product, buy that product, use the product, you know. So that could completely change how moving is done or how I don't know, buying a car is done if the interface changes from a website to this new interface.

Yeah, that's definitely coming. We did unpack this topic in some of our other episodes of, you know, what the future with AI looks like. And, you know, that future, to be honest, is like 5, you know, 10 years away. Just for an intro.

The trick, I think the hardest part is going to be what is a new paradigm in terms of interaction. So like my little kids like Alexa, the, you know, Alexa, Speaker. And, you know, the 6-year-old understands most of the commands, the 2-year-old not yet. But they don't— they are frustrated a lot because there are things they're asking of Alexa, both of them, that Alexa doesn't do or there isn't a command for. So I think getting to the point where there's some sort of common set of commands and sort of understood way of handling it when it doesn't understand your command will really, I think, drive a lot more value. Like, you know, in an app you have some buttons and it's become relatively standard. But there are still apps where they've gone way beyond what are the normal interfaces. So if you think about maybe TikTok or Snapchat where the interface is quite different in fact and it takes quite a bit of learning to figure out how to use it. So I don't know, we'll see how it works with voice.

So it's almost like I'll come to the website and instead of seeing an e-commerce website and instead of seeing a website, potentially I'm gonna get greeted by you know, a bot, he's going to be like, hey, welcome back, Mike. You know, I know all about you. You probably came here right now to buy some kitchen towels and some milk.

Yeah. And yeah, you're, you know, now you go and it says, hey, do you want to continue the website or go to the app?

Yeah.

It may have a new pop-up which says, do you want to go to the website, the app, or do you want to talk with us?

Yeah.

And you'll just choose, you know, talk, and that'll be the new interface.

Yeah. And then instead of really typing, you can just actually talk like we just talked right now and just you know, do the shopping like you would be almost doing that on a phone call.

Yeah, I can actually— there's still, there's still reason to do texting. I'm not saying texting completely go away. Like, if I'm in a meeting and I— it's helpful to text, not talk. But talking is so much faster than texting, and there's so much more expression that can be used in talking. So I actually think that this direction of like the, like, AI as a chat agent is not that interesting. I think AI as a voice agent is much more interesting as an interface.

It is already there. I mean, I personally have a Bridge app which allows a lot of AI-driven automation takes place. I do tend to type to it, but it has the building capabilities to actually chat with it, which is fairly good. So, you know, if I'm having lunch, you know, I could be like reading an email and I need to put something in my diary. I would just click a hotkey which opens the interface with the app. I would be like, can you add a meeting with John 2 PM on, you know, yes, tomorrow, and it just adds in the Google. Of course, that takes some time to set up, but it's very easy to see how there's going to be markets coming out with these micro agents or whatever pre-baked agent-driven solutions, and people will just, you know, onboard a fleet of agents for different business challenges to just, you know, like, it's absolutely doable. So one question, Alex, What Regal does right now in the backend with AI, do you know from technical perspective if you have your agents on top of the open source like Ollama agents or on top of ChatGPT, or do you actually train your agents from scratch?

Yeah, it's a good question. We basically have decided not to invest in LLMs ourselves. We have integrations to the major LLM providers and some of our customers even bring their own sort of LLM. That's all fine for us. What we've built is largely how the layer of how you use that LLM to build an AI agent, and then how do you take an AI agent and put them into production in a contact center environment and, you know, monitor them and QA them and know what they're doing and see if it's better than the human or not. So we've really focused on, on that side of it. We are not going and spending billions of dollars training LLMs. Like, that's not our business model.

Okay, it's lovely. All right, we're just at the time, so thanks everyone for listening. And, you know, hopefully that was useful from Alex and some observation from my end to understand customer intent. You know, it isn't a trend, it's not a new thing, you know, it was there all the time. And you start effectively with good data capture, personalization, understanding behavior patterns, understanding your customer intents, and then actually being very proactive with what you do with these intents, not reactive in the sense of waiting until the customer reaches out to you. If you need more information, of course, you know, reach to Alex. All the, you know, all the links are in the show notes, just scroll down. And before we close, if you find value in the show, please do leave us a review, a comment in Apple Podcasts, Spotify, CastBox, or wherever you're getting your pods from. And as always, until the next time, stay safe and see you online.