RICK (00:01.058)
What's up this week, Tyler?

Tyler King (00:03.218)
Dude, I'm discombobulated. It's a Friday afternoon. No, I don't mind it. It's just like, it's breaking the routine and it just doesn't feel like podcast time to me.

RICK (00:06.803)
yeah, sorry about that. We were supposed to record on...

RICK (00:14.828)
Yeah, I'm way too energetic and not sleepy. Usually Thursday morning, I'm like groggy. It's been a long week. It's 8am. Now it's like noon on Friday. Yeah.

Tyler King (00:18.919)
Ha

Tyler King (00:23.73)
Yeah, I'm drinking my second Dr. Pepper of the day instead of normally my first Dr. Pepper of the day.

RICK (00:31.414)
Yeah, yeah, third coffee versus second for me. It's bad.

Tyler King (00:33.97)
In terms of how things are going, I'm good. Pretty even keel. Things are feeling normal and stable and all that. I've got various things to talk about, but what's up with you?

RICK (00:45.646)
We usually record from your office location and I believe you're working from home. Is that a normal Friday for you?

Tyler King (00:51.11)
Yeah, Monday, Tuesday, Friday, I'm at home. So you normally get me, which is better for variety of reasons, one of which I think this, gets removed by like automatic AI post-production stuff, but there's like a loud fan running in our whole office all the time. So you hear a bunch of background noise. So this time you get a little reprieve from it.

RICK (01:11.488)
Mm hmm. All right. Yeah, it's nice. It's nice. I just got back from San Francisco. I was in there from Monday through Thursday for our revenue kickoff, which is the annual sort of sales and go to market event to kick off the year. It's a lot of work in the month of February to pull that off. And it's a lot of it's very intense for four days of, you know, sort of like the day before.

two days of execution and the day after follow up. I...

Tyler King (01:44.402)
This is like a Rally the Troops rah rah type of thing.

RICK (01:47.372)
Yeah, we, we, we have an annual theme that we rolled out. share our North star goals for the year. we break it down into key themes and bets. and ultimately that permeates the various sessions, which we have sort of global good to market sessions around vision and the bats. And then we break into like divisions for, you know, certain content and then even into smaller teams for very specific content. That's more tactical. but everything you, we cover everything from like, you know, here are our goals to.

new demo tracks for the software demo to, new sort of strategies for onboarding, you know, upsell renewal, that kind of stuff.

Tyler King (02:27.762)
Cool, do you like it? I mean that sounds boring as hell to me, but do you like it?

RICK (02:31.406)
Well, it's not one of it's not one of those things you can love or hate. It's like it's it's it's just the cadence. You know, it's the operational cadence. And so you got to do it. And it.

Tyler King (02:42.29)
You just got to do it. Yeah. I mean, that makes sense. We do the six month presentation every six months and we've got one of those coming up in two months and I'm sitting here like, I really don't need this. Like I'd rather skip it, but I guess it's good to do with some frequency, kind of get everyone on the same page.

RICK (03:01.934)
Yeah, it's the it's the rhythm get on the same page. What are we gonna stop doing? What are we gonna start doing? What are we gonna keep doing? Where do we need to have more urgency? I mean, AI is a big topic. I a lot of time talking about AI both externally and internally, as well as our product team. So it's pretty fun. And then know, you do social stuff at night, and I'm remote. So like, it's nice to get in person and then go have go have, you know, some

Tyler King (03:13.798)
Yeah, of course.

Tyler King (03:26.78)
Yeah.

RICK (03:30.08)
alcoholic adult beverages with your co workers that you don't get to see every day.

Tyler King (03:32.487)
You

Yeah, nice. Cool. All right, well, I guess you want to just dive into some updates here? You got anything else? Yeah, I thought maybe I'd start with some non-AI stuff since I just can't imagine how bored certain people must be of listening to it, but just normal business updates. The two big things we've been working on at Less Annoying are both soft launching now-ish.

RICK (03:42.85)
Yeah, like what's on your mind?

Tyler King (04:04.954)
intentionally and one unintentionally. So the intentional one is Kanban boards. We've been working on this for since maybe midway through last year. We've had a pretty decent prototype for months now, but as I've said before in the podcast, it's just the devil's in the details here. It's so many little edge cases to handle and stuff like that. There are still a good number of those to work on, but I think it's at a point where

98 % of our customers can use what we've got now. People with a lot of data or lot of users or certain specific use cases, it basically just wouldn't work for them.

RICK (04:45.743)
And this is Kanban is like updates in the front end. Like you've to be able to move the thing and not have it break.

Tyler King (04:52.452)
Yeah, yes, a single person can move the thing. One of the things we still need that is like if another person's looking at it at the same time, having that push to them, stuff like that. We're almost done with that. But yeah, so basically what we did, I say soft launch. Previously, if you wanted to look at a pipeline, like a lead pipeline or something, it was what we call list view, just like listing all of the items vertically. Now there's a little button at the top right of the page that says display. And if you click on it,

RICK (05:00.748)
Ooh, yeah.

Tyler King (05:21.444)
It says, you're currently in ListView, do you want to switch to BoardView? And if you do that, you've basically opted into the beta test. Yeah. Yeah, not a lot of people have opted in yet. Hang on, let me refresh it real quick. No, no, not, well, for right now, we had maybe five or 10 people using it like for a couple months now. But for now, want to see, we want to get bug, like give ourselves time to get bug reports and stuff like that. The plan is to wait two weeks.

RICK (05:28.014)
Ah, cool. Do you have stats?

RICK (05:35.416)
Have you marketed it or is it just more like discovery?

Tyler King (05:51.056)
and then turn it on for new users by default. Again, anyone can opt in. So by turning it on, mean, when they go to their pipeline report, they just see the Kanban board. They don't have to switch to it. But yeah, we have 52.

RICK (06:01.902)
What's like the typical, what's like the, I would just be curious, like what's like the mean, median and like range of number of pipeline items? Yeah. Okay. That's where a combined view makes sense. Like if it's tens of thousands, like it's like, why are you even looking? It's not even a useful pipeline at that point.

Tyler King (06:11.376)
Yeah, I definitely don't have like actual stats on that, but I think a huge number of people have, you know, dozens to hundreds, like not a lot. And then...

Right.

Tyler King (06:30.946)
Exactly. And that's been one of the challenges with kind of like scoping the project. I said that if people have a lot of data, it doesn't work very well right now, but it's also like, why would they even want to use this view? But there's a difference between like it breaks your whole CRM and you can't load the page versus it gracefully degrades and only shows you a certain amount of stuff at a time or something like that. but yeah, I agree. It's, you know, something like five.

RICK (06:54.029)
Yeah.

Tyler King (07:00.732)
pipeline statuses with one to ten items per column is probably like the sweet spot of where this is a really good way to look at.

RICK (07:11.758)
Cool. Congrats. Big stuff. Velocity of shipping.

Tyler King (07:12.749)
thank you. Yeah.

Yeah, now it sucks because we're still going to spend two more months finishing it. like, you want to be like, all right, let's move. Let's move on to the next thing. But I don't know. I do kind of think in the world of AI, like shipping quality stuff and actually finishing features and not just putting out slop is going to be important. So yeah, 52 people using it right now. I may give updates on how it goes, but otherwise we'll just assume it went smoothly. The other soft launch is the mobile app.

RICK (07:42.798)
Cool.

Tyler King (07:46.694)
This one is much less ready. let me re we have a mobile app that it's just embedding our mobile website. It works fine, but it doesn't like offer anything that the mobile website didn't. Yeah, well, no, not yet, but we've had it in like like iOS has a thing called test flight where you can like invite people and they can. So I have it on my phone, but you would not find it in the app store. What happened is so one of our main marketing channels is getting like listed on various directories and trying to get reviews from.

RICK (07:57.34)
And you can download it on the app store. Okay.

Tyler King (08:16.432)
notable websites. So we have a PC Mag review. I think this matters even more in the age of AI because as we've said before, AI is like reading these reviews to like, if you ask Chachi BT about LessonWing CRM and then you say, okay, what's your source on that? It will often say PC Mag is the source. They have a policy of like re-reviewing apps periodically. Like we can request a re-review, but there's this like tension of like,

RICK (08:22.67)
Hmm.

Tyler King (08:46.586)
Okay, we're always about to launch something big. anytime you ask for the re-review, it's like, but shouldn't we wait for us to ship this thing? But then if you wait, then it's like, well, now we're another three months away from the next big thing. So when do you actually ask it? I think there's like a every three years or something like that, which I understand why they do that. That's totally reasonable. But so we asked for this and then the reviewer kind of got back to us and was just like,

RICK (08:57.772)
Can you do another re-review?

RICK (09:02.57)
okay.

Tyler King (09:12.1)
I just want to make sure I didn't miss anything so I don't give a bad review. Like, you don't have a mobile app, is that correct? And we were like, well, we're like about to. And they're like, well, can't, I can't, I can't include it in the review if it's not live. And we were like, if it's live two weeks from now, could you include it? And they're like, yeah, probably. So we're like, okay, we're just going to ship it.

RICK (09:31.839)
Ooh!

This is like the most pressure I've seen on less annoying CRM to get a feature out that I've ever.

Tyler King (09:36.69)
Yeah, yeah, I mean, we've never really had deadlines before, but yeah, this is a real deadline.

RICK (09:42.166)
So I'm curious if like people like this.

Tyler King (09:45.49)
I mean, employees like it or users like it. It really only affects one employee, but yeah, I think he's, he and I are both a little nervous. you know, I don't know. think I've talked to myself into this is no problem. Like it's so hard to find apps on the app store that no one's going to see it if we don't promote it probably. But the nervous, the thing that makes me nervous is someone installs it. It's not great yet. Bad reviews, right?

RICK (09:46.634)
Yeah. Employees.

RICK (10:11.16)
Bad reviews.

Tyler King (10:14.958)
I've talked myself into just like, don't worry about it. Like it's so easy when you're running a business to convince yourself that let's go really far out of our way to avoid a potential problem. And then you realize later that wasn't even a real problem.

RICK (10:26.306)
Yeah, you just make sure you call it something that you can throw away and then come up with V2 and start from scratch. really? Call it like the less annoying mobile app and then the next one can be less annoying CRM. Yeah.

Tyler King (10:31.132)
think they are pretty good at preventing that. I could be wrong. I was actually surprised.

Tyler King (10:42.864)
Yeah, LACR, yeah. One of the funny things about this, so, you know, I've always heard that it's so hard to get an app approved by Apple and they just reject for all these reasons and our app's nowhere near done. One of the things I know you need is you need native features. You can't just wrap a website. And another thing is you can't have the full billing system available in there unless you give Apple a cut. You can in America, but you can't in other countries and we're listing it globally.

We submitted it expecting Apple to come back. So I knew those two things would get rejected. But I was like, there's probably 70 other things I'm not thinking of. So let's just submit the app for review and see what we need to do to get it approved. They just approved it. So we're just ready to go. Yeah. So yeah, in the next week or two, we'll probably have a public app. I know, we just need more external pressure on us.

RICK (11:29.186)
Let's go.

Tyler, this is high velocity shipping. If you knew how many times Tyler and I over the last decade, no, how long have you been doing less noise serum? 20 years? No, 15, okay. 16, 17 years I've asked him about a mobile app and he has told me we do not need a mobile app. So this is 16 years in the making.

Tyler King (11:46.29)
1617.

Tyler King (11:55.314)
Yeah, we're a little behind. I haven't said that for several years. I know we're behind. I think in probably 2020, I was saying, yeah, we don't need a mobile app. Yeah, we'll see. Is it going to affect anything about the business? No, nothing does. I'm a total nihilist about it. Nothing matters.

RICK (12:05.833)
Yeah.

RICK (12:18.07)
I can't wait until someone just like writes you and saying, the mobile app has changed my life.

Tyler King (12:23.514)
they certainly will do that. Well, yeah, every little thing you do changes someone's life. just have to, if we're to affect the metrics, you have to change a lot of people's We did, because, okay, we talked about AI Week last time,

RICK (12:37.836)
Yeah, you said you were going to do it. Did you do it?

Tyler King (12:40.388)
Okay, was both, yeah, we did it. Wait, we must have been in the middle. We must have been in the middle of it, okay. Yeah, we did it. Can I just roll right into that? Looks like you've only got.

RICK (12:43.959)
Yeah.

RICK (12:48.278)
Yeah, I'd love to hear about this. I've been I've been chomping at the bit about this.

Tyler King (12:52.018)
Yeah, so as a reminder for anyone who didn't hear last episode, was kind of feeling like we'd been... Lesson on Serum had been waiting on the sidelines with AI. mean, we were... Everyone had a ChatGPT or a Cloud account, like paid. Many people were using Cursor. It's not like we weren't using AI, but we were behind on the Cloud code side of things. So to create a little urgency, we brought Bracken in town from remote.

whole company had to come in five days a week instead of the normal two. We catered every day. Try to kind of, kind of what you're saying the RKO thing was at windfall. Like just try to make it an event of it. And we made the rule, you're not allowed to write any, human code for a week. Like all code has to be AI written. and we threw out or paused all of our other priorities. And I just gave people a bunch of little like features to build to see what we could get done. you, you knew all that already, but that's my summary.

RICK (13:44.886)
Yeah, so what happened? What was the I mean, I want to know like, was it was there any like good or bad things that came out of this?

Tyler King (13:51.142)
Yeah, I so for, I think there's two sides of this. Like one, what did we build? And then the other is like, what did it do culturally? The cultural thing is more important, but I'll start with what did we build? I think we shipped 12 customer facing features that like, this is what we traditionally have called a cup holder. Just things that customers ask for pretty regularly, but things that are very minor iterative improvements. What?

RICK (13:57.624)
Mm-hmm.

RICK (14:15.64)
Dude, you've done more in the last two weeks and then, okay. I mean, like this is, can we just take a second? Like, this is the most stuff that you've shipped in a two week timeframe than like ever since like the early days.

Tyler King (14:21.074)
I mean, Kanban and mobile app were built prior to that. We just pushed deploy, right?

Tyler King (14:35.634)
Oh yeah, absolutely, absolutely. And if, so you have to then ask the question, should we just be doing this all the time? I want to say we built 12 of these features during AI week and then had another almost full week of like bug fixes actually deploying stuff. You know, it's one thing that it's working locally. I have to actually get it deployed and all the developers are like tripping over each other in the deploy queue. Um, so really we're talking more like two weeks.

And like even now, even this third week where we are now, we had our dev check-ins this week and the developers are kind of like, yeah, I have like three times as many bugs as I normally have, primarily because of AI Week. So like it comes at a cost, you know?

RICK (15:15.672)
So the new development cycle is ship, ship, ship, ship, fix, fix, fix, ship, ship, ship.

Tyler King (15:20.166)
Yeah. Well, it also means with that much code shipping, that's a lot of code to review, which sucks. This is the number one thing on my mind related to like AI coding right now is just how do you, what do you do about that review burden? Cause I don't want to just ship slop like I said earlier. but also I, I can't ask a developer to spend 80 % of their time reviewing code written by AI. mean, that's just the worst job experience I can imagine.

RICK (15:26.062)
Mm.

Tyler King (15:49.212)
So it's not all good, but man, these 12 feet, these are, again, these are not earth-shattering features. They don't change anything for us strategically, but I posted about it in our user forum and like, the one that I even forgot to, I recorded a Loom video walking through them and I forgot one of them because I felt like it was so unimportant. And then I left a little follow-up comment like, also this one I forgot. And then someone replied like, my God, the last one changed my life. This is the best thing ever. Yeah, it's cool.

RICK (16:15.651)
That's awesome.

Tyler King (16:18.714)
The cultural side is like, what does it mean for the company strategically? How did the developers feel about it? That's probably the more interesting side. I think like overall positive, but definitely a combination of positive and negative. it's not you go online on Twitter or wherever, and you can see very, very all positive reviews of AI coding. And that, was not our experience. lots of bugs, like I said, I think some of the developers were like,

Wow, I have superpowers now. This is incredible. I think some of the developers were like, this feels like a useful tool, I don't get like the other people around me are moving much faster and I'm only moving a little faster. So this is the thing we still have to dive into is like, there's clearly more skill to it than, there is skill to it, but none of us know what that skill is. I think just some people are more intuitively good at it than others.

And, they don't know, like there's no theory behind any of this, right? So like, how do you teach them? How do you, like, we don't have that solved yet. I don't know. And then I think one or two people are like mourning the loss of human code. Now, of course we've gone back to human coding. I'm not saying the, the all AI code rule was only for that one week, but there is a mourning period going on for some of the developers as well.

yeah.

RICK (17:40.108)
Well, it sounds like you've created a necessary moment of reflection and change.

Tyler King (17:47.29)
Yeah, yeah. Now, okay, so here's one of my challenges now is like, so I sent this whole email out to the company and you read it, right? The, we need to take AI seriously, a newsletter that I sent. Thank you. It was pretty like dramatic, right? I'm like, this could be the end of our business, like that type of language. And I think it did a good job of getting people to buy in. Like even if I'm mourning this, get that.

RICK (17:58.876)
yeah, that's good.

Tyler King (18:14.278)
the business can't wait for me to finish morning. I need to, we need to go. I think I got everyone on board with that. I'm a little now like, okay, now what? Like we did AI week. It felt, that felt like big action, right? The point of this newsletter was we haven't been doing enough. need to get in gear here. And then we did AI week right after that. And everyone feels like, okay, we're in gear now. But like the strategy of the company has not changed. The strategy of the company is still build the same things we're going to build. Just do it twice as fast. It does kind of feel like

like deflating a bit now that I don't have a next thing, you know?

RICK (18:53.098)
Hmm. Well, I think maybe you just observe for a little bit and maybe you need to, know, if it's not going as fast as you think it should, or people are lagging, you talk about it and you do something different, but like, I don't know what I feel like this is a let it play type situation.

Tyler King (18:56.124)
Mm-hmm.

Tyler King (19:08.658)
Absolutely, and that's where we landed on it. I basically sent another message out that's like, we're going to give it a month. What we hope to see is it kind of feels normal in the sense that we're working on our normal projects, but that everything's getting done faster and that certain things will break around the edges, right? Like the code review problem I was talking about. So let's give it a month. Let's figure out what the problems slash opportunities are. And then let's do another round of like, okay, let's change stuff.

RICK (19:23.105)
Mm-hmm. Yep.

RICK (19:31.244)
Most of the things that you share around AI Week are product specific. Do you have any go-to-market or internal operation type AI wins to share?

Tyler King (19:41.218)
Yeah. So our marketer was already more AI-pilled than other people prior to this. I think she hadn't used Cloud Code before though. I think people like, it's all the same thing, which is people building software tools with AI. It's just the CRM coaches are building their own tools and the marketer is building her own tools and the devs are building like the product.

Yeah, so let me run through a couple things. Right now, Eunice, the marketing person, is working on a tool to basically pull all our emails down from front, throw them into a vector database so that we can search for feature requests really easily. And then we're in the early stages of this. A lot has to go right for this to work. But the idea is every time we ship something, she'll be able to say, give me a list of everyone who's ever requested this, and then we can follow up one on one with them.

RICK (20:33.462)
Hmm. That's cool. Yeah, that's really unique experience in terms of a product release, one-to-one outreach and closing the loop. Wow. That would be really cool. I like that one.

Tyler King (20:38.311)
Yeah.

Tyler King (20:42.706)
And I think like long term, if I mean, maybe AI revolutionizes everything again and everything we're talking about now becomes irrelevant. But like in the current world, I feel like that's what it looks like long term is just marketers have a bunch of these little tools that they've built for themselves and the customer service people have their own little tools and you can just do a lot more weird bespoke stuff that would have been impossible before.

RICK (21:05.974)
Yep, I buy that. On this sort of good.

Tyler King (21:10.01)
Related to that, I was going say related to that, think, so one of the big points of friction for enabling marketers and customer service people to five code their own tools is like, what access do they have? They have no training on how to build secure software. And like to do so would be so onerous that I don't think it's worth doing it. So I don't want them writing code in our real code base. What that means is they don't have access to live data.

which really limits stuff. So what I have been thinking I need to do is probably build, and I think this is probably a pattern that would make sense for a lot of companies. I need to build internal API functions for our team to use. So for example, I say we have one marketer, we kind of have one and a half. The half marketer, because she's halftime customer service, she does our affiliate program. She wants to pull down data about affiliates and do all kinds of stuff like vibe code an app that does XYZ.

So I need to just give her an API function that gives her live affiliate data. I need to give Eunice a function that gives her like a way when she's doing this function thing or this, sorry, this searching for feature requests, she then needs to see like which of these users are still using us. We don't want to email a bunch of people that canceled or maybe we do, but whatever. So I need to give her a API function that lets her pull down like our live customer list and compare it against that data. I think that...

RICK (22:09.71)
Mm-hmm.

RICK (22:30.947)
Yep.

Tyler King (22:32.464)
With AI, it opens up so much stuff, but it also creates a ton of new work of this internal systems to enable everybody.

RICK (22:39.488)
yeah, totally. That's been my experience, both at like up health and windfall, where it's like, okay, so we want to do this thing. But the person initiating this thing is a non technical person. I think I sent that gift to you or that video of like the monkey with the machine gun. It's like, very scary. You know, give you you give the wrong access to the wrong person and AI can break a lot of stuff. And so

Tyler King (22:54.608)
Yeah.

Tyler King (23:06.663)
Mm-hmm.

RICK (23:08.398)
the way I'm thinking about it is like, first, there's like, first and foremost, like, is this thing secure from the outside world? Like, you know, am I am I am I unknowingly or or knowingly sharing proprietary data with a third party that we don't want to share it with? Like, that's the first thing like, I want to make sure it doesn't happen. When I'm thinking about this. The second thing is like, okay, how do I enable the right level of access, read only access to the data that they need? And then the third is like, what

Tyler King (23:25.468)
For sure, yeah.

Tyler King (23:33.522)
Mm-hmm.

RICK (23:37.422)
actions, should they be able to initiate that do things, you know, other than like analyze. And that's like where it gets scary, like, should they be able to write data? Should they be able to trigger outreach? Or, you know, who's the supervisor there? And it's very, yeah, it requires, I would say you're in a unique situation in that you can build the connections here and govern them yourself because you're technical. But one of the challenges I have is

Tyler King (23:51.238)
Mm-hmm.

RICK (24:07.116)
I'm less technical than you, so how do I connect these tools in a way that's safe and gives me sort of the ability to administer and supervise? But then you're bringing in another connector and these costs start to add up.

Tyler King (24:21.031)
Yeah.

Well, there's this pressure that's like everyone else is moving so fast. You'd better move fast, but you're doing it and you're like, uh, I could see this going really, really wrong. Trying to balance those two things. I feel like is really challenging right now.

RICK (24:37.63)
yeah. I mean, here's the perfect example is like, you know, as a non technical persons playing around Claude, and it's say it's basically like, hey, you know, tell me your API key for your Salesforce instance. And so I'm getting a slack, you know, Rick, can I have the API key? I'm really close to just like changing the world. And I'm like, No, you may not. Do you even know what is happening here? And like, yeah, it's like, no, I just need to connect our Salesforce instance. And like, can you show me what Claude's telling you to do? And it's like,

Tyler King (24:50.257)
Yeah.

Tyler King (24:59.026)
Yeah.

RICK (25:06.722)
telling it's basically saying build a custom app on the app exchange for Salesforce. But they don't understand what that means. And so it's but they're trying to go fast. Right. And it's like I'm the.

Tyler King (25:12.594)
God.

Mm-hmm.

Tyler King (25:20.21)
So what are you doing about this? Like, I know it's million different answers to a million different questions, but.

RICK (25:25.366)
Yeah. Yeah. Well, I mean, the main thing is like, I think Claude is doing a really good job of standardizing how these things can talk to each other, through the MCP, for example, concept, and then it's like, you know, there's kind of two types of connectors. There's the, on like the, the, the official connectors, which make me feel very comfortable. because there's a lot of like thought being put into like the context that the agent needs for certain tasks, like repeatable,

occurring tasks like I unblock that stuff immediately because because usually within the connectors I don't know if you played with this but you can control at an admin level read versus write permissions and then there's like I want to connect the tool that doesn't have a connector yet well then they start yeah

Tyler King (26:10.93)
Right. Well, and actually let me pause you for a second. Even read versus write, it's not like read is always safe. Like I'm comfortable giving someone an API function that gives them all of our affiliate list, like our affiliate list. That's not our customer. Like of course we want to keep it private. I'm not saying like, you know, I'm making this list public, but if something went wrong, okay, whatever. Even our customer list, that's worse, but it's not the end of the world. But our customer's data, absolutely fucking not, right?

RICK (26:39.468)
Yeah. Yeah. Yeah, totally. I mean, that's the same, same issue, you know? and then if you add in something like Lego palette that has HIPAA privacy concerns, like there's a whole nother layer of, of risk and compliance that you've got to be careful about, with the read, access. So, so yes. but like when, when it's a nonstandard connector, it opens up this question of like, okay, this end user wants access to this, the, the, the, data and the capabilities within this tool, but there isn't an official connector yet.

Tyler King (26:48.337)
Yeah.

RICK (27:08.906)
Okay, what do you do? Well, you build your own connector, which is what you're doing, which I don't feel comfortable doing yet, at least. And then the second is there are third party zappiers basically, coming in play that are becoming the MCP layer. And then you're introducing another party with a whole with exchange fees. And then, you know, okay, it exactly and

Tyler King (27:32.934)
and they can get hacked or they could have bugs,

RICK (27:36.608)
And what's what I like about the official connectors is usually they're building the sort of the the the they're building the connector in a way that like is built out of the box to handle the common things that a that a end user wants to do with AI with a lot of these MCP layers. It's like, you have to build that like you connect it, but then you have to build the actual like here's the skill here's

here's the context you need for this action. It's not like, know, Claude makes it to the end user saying, this is a piece of cake, like, just go do this. And, you know, I'm going, no, no, this is a lot of work. So, and it costs money. So like, how do you know where to invest? like a lot of this stuff, like you could spend a week or two on of wasted work, and it not and then blow through lots of credits and dollars.

Tyler King (28:18.385)
Yeah.

RICK (28:32.886)
And all of a sudden you're like, why did I just waste two, three weeks and you tinger in on this?

Tyler King (28:39.058)
Yeah

Come.

RICK (28:44.556)
But you're lucky because you can build this stuff internally and you you understand it.

Tyler King (28:47.302)
Well, yeah, can within reason though, but I'm still like, again, the people who are just blast shields down, I don't care about security. I'll let open claw do whatever it wants until it bites them in the ass. Those people are winning. That's the, that's the problem, right? Is there moving so much like, yes, I can build these API functions, but I'm looking at it like, well, how much data should I expose? And

I don't want to have to keep redeploying. Like, every single time they want to build something new, have to go in and write more code again. So I got to think through future use cases. And I just feel like I'm being too conservative, but I'm not. I want to be clear here. Like, especially if a customer is listening to this, we are not being reckless here, but you've got to understand that FOMO feeling of, like, am I going to lose to the people who are reckless?

RICK (29:41.166)
Yep. Yep. And then, you know, the conversation that is you get larger as an organization you have to have is like, how do you explain this to non-technical people so that they understand how prioritization is happening? And we're not saying no, we're saying not now or let me find out more information first. And it's so it creates internal friction. It's a it's a whole new paradigm.

Tyler King (29:49.468)
Mm-hmm.

Tyler King (30:01.148)
Yeah. Yeah. Do we have to start training every employee on not, we already do security training with every employee about like how not to get fished and stuff like that. Do we have to start doing like technical like infrastructure type security with every single employee now? Maybe. Yeah, I guess so.

RICK (30:15.394)
Yes, if they have access to these tools and we want them to have access to these tools, we absolutely do. So speaking of like just kind of containing the AI thread for one more application is like I have I am very excited about I mean, and when on one end, I'm like, okay, this is bad. On the other end, I'm like, this is super exciting. I have lots of content on right, there's not good content out there for health insurance. And our space like a health and between like notion, Claude Webflow, like

the, it's pretty cool. Like I can write really good content, do the final edits, through prompt, like get to a really good final draft and then do like human editing. And then I click a button and it goes live on the website. It's so cool. Yeah.

Tyler King (30:58.768)
Yeah. So let me dive into this. So what is your flow for writing something? Do you write a draft and then have AI review it? Do you have AI write the first draft and then you review it?

RICK (31:08.494)
Yeah. So, um, I have a pretty good sort of prompt and notion for drafting content. So I'm using basically the AI interface and notion to get to get from like a topical concept to a first draft. Um, you know, whether it's short form or long form, um,

Tyler King (31:25.778)
Do you tell it, I want it to be about this topic, here are my insights about this topic, write something.

RICK (31:29.678)
Yep. And then and then, you know, if I have source material, you know, third party source material, or, you know, internal source material, I'll provide it access to that. And then I'll ask it to, you know, build a meta description, I'll ask it to draft a description of the right image, you know, that it should have. And then it drafts it, I then go through around a human review. In notion, kind like an icon bomb board.

Tyler King (31:46.706)
Mm-hmm.

RICK (31:55.212)
And then once right now, what I'm doing is I'm going into Claude and I'm saying, Hey, here's, here's a link to a notion document. Can you, can you get this staged in web flow basically? and what I'm going to have is when it hits the cert, like the publish button and notion, it just automatically goes to Claude and then Claude just handles it from there through an agent. but the, the, beautiful thing here is it to actually like,

Tyler King (32:13.33)
Mm-hmm.

RICK (32:20.29)
The two things that take the most time, well, there's three big steps. The first step is coming up with the topic and getting the right sort of like brief in place for the AI to write the content. The second is the final, getting to the final draft of editing, which takes some time and doing the interlinking of other pages and that kind of thing. And then the third is actually publishing the thing, finding the right image, getting it staged on Webflow, going through like HTML formatting issues and...

Claude is really helping to address number three, which was a huge bottleneck for me. And like, I would get a lot of content written, but like, I've got to go find the image that matches this. I've got to, you know, I've got to go get it, copy and paste it all into Webflow.

Tyler King (32:54.716)
Yeah. And you're using Cloud Code or like Cloud the web app? Cowork? Okay.

RICK (33:00.622)
Claude Cowork. Yeah, well, I actually, I'm even using Cowork. I'm just using Claude the interface. And so basically it's saying, okay, cool, yeah, I can do this. And Chat GBT is doing the image creation. I've basically got a project in there where it's like got standard illustrations that I've created that depict these. I didn't know Claude could do the image. So, but I had already set up the chat. Before I had Claude.

Tyler King (33:18.864)
Why chat GPT rather than Claude for the image?

RICK (33:28.982)
I had already started using chat GBT for image illustration. I just.

Tyler King (33:31.602)
But so, Claude, and sorry, we're getting into details here, but I do think like, we're all, us and all our listeners are trying to learn the specifics of this, so it's worth it. So you said you're not using Claude, so you're using the Claude desktop app, there are three tabs, the web app, okay. And that, the web app can call the OpenAI API, how's it interacting with OpenAI?

RICK (33:36.204)
Yeah.

RICK (33:45.486)
No, I'm using the Cloud Web App. Yeah.

RICK (33:53.998)
I am, I am playing that role currently. So what's happening here is it is the, the one, uh, there are two things that I don't have connected. The first is when I would like it from notion, uh, to have everything ready to publish. Uh, so like, would like notion to go get, um, the image, uh, from chat, GBT, um, then hit publish and then Claude just takes it to webflow from there. Um, right now what I'm doing is I'm getting it to final stage and notion.

I'm going into Claude, I'm giving it the notion thing. I'm going to chat GBT, I'm getting the image and I'm feeding Claude the image and that stuff. And then it's pushing it all to Webflow.

Tyler King (34:27.282)
Gotcha.

Tyler King (34:31.474)
All right, I have a few thoughts here. One, is it worth the next step of automation? Like it sounds like you need to spend two minutes per post moving stuff around. You might spend hours getting this automation built. Meaning you'd need to write 200 posts or something for the automation to be worth the effort. What do you think about that?

RICK (34:54.038)
Yeah, this is working for me right now. It's what's what's happened. What typically happens is if I'm doing something over and over again, and I'm like, I could spend 10 minutes right now to make this better. It's like it's very iterative, like, so so right now I'm like, I'm pretty happy. But yeah.

Tyler King (35:05.522)
Yeah.

Tyler King (35:09.458)
It does sound like the next step. I could be wrong about what Claude, the web chat tool can do, but I don't think it can take this to the next level. I think you'd need coworker Claude code to do the next thing, which is a little less iterative. It's a little more of a step change, I think.

RICK (35:26.722)
Like having a full-fledged

Tyler King (35:29.338)
Yeah, and I don't think it'd be hard. not saying it's like a big project. I'm just saying, I don't think there's like a really, really small step to get the next thing and then the next thing. then I think there's like, need to move the whole thing into Claude cowork.

RICK (35:33.9)
Is it worth it?

RICK (35:40.356)
Can I tell you what my next step is? It's not actually work improving this process. It's more on the front end. So right now, like I'm the one coming up with the topics. So I've signed up for a tool called AirOps. AirOps is basically an SC like a modern SEO tool that monitors chat, GBT, Gemini, and does like prompting to see what like, if you type in like, what are the best health insurance providers in Utah? Who's getting referenced?

Tyler King (35:45.564)
Mm-hmm.

Tyler King (35:59.794)
Mm-hmm.

RICK (36:05.186)
Are you getting referenced or your competitors getting referenced? Are you being mentioned? and like, how do you, what content should it actually prescribes? What content to write to get mentioned in these things. So what, one thing I would like, what's next on my radar is in addition to my sort of human driven workflow is to get an automated cadence going where air ops is triggering briefs to notion in a certain format. Notion is getting it to a final draft. I'm reviewing the final draft and picks up from there. That would be my next sort of iteration.

Tyler King (36:34.258)
Cool. Yeah.

RICK (36:34.945)
Yep. It's pretty cool. But I mean, at the end of the day, like if everyone's doing this, eventually there's just content everywhere. And I'm like, what is the point?

Tyler King (36:43.698)
So, okay, I had this at the bottom of my list, but this segues too perfectly, so I have to say it. I think every single marketing channel is gonna get destroyed in the next year or two. Maybe every single one is an exaggeration, but anything that can be destroyed by overdoing it is gonna get destroyed. Which A, I have conflicting feelings about, right, so does it mean you wanna be first, or does it mean don't even bother because like,

RICK (37:05.134)
That means we need to race towards destroying it. Take advantage of it before it gets destroyed.

Tyler King (37:12.37)
it'll stop working so soon.

RICK (37:14.988)
My, I mean, my, my bias is like, there's going be new marketing channels and the, the, the first to squeeze the juice gets the juice and onto the next one.

Tyler King (37:25.766)
Yeah, but I don't think being really good at automating AI slap on a blog puts you in a better position to take advantage of the next channel necessarily. I agree, there will always be new channels. People are gonna buy stuff, especially health insurance. Everyone's gotta buy health insurance. They're gonna find it somehow. Yeah, I don't have a lot of action to suggest for this.

RICK (37:45.132)
What's the buying journey? Yeah.

Tyler King (37:53.19)
But like, just, seems obvious. Like cold outbound emails cannot possibly keep working. I don't know, our email inboxes are just gonna get destroyed. I don't know what's gonna happen. Even calls, right? Like AI can do phone calls now, so you can scratch that channel off.

RICK (38:08.342)
Yeah, I mean, the amount of unknown callers calling my phone every day, I just filter them out. I don't even look at them.

Tyler King (38:11.738)
it's Mm-hmm. And like, one scary thing about that, that's what JD does. what are his... So, okay, and then there's inbound, various forms of inbound, but even like, well, another thing like UpHealth does is buys leads. So some other company makes websites that get people interested in health insurance on them. They have a lead capture form. I feel like that stuff's going to get watered down.

RICK (38:37.589)
It's in-person events,

Tyler King (38:42.245)
But is anyone buying health insurance as a result of in-person events? I get like windfall can go to...

RICK (38:47.008)
I not know, no, they're not buying, but that's where relationships get developed. And then you're like, when you have a need, I'm here and there's association, you reach out to someone you trust and get a referral. That's like the future. Like, that's like, if you want to own something, own the trusted relationships, and be the first recommendation for someone when they call call for a trust recommendation, like that's what like long term, I think we need to invest in. Yeah.

Tyler King (38:50.918)
Yeah.

Tyler King (38:54.808)
Okay.

Tyler King (39:06.436)
I hope that's true. mean, there could not be a better answer to this for less annoying CRM. It's harder for like, what that favors incumbents over new entrants and it favors basically companies that focus on user experience and like making their customers happy. Yeah. Well, word of mouth, like, everyone uses Salesforce, but no one likes Salesforce, you know? They have a way where you do, but most people don't. Did I hurt your feelings?

RICK (39:22.264)
Word of mouth, yeah.

RICK (39:33.492)
No, I just like what someone says, like, should I set up Salesforce? I'd be like, yeah, if you want to scale your business, it's the best tool for that.

Tyler King (39:40.12)
Right, but you're not going to get full-throated like, okay, Salesforce has the decision makers who choose to use it like you. And then they have the no, no, no, but okay. But then there's all the people underneath there, all the people, the sales reps and the customer service reps. If someone asks them, they're going to be like, no, do not use Salesforce.

RICK (39:47.198)
trusted referral. Like I'm the trusted person. Yeah.

RICK (39:56.172)
I think this favors Salesforce too, is all I'm trying to say. Just a different market than Less than One CRM. I'm not arguing with you. I think what you're saying is like end users that are delighted are more likely to refer Less than One CRM than a Salesforce end user who doesn't like Salesforce.

Tyler King (40:09.734)
Well, and Salesforce has a core competency of doing sales. Like, they're not a product, they're not a tech company. They are a giant sales team with a tech product attached to them. That's about to get, yeah.

RICK (40:19.906)
They own Slack.

I mean, that's a tech.

Tyler King (40:24.466)
No, they have tech products, but I'm saying like Salesforce, like I worked at General Mills for a while and they were very open about this internally. They said, we're a marketing company. The next CEO is going to be from the marketing department and the CEO after that is going to be from the marketing department. It's not, you you think of us as a serial company, we're not a serial company, we're a marketing company. Salesforce is not a CRM company, they're a sales company. They're selling the exact same CRM that Oracle sold before them. I mean, that's not literally true, but you get what I'm saying.

RICK (40:28.567)
Mm.

RICK (40:49.514)
Yep.

Tyler King (40:53.05)
One of the reasons this word of mouth thing advantage is companies like Less Knowing Serum is because we suck at every other form of customer acquisition. If it's true that all those other ones are going to get watered down, this is going to equalize things in a way that favors us.

RICK (40:59.603)
Yeah.

RICK (41:04.13)
Yeah, I see what you're saying. Yeah. Yeah. I think I do think that like the thing that less knowing serum will probably never do is intimate, like field events, like with high level decision makers. And that is something I think Salesforce does. And we'll continue to do more of is like, Hey, Rick, I'm to be in Utah. we're getting, we're getting a collection of VP of rev ops together for dinner. Would you like to come? Like, I think that kind of thing, if you have a high enough ACV is very relevant. you know, like I've helped this little borderline on that. we would need to.

Tyler King (41:13.648)
Yeah, true.

RICK (41:33.826)
we would need to move up market a little bit to be able to support that.

Tyler King (41:36.348)
So yeah, the ACV thing is exactly where I was gonna take this as well, which is like, maybe there should be two different questions here. How do high ACV companies grow and how do low ACV companies grow? The things that are getting blown up will affect both. When I say blown up, like water diluted by so much AI slop that you can't, the channel doesn't work anymore. It's especially gonna hurt self-service companies. I know I'm talking out of both sides of my mouth. I'm saying this is gonna be good for us and bad for us, but.

RICK (42:00.371)
Mm. Yeah. Your affiliate channel is important.

Tyler King (42:05.904)
Yeah. that's true. But yeah, like anyone who is doing inbound, like HubSpot built a whole, I think at this point, they're probably more sales driven than they used to be, but like the whole inbound marketing thing that they invented.

RICK (42:17.582)
Have seen their stock?

Tyler King (42:19.504)
Well, every tech stock's way down, right?

RICK (42:21.582)
Well, HubSpot has gotten creamed this year. They're down, uh, over since January, 2025, they are down from 700 to 260. Oh yeah, it's bad. Really bad. I haven't, I just heard about this recently. I don't know fully why, but I think it has to do with like their, their inbound sort of marketing piece. It's just getting blown up.

Tyler King (42:23.81)
Yeah, they're one of the...

Tyler King (42:34.342)
Oof, I didn't know it was that bad. Do you know why them specifically versus...

Tyler King (42:46.002)
That makes sense. So, iView HubSpot is not our main competitor in the sense they're a much more robust product, but they are the main freemium CRM. when our customers are buying a CRM, they don't want to pay anything if they don't have to. And HubSpot is the best free one. They made pretty significant, like they pulled back the utility of their free tier pretty significantly at the end of last year.

And if they keep getting creamed, they're probably gonna have to keep doing that more. That would be fantastic. I would love that.

RICK (43:18.668)
Yeah. Yeah. And that's, mean, I can't say more, but I'm happy to talk to you offline about like our experience at both like health and windfall. It's like complete, like it's getting less utility for like a health, at the price point that we're at. Yeah. Yeah. It's been bad. It's been bad. and then, on the, on the windfall side, it's a totally different type of experience. which I can tell you about. Yeah.

Tyler King (43:28.37)
Mmm.

Tyler King (43:31.762)
You can't talk about windfall, you can talk shit about your leg up health experience on air. Okay.

Tyler King (43:44.646)
So wait, sorry, I know you can't say as much about WinFall, but you use both Salesforce and HubSpot at WinFall? Okay, at like of health, when you say HubSpot's bad, would you consider, I know it's not a priority to like use anything right now, but.

RICK (43:49.262)
Mm-hmm.

RICK (43:56.91)
The gap between like unlocking functionality from a price perspective is almost $1,000 a month, like to step up and it's like, okay, I'm going to go to another tool.

Tyler King (44:05.776)
Yeah, this is.

Tyler King (44:10.706)
Yeah, this has always seemed crazy to me about HubSpot, because this has never not been true. It's always been the case that they have a free tier, they have like a, whatever, $25 a month tier, and then the next one up is $1,000 a month or something like that. 800.

RICK (44:24.514)
say a hundred. Yeah. Yeah. For annual pricing. so, you you get, so it's, it's, really bad. and so what, what I'm starting to question is like, but, HubSpot is really, really good at what they do. So, but, but I believe what the market is saying is a new increment founded on AI is going to beat them at this. and you're gonna be able to go ahead.

Tyler King (44:30.406)
God.

Tyler King (44:38.94)
Mm-hmm, mm-hmm.

Tyler King (44:48.678)
Yeah. I do think all this like the SaaS market drawdown is, it feels wrong to me, I think. Or tech and, it feels very premature to me. A lot of these companies are making more money than ever before and seem cheap. I've never bought an individual stock in my life and I'm never going to, but it just feels like a buying opportunity right now to me. Sorry, that's not the point of this podcast, I guess.

RICK (45:12.493)
see. This is not a financial advice. not, you know, do not interpret it that way. The only thing I'll just sort of give anything else you want to cover before on this stuff? Yeah. Okay. The only thing I just wanted to provide an update on is we at leg up have been going through an intentional sort of nature walk planning process. I think in the you know, in the month of February, it sort of reached the point of like

Tyler King (45:20.274)
You

Tyler King (45:26.908)
I have other topics, but yeah, they can wait.

RICK (45:41.61)
maximum uncomfortability and the result was like, okay, let's start to narrow in because this is like, this is making us all uncomfortable. so we're starting to narrow in on some themes. and, I'm excited. I think in March, it'll start to take form into like some clear, like priorities. the two things that are standing out that we need to. Like absolutely do is one is, help JD with, with like the initial sales experience, whether that's with tooling or, you know, just

Tyler King (45:46.514)
Mm-hmm.

RICK (46:11.01)
process improvement, helping him like, you're intake a small business quote request and very quickly provide them value on like what their options are, benchmark them against their competitors, that kind of thing. so that they can make a confident, like directional decision on what weather to offer benefits at all. And then what, what chassis to put them on, to put the dollars into and how much. Yep.

Tyler King (46:17.052)
Yeah.

Tyler King (46:33.522)
Can I expand on that just for the listener that like just to add a little specifics, like the normal health insurance quoting process is, you're an employer looking interested in group insurance. Give me all of your employees information. Give me an estimate of who's going to opt into the plan and this and that. And then I'm going to take that and go get quotes for you. But it takes like a ton of work for the lead to get the information they want, which is what will it cost? And so.

RICK (47:01.856)
Yeah, and even worse than that, like, it jumps typically into like, they're getting quotes from an HRA provider, they're getting quotes from a group health insurer provider, they're going to a PO and they're going to three or four different places, they're all getting sold slightly different things, and they may have been working with a couple of brokers. And they're all asking for the same information, which is like, massively time consuming. And so like, sorry, keep going.

Tyler King (47:21.926)
Well, just to say, so like when you say help JD with this process, it's like, how do we give them a rough estimate of a price without them having to jump through all those hoops? That's how I understand it.

RICK (47:32.718)
Yeah, and I, yes. And where I'm taking it a little step further is like, the only reason to do this is because your employees are asking for it and you're going to lose them or our data says that your competition is going to beat you. the, the, the whole point is to recruit, retain people, right? Like how much should you invent? Like the first question is how much budget should you invest into health benefits? First and foremost.

Tyler King (47:59.826)
Mm-hmm.

RICK (48:01.504)
Okay, once you have that number, it's like, okay, you might know what that is what you're willing to invest. What is the best pathway for you? Like, can you even afford group health insurance?

Tyler King (48:09.298)
That's not how we, like what you're saying sounds logically true, is it emotionally true? Like I feel like some people will go into this like, we need group insurance. I don't know what it costs, but we need it. Okay.

RICK (48:19.522)
great, then we'll go then. Yeah, let's do that. Like that's that's I think that's part of the question here is like, where are you in this paradigm? Are you are you looking for group health insurance? Are you are you unclear on like what you should be doing with health benefits? And it's like, we're helping them recommend there. And then yes. So that's one and then the second is just getting more, you know, inbound leads for JD and whether both of those things if we do both those things, we'll grow the business. The third sort of thing that's emerging for me

Tyler King (48:25.66)
Mm-hmm.

RICK (48:49.07)
that could be a kind of a fun big bed is if we can figure out how to expose a lot of our structured plan data, cause we, we track plan data for all of our customers and then, centralized, some sort of, chat experience. I don't want to it chat experience, but like interface for what it could be text, text message base email base where they could ask questions to the AI, in a lot of ways they're going to get better answers than JD can provide in real time.

based on the structured data because of how structured health insurance is. I think there is something there. I just don't know it's gonna solve a problem for us this year. It's more like a long-term vision.

Tyler King (49:28.433)
Yeah.

Well, you know more about this than me. That doesn't thrill me for two reasons. One is I don't think that's why people buy. Like it might be a nice value add after someone buys, but like that's not the bottleneck. But the second is, man, you don't want to get it wrong. You really don't want to get information wrong with health insurance. And even if it's right 98 % of the time, that 2%, yeah.

RICK (49:51.17)
The one wrong. Yeah. Lose the trust. yeah, it's, mean, it's yeah. Yeah. Yep. Yeah. No, I get it. I get it. I, I, I just, I guess what I'm trying to say is there are lots of questions people don't ask about health insurance that they would maybe ask if they had more real time answers and they trusted the answer source. and I think like everything from, you know, what's a,

Tyler King (49:54.982)
Not just, I mean, get sued, I think.

It makes me nervous.

Tyler King (50:16.242)
Mm-hmm.

RICK (50:20.652)
like what's a deductible? What's my deductible? What does that mean? To like, you know, help me understand what a health insurance plan is, like, like basic stuff that AI could do a pretty good job of making them feel self safe answering and educate. And then eventually like, get to like, okay, we need, this is an escalation to a human for, for, you know, for your, for your safety. But, but those are kind of the three themes that are emerging for me from the planning. And I, guess I'm excited to.

Tyler King (50:22.012)
Yeah.

RICK (50:50.816)
not be doing this anymore and start executing.

Tyler King (50:53.18)
Yeah, for sure. Cool, yeah.

RICK (50:54.68)
Cool. Well, if you'd like to review past topics and channels, visit StarvedForLess.com. See you next week. Bye.

Tyler King (51:01.596)
See you next week.