LaunchPod - Brian McMullin
===

[00:00:00] 

Jeff: All right. Brian, welcome to the show, man. Thanks for coming on

Brian: Hey, great to be here. Uh, excited to have the conversation today

Jeff: Everyone likes to talk about, , , the YC technocrati building all sorts of fun AI features for folks like you and I who work in tech. , But I mean, the reality is,, I think most of the consumer population across the country and the world is probably more like someone like my father-in-law, who owns a painting contracting company, and he has an iPhone, but, uh, still iffy on how he does it or not.

He's still, you know, I think I'm his in-house tech support. So, you know, most people are not thinking about prompt engineering and, , writing evals for their AI outputs. And that's what we're gonna talk about today is how you, build, you know, great product for the other 95%.

But before we do that, maybe can you just give us the kinda quick TLDR, like, you're at Network Solutions running product now. How did you get to here and, and what's the background and why are you authoritative on this subject?

Brian: All right. , So yeah, uh, a relatively, typical, uh, product journey for me. I, I started off as a, as a developer, [00:01:00] realized after a few years that I was much more interested in the why and, and the what, and working with customers to find the right solution than, you know, building the best possible, software.

So I used business school as my jump into product management. , Been in product for about 20 years, 15 of that focused , on SMB and, , small business products. , Started with, , with Yahoo, , doing, uh, media products, then spent, , really the, the formative years of, uh, kind of my product leadership journey at, at HubSpot, helping them grow, you know, kind of 10X during that time , from , pre-IPO through to, , a, a very successful public company.

, From there, product leadership roles at places like ezCater, Wayfair, , led product for a creator economy startup, , called SamCart, and then spent the last two years at, at Network Solutions. At Network Solutions, , we like to consider ourselves the OG of the, , online identity presence and domain space, , and it's pretty humbling being, you know, part of the story of the first public domain registrar and all the businesses and websites that it's helped launch over the last 30 [00:02:00] years

Jeff: Yeah. I mean, you, you have done your time at several kinda notable companies there. First, whole tour through almost a who's who of Boston, it companies. Uh

Brian: got a couple more from my, from my, you know, checklist, but we'll, but we'll save those for

Jeff: Life's long. Life's long. , But I mean, now to go, , you know, the first domain registrar is a pretty good one. And I think that all that goes back to, you know, across the board, these were not serving technical customers. They were serving people who had a problem. , You know, ezCater, you just, you just want lunch, man.

Your office is just hungry. , Wayfair is just you're buying, you're buying furniture and stuff. And that kind of brings to a great point. , I am a huge consumer of cloud code and, and early on I used Lovable and all these kind of tools where you have very technical people building, , tools where you need to really, really think about your prompts and how do you get great output and, and all that kind of stuff.

And while the market cap of these things has exploded, , you know, most people buying a domain or buying a couch or, or buying lunch, , AI is great in that you can serve them better, , but they're not thinking about that end of it. how do you [00:03:00] kind of approach that world when the product is just a solution?

They don't care if it's software or not, let alone AI or not. Like saying AI is not going to make them want it more.

Brian: Yeah. It, it's been really fascinating to watch over the last year and, and, and how sentiment is really accelerating even faster than, than, uh, I, I personally was, uh, was anticipating. So, in all of our customer outreach and surveys, you know, what we're seeing is, people are going from kind of a, a sense of almost distrust in tw- 2024 25 to almost, you know, in the SMB space, like one of FOMO, where it's like they're hearing enough about, , how valuable it is, that they should be using it, , that they're missing out if they're ultimately not using AI and agents to do what they do every day, but they don't understand how, and they're really looking for someone to help them kind of bridge that gap between, you know, what they think they should be able to do and, what they really understand , that they're able to do today.

In, in a lot of ways, it sort of feels like the first days of, uh, [00:04:00] of having an Alexa device in your home, right? Where you're not quite sure what it does. You kind of try things out here and there, but it's a little bit of a black box of, and literally a black box of, you know, what should, , this thing be able to do for me?

So, , I think it's , been exciting to see , that transformation because I think there's a lot more openness and willingness to try things and, than the desire to learn and incorporate that , into what, , folks are doing day to day. , I think what's become incredibly clear is that as we're building these types of products, , and this is where I do feel like , we've, vary , pretty drastically from, , attracting a more tech-savvy, , customer target, is the answer isn't just a big empty text box. So

Jeff: But they're so pretty, just having a UI where you click into and, and it can do anything

Brian: gradient behind it and like, you know, your, your valuation's gonna be amazing.

Jeff: Through

Brian: a-a-as much as, you know, we want to put the control into customers' hands, for most of them, they might as well be staring at a [00:05:00] terminal, , window

Jeff: right. Well, I think there's this, there's this , overarching view of if you have these kind of like technical, more experienced, really focused customer set, you know, you're selling to engineers , or even product managers and stuff like that, , the prompt box is fine, right?

'Cause they know what they want out of the thing. , , I would argue, I don't even know if that's true because I've talked to, , tons , of people across multiple, you know, verticals and industries and, , using different software types, and even some of the most experienced forward-thinking AI-pilled product folks I know go into some of the tools they have and they just say, " "I-- It ju- it's just a box.

I don't, I don't know what to do with it." If you have experience with the tool and already a well-opinionated idea on how you use it, I think the box can be great 'cause you like, you have the things you wanna do, now just make it easier. But I do think generally that's not just a problem of, selling to non, techy people, but it goes back, I mean, this is a consumer thing we've seen for a long time, right? Like, more choice is not always good. , E-commerce figured this out a long time ago. You don't want to give them eight choices. They're gonna pick none. , Given that this is not a new problem, it's just a new flavor of the same [00:06:00] problem, what does that mean?

So you can't just put the box on there.

Brian: Yeah, yeah, definitely. Well, I mean, I think you were hitting to it earlier, like, , I've spent 15 years working , with SMBs. I have yet to meet a, a business owner who wakes up in the morning and, and thinks, "I can't wait to use AI today," right? Or, you know, rewind five, 10 years, "I can't wait to use my SaaS, , product today."

Like they're waking up, just they, they need to figure out, "Okay, how am I gonna close, , this deal? How am I going to get paid on this invoice? How am I going to, manage, , this employee situation that I have at this location, you know, of, of my business?" , especially in, in the context that I work in when it's like your online presence and, your marketing and how you're driving sales through all those different things, , they have very specific things that they're trying to accomplish there and especially , for SMB, they have an incredibly small amount of time to, to do that.

, Looking across, you know, the restaurants I worked with at Ezcater or, , the creators that I work with at SamCart, like they have like a couple hours a week where they do their marketing stuff or they do their kind of their, their business [00:07:00] stuff, you know, that is taking them away from actually doing the things they are great at, that they love doing, which is serving their customers and whatever their business ultimately does.

So the tools need to be, things that, I mean, at the end of the day, they are tools. , The AI and the agents and everything that , we're building for them ultimately are there as a means to an end to help them achieve the thing that they're trying to achieve, , the thing that they hired , our product and our business to actually do for them.

know, AI is really just the next transformative technology, , it's, how do we, bridge the gap between what they ultimately wanna do or maybe even don't know they need to do in a lot of cases to doing that thing for them.

The way that we've been approaching it or the way that we-- I've seen it to be most effective is really think about that through the lens of, of personalization, , providing context and then ultimately taking very kinda educating actions. I do believe, you know, especially for this audience that, , we're going to be most successful where the, the AI is effectively [00:08:00] opaque, that we show them the value versus the prompt.

They're less interested in, you know, the amazing technology that lives behind the scenes and more what this can do for them, how it can, I think, initially save them time and money and then ultimately help them grow when it comes to the things that it can do to drive new revenue for their business And what I think that looks like in, in practice and what we're really starting to experiment with, , with our products , is educating through the experience itself.

So again, rather than showing them the prompt box or even rather than showing them, "Hey, we have these agents, you know, come use a, a marketing agent or a sales agent or, front desk agent for your website," I think we need to take one or two steps further and say, "Wow, we've used this agent to, map out your next two weeks of social media posts.

Go, you know, schedule them, and then if you want us to keep doing this for you every two weeks, , turn it on right here. Or [00:09:00] if you want us to go update your website, for the holiday season that's coming up with, , the, promotions that you were already planning, like we've already done it.

Come and claim it or turn it on." So how can we proactively show them the value that these tools can, can bring to them in a way that it, it feels more kind of predictive and personalized to them, , and in some ways almost be, be that gateway drug , into AI agents and all that. Show them what's possible so that the next time they do see, , a big open text box, that they have a much better idea of the types of things that they can do in order to,, get the most out of, of those capabilities.

Jeff: This is not a new concept, right? It's make the action that you desire them to take or you think will be best for them, the easiest action for them to take. people don't want another tool to go dig around in. They just want the outcome. the same way we look at, you know, we, we do, , you know, session replay and digital analytics and [00:10:00] part of how we've looked at it is, okay, we built the AI agents that can watch all the sessions for you and can tell you what's important , and surface the little bit of signal from a lot of noise.

If it was just, "Okay, now come into the tool , and put on your filters and ask us what it is," only so many people are gonna use it. We've seen astronomical growth though in like sessions analyzed and sessions watched and, and all the kind of issues and takeaway stuff because a lot of that just gets pushed to the user.

, We just automatically tell you, "Here's what's important. You should check it out," or, " here's what you should do about it and here's how you can connect it to just auto code it." , when we started to make that switch, we saw a lot bigger uptick. So like these lessons are, , if you're selling to a techie audience or, or you know, my, my Uncle Steve.

Brian: Yeah, yeah, exactly. But I- I think what's interesting too is, , that can fall flat if it doesn't feel personal. And , if you push something as generic where it's, it's less, "Hey, I'm gonna move you to this specific time zone," I mean, it could maybe a bad example versus, "Hey, we're gonna set a time zone."

And I think, , the more personal and direct it can be for a specific [00:11:00] business's needs, the more adoption and interaction and engagement you're gonna get. And I think the, the tricky part there is what, what you need to drive that level of personalization is ultimately context. And I think this is also where, software, , companies that are investing , in these tools can ultimately win and create differentiation is in building that context, not only within their, kind of domain or, or kind of business space, but with that specific customer. , The more that you understand about them, their business, , their behaviors and habits , and everything that you can glean from the things that they're already doing with you, the more personal each of those recommendations or preconceived actions can be, , which will kind of pull them further down the usage pipeline.

Jeff: Right.

, So I, I think the other problem comes up, and I'm, I'm curious how you guys think. So like I get the personalization just kind of the easiest path needs to be the path you want. But, , I keep coming back to, you know, monetization kind of comes up and, and you can package , all these AI functions in, and sure it's easy if it's just, it's just now included in what you do.

There's no cost associated, you know, whatever. [00:12:00] If you can roll it into your subscription, wonderful. but at the same time, , the value of software used to be that the opportunity cost of serving, , n plus one customer was, was very low. But now that we pay for inference, that's no longer true.

- I can explain tokens, , and consumption-based, , billing to product managers and engineers , and even marketers all I want, and that'll, that'll go over pretty well. You try and tell a restaurant owner about token-based billing, I feel like it, it might be a little tougher sell. How do you kind of look at this piece?

'Cause that's a, you know, you can build the best thing in the world. If, if it's not the right cost structure, it can still fail.

Brian: Yeah, no, definitely., Not the right cost structure and if it's, , confusing and difficult , for the customer to conceptualize,, it's gonna fail as well. I mean, like as you're describing, , the relationship between prompts , and skills and tasks and what model you're using and tokens is a complicated one for us to even talk about.

, Not to mention, a restaurant owner or, , somebody, working at the front desk at a law firm trying to kind of manage that. ,, I think , we need to figure out a way to, to ultimately [00:13:00] simplify that , for our small, , business customers and, find the way to, to map the value that's being created, through these tools, to, you know, the amount of value that's being exchanged, for that, , when it comes to, you know, however you're ultimately trying, trying to monetize that.

It's an especially hard sell, to your point, because up until this point in time, the prevailing SaaS model, , they're paying a certain amount of money to get value out of your platform, and to tell them all of a sudden now they have to pay more to, you know, arguably get that same value is just a really difficult thing , for them to, to understand and for, for any of us to really kind of conceptualize and say, "Yes, I, I'm willing to, to ultimately do this."

So it's interesting and it's actually a- almost a little bit of a warning of like how deeply to embed AI into your core product , and value proposition. , I think until , we get to like more stability and token pricing , and understanding and predictability of what that looks [00:14:00] like, the more you bake it into, you know, the, the core of what you do, the more you're gonna have to, you figure out, you know, how you, manage that and have a predictable, , cost structure , for your customers.

Jeff: , that's an interesting point you kind of bring up is it feels like this is probably easier where if you're providing domain registration, , you can probably still do that without AI. The core, you know, dependency there of what, what, you know, group of people are just gonna come to you, you're not adding inference cost to every single one of those.

So that, that's great. , But yeah, if you are looking at, you know, this is our core product and now we're building inference into just the basic building blocks of it, it makes sense in that, yeah, you wanna be AI native and, and bring it into the core experience. But at the same time, if you're not, if it's not enhancing, you're just adding basically, something that's dragging your margin down, gonna cost more, and you have to either re-explain pricing to people who don't know or don't care about it, or , at worst, you're gonna feel like they're just getting ripped off.

You know, it's another thing inflating. " How, how have you guys looked at it there? I'm

Brian: Yeah. So , in all the customer outreach and surveying that we've done, I guess , , not surprisingly, , we're not seeing a whole lot of [00:15:00] consensus. I mean, I think it's sort of, across the board and, and mixed between, , do people want more control but potentially more cost or less control and, and less cost?

, Do people want to see things, you know, as pure token usage or abstracted into credits or even per task, type , of model , of mapping, you know, the cost to the value it's being driven? , It would be really great if there was a prevailing opinion there, but we're not seeing one arise yet.

I think we're just early. , So I think ultimately what we're doing is we're, we're trying to experiment into, a few of the different models, and I think the, the two that we're, we're looking at right now that are, pretty common in the market space is, you know, tiered usage models.

So, , I think for SMB trying to move to a, a true usage model of, you know, hey, you're paying for a certain number of tokens or increasing it and decreasing it every month , is gonna be really difficult to conceptualize. So we're starting with, "Hey, we're gonna give you a certain number for free to really get you in there [00:16:00] and, and trying things and see the value of the, the capabilities that we have that use these, tokens," and then give you the ability then to say, "Yes, you know, I've hit the threshold.

I want this. I want more of it. Uh, let me start paying for some higher, , threshold or, or tier of usage that then I can operate in, , on a month-to-month basis."

The value that we're seeing and anticipating there is that, , I think it's gonna allow for a, a lot more of a true kind of freemium to PLG type of motion and an introduction into what agents and these types of skills , can do for you as a business. It will allow customers to really try before they buy, , versus having them to, emotionally and kinda mentally elect to use an agent or a, an AI product and go through a paywall.

Jeff: curious to see, because up until very recently, for the most part, SaaS your valuation was very heavily dependent , on your ARR, right? Because across the board, it's very common, you signed maybe monthly contracts or annual contracts as a c- as [00:17:00] a company, there was probably no way that mid-contract someone was going to really, really spike for one month and come down, but you had them for that year at least, and you could predict churn and, and all that kind of stuff.

I remember 'cause this came up years ago. I was at a different company before LogRocket, and the way we billed was, we would help you manage your AWS cloud spend or, or your, your Azure, whatever cloud spend. And so it was whatever you ran through our system, we took a percentage of whatever you paid the cloud provider.

which was great 'cause, this was ZRRP. Everyone's cloud spend was going up , and it was a wonderful business model. And as everyone grew, your cloud spend grew because you served more people, and we had, you know, this kind of built-in growth mechanism for all of our customers. , But the downside was, sure, you signed on for a three-year contract.

If you just stopped routing it through us, though, your three-year contract was three years of you paying us nothing,

, so there was no - predictability to it, and I remember this kinda coming up once or twice talking to, people in the capital markets and stuff of, "Well, you're not really ARR, so can we put the same multiple on it?" And in the end, we actually had a, you know, great exit. [00:18:00] But it feels like back to this again a little bit, where, okay, you have a customer paying you, you know, maybe in your guys' case, right? You have people pay for domain registration.

, The agent and the token cost and stuff, you can pay for rights to it or, or, you know, access to it. But if you're paying token-wise, you might have one month , where we have, you know, I have a ton of stuff I need to do, and next month I have nothing I need to do.

So the predictability of this is, is greatly gonna decrease. And I do wonder how that affects how we value businesses, how we look at success of, of our own efforts.

Brian: Yeah. Uh, that's where I, I, I, I do wonder if more of a tiered kind of limit model is, is going to be, going to be useful there. It, uh,, It provides that level of predictability where, say, you get credits a month or whatever and just that re-ups every month. I think the, the concern there is that it, it is more difficult for, customers to kind of conceptualize when it comes to, you know, SaaS product usage.

Like, why do I need to do this? Or [00:19:00] do they, they start viewing it like Planet Fitness where it's like, oh man, I'm paying for a gym that I'm not going to. Like, I bought this when I had a bunch of spend that I needed to make 'cause I was doing a website redesign that month and now I'm on the hook, you know, for, 10 bucks a month, 20 bucks a month for a bunch of credits that I'm not using , creating that kind of, , anxiety around,, , token usage over cost.

There's definitely something there and I think if anything we're, we're, we're all trying to figure it out. the other model is the more natural one , for a SaaS market where it's just, you know, a paywalled, , premium type of, of, of agent or, or app model where I'm just going to pay you 30 bucks a month for this, agent that's going to do a very discreet thing and maybe there's some very high, , token limit that you have within there.

But the expectation is I'm just paying for this agent and it's going to do what I expect it to do month over month and really the software provider is, , the one that's mostly on the hook for token usage and, and controlling that margin. What's interesting about that is I, I think that one is the one that is going to be [00:20:00] easier for SMB to understand because it's a more direct, you know, pay for value.

But I think it's also going to be the one that SMBs are more resistant to adopting because if they don't necessarily truly understand the value of the agent, then they're not gonna go through a paywall in order to actually get there in the first place. So i- in a lot of ways we're, we're experimenting with both here because I think that the market right now is trying to figure out what the right model is

Jeff: , Yeah, I don't think anyone has the answer here. , I mean, we're all still arguing about, , how much we pay, , Anthropic for Claude access. So if they, I mean, if they haven't figured it out, I don't know what hopes do the rest of us have. But yeah, I, I think I have to agree. Like, there, there's gotta be some level of predictability.

probably gonna end up somewhat friendly for customers, somewhat friendly for us, the SaaS providers, or not us. but, you know, at the same time, this conversation literally just reminded me I need to go throttle down my Lovable subscription quite heavily because I had a couple months where I was super high and I kind of moved away from building there.

, But, in between you want some access, you want some regular subscription, but at the same [00:21:00] time, like once you find value in it, , you're much, much more willing to go, "Okay, I need to do 10X more this month than any other month. Just take my money. I know the value, just, just have it."

Brian: totally agree. And I think it also brings up a, a really interesting point and something that we've been discussing a lot is, , I think especially a- as businesses or a- as, software products, , move into this space, the conversation about monetization , is a, a , very important one.

, But at the end of the day, it, it isn't the most critical when you're getting started. The biggest risk you have is that nobody actually uses the product in the first place. So I, I... There, there is a, a temptation to, to build complex, you know, monetization models and put in tons of limits and, and over-engineer that aspect for it before you've truly found product market fit with the, the product that you've, you've created.

So I think alongside with our experimentation, we're actually, like, trying to, you know, back off a little bit on enforcing a lot of things early on while we are, , [00:22:00] gauging interest, educating our customers, and ultimately getting them to, to use the new AI and agent capabilities that we're delivering.

Jeff: it comes full circle to, I think what the, one of the earliest things you said today so far. Just make the thing you want them to do the default easiest choice to make.

Brian: Yep.

Jeff: It just gets com- we, we make it complex and implementation's hard, but the idea is easy. to that, to that point, I think there's, there's another aspect to this is like you can get people doing all this stuff. , You can push them, you know, make it the easiest path to go, , any of that.

this has come up several times recently and, and I wanna hear your take on it 'cause I think it, it kind of matters across the board. The AI wars and, and all the, you know, SaaSpocalypse and all the fun wordplay going on right now. But in reality, we're solving problems, and we wanna do it in a way that, the choice we want them to make is the easiest, most default choice they can make.

You can add features, you can do all that, and that's wonderful. The other side of it is churn has never been less far away for an aggregate group of users, I feel like, than it [00:23:00] is now. you need to serve them the, features and outcomes that they want, but at the same time you have to not annoy the crap out of them with the things that you're providing.

Like, it has to do the thing they want to do. It has to be nice and easy someone put it really well recently who I talked to. Churn is not a one-time event. It can be a collection of kind of micro grievances , and micro, annoyances. And when you get all those things, there's usually one event then that tips them over.

It's the same as like if you wanna retain your great employees, it's not keep them from taking the new job. It's your, my goal has always been, how do I stop them from taking that first phone call from a recruiter? , how have you made that work or how do you focus on that? 'Cause I think it's an interesting topic right now.

Brian: No, I mean, I think it's a fil- familiar one to probably everybody who's listening to this that's ever had to do a roadmap review with a leadership team, right? Like the, like you, you have all of the, the really exciting new stuff that, that you're, that you're building that's either strategically important or, you know, very directly revenue driving.

Jeff: sexy. You wanna, you want to put out the new [00:24:00] feature. Like, everyone wants to be the hero

Brian: E- exactly. And then you get to the point where it's like, "Oh, and we need to make user management work," or, "We need to make sure that the login page is doing all of these things that it needs to do for, you know, compliance reasons," or, "Hey, the homepage needs an update because, you know, it's, it's really not, driving people to the right places , within the application," or, "Our CS team , is really hitting us because this part of, the product, you know, customers are calling in about it all the time."

it can be really hard to, have those conversations with your leadership team where they're like, "Okay, well, like, show me the revenue. Like, what is the ROI of this? You know, we have financial targets, , how does this drive direct revenue?" It's difficult. Um, I think, uh, it's not like, uh, if, if I build it, you know, I will make, you know, 20K in, in MRR or anything along that line.

But, , to your point, these things add up over time. the more that, the customer struggles to do basic things within your product experience, , the less likely they are to even try out the new things that they might have bought for. I think, [00:25:00] experience I had, uh, at SamCart, you know, we were building really exciting new capabilities, , that allowed creators to,, monetize , their audience.

great converting checkout experiences with a, a bunch of tools to be able to really maximize , the amount of revenue capture from each of those interactions. And we had a marketing team that was, you know, second to none at talking about that and acquiring customers based on the new features that we were driving.

, But then it turns out that,, when we look at the churn reasons, none of it was about these new features not driving them revenue. It was about, "I can't get this basic thing to work. I need to add my team and I can't do that really well. I, I don't understand. I, I don't have visibility into how these things are performing.

I can't connect my, my payment processor in, in the way that I want to and, you know, I'm not able to take payments." It was those things that were ultimately, you know, driving the, the churn in the equation, even ahead of them adopting, , those new features. I think that's also just another thing, to keep in mind as we're, as we're building all these great new things.

, , a lot of [00:26:00] times, especially within customer acquisition, people buy for the optionality to have the capabilities that they're buying for, not necessarily to directly use them. , But the number of times that I've seen, you know, customer cohorts come in that, you know, it's clear that they bought on a specific marketing campaign, you know, targeted at a feature, and they never actually use that feature and they stick around.

It's, it's more just the concept of like, "I, I wanna know that I have that in my back pocket if and when I wanna use it, but you know, the core of what this product does is really enough for me right now." So- making sure that those core things ultimately work, you know, i- in, in that case was our biggest lever to actually, you know, addressing, you know, churn and, uh, and retention.

I think getting back to your kind of o- original question of like, well, how do you have that conversation? Like, when the CFO's looking at you and saying, "Okay, show me, show me the money here."

Jeff: Why are we spending - a third of our development budget on, on this thing that you're doing that is not a new feature

Brian: I think you need, you need to, to put it in, in their language in, in a lot [00:27:00] of ways. Like, how do all of these things map to, you know, 1% , of retention? and, and ultimately backing out,, further from that because retention is such a lagging indicator. Go figure, you know, in just about every SaaS company that I've, I've worked for, , that I've worked at, there's always one thing, that, , drives retention and, uh, and customer, customer engagement.

It's frequency and breadth of product usage. Like, can we get people to actually use the product and use it in the way that we know creates value? And then if so, that is, you know, incredibly predictive of them sticking around with us. So looking at that, building, building a model where it shows, okay, here are where we feel the gaps are, here's how we're seeing that map to, to customer retention, and these investments are really meant to drive those things that we can show you on a monthly basis that they're, they're moving, you know, that they're moving them up.

Um, but, uh, uh, I think needing to kind of put it into those types of terms, I think is, uh, is [00:28:00] incredibly important. Kind of back to this concept of, you know, this type of work is, is never really done. You know, I think, you know, you get that question of like, "When are we gonna be done with working on the customer experience?"

You're never really done. You will finish up, , discrete features within those areas, but this is a, a consistent level of investment that you're gonna make towards making sure that, uh, you don't hit that first, uh, issue that's gonna start the cascade of, "Okay, now I'm getting to the straw that, that broke the camel's back and I'm, I'm looking for, for a new alternative."

, And, and as you mentioned earlier, given the cost of acquiring customers nowadays, like keeping a customer and, and spending, money against that is so much more, , productive , and efficient than having to go out and, acquire new 

Jeff: exactly. Well, I, I am, uh, you know, kinda gonna be honest, I'm sad to say, , churn happens in all, all places, right, Brian? , It happens in your, in your SaaS product and, and people leave. It [00:29:00] also happens-- We have 100% churn rate on recording this podcast that at some point every guest churns off of talking to me and leaves.

, And, you know, as much as I'm sure I hope you're having a good time, I do understand you have, , a whole business that you need to help run. But it's been fun, man. I appreciate you coming on. This is really, you know, interesting to see how you guys are thinking about building, , AI products when, the cool sexiness isn't saying it has AI, but it's just delivering outcomes.

and, and, how do you do the unsexy stuff that actually pays the bills, . So yeah, thanks for coming on. , I hope you can come on again at some point, and we can catch up and, and see how the world has changed.

I mean, maybe, maybe AI will be cool for restaurant owners down the road.

Brian: No, definitely. Well, it's been a pleasure, Jeff. I, I think, I think AI is becoming cool. I'm so excited to see even over the next three to six months, you know, how things, how things evolve. , we might be having a very - different conversation in six months,

Jeff: Yeah. So we'll have to see. , That said, if people want to reach out and ask anything or, or follow up on the conversation, is LinkedIn the best place to do it, or

Brian: Yeah, please do.

Jeff: Perfect. All right. Well, Brian, thank you so much, man. It [00:30:00] was a pleasure, and, , look forward to talking again.

Brian: I like rice. Thank you

Jeff: Bye.