Whoops! Our last recording got botched halfway through and only Andrew's audio was saved. So we recapped what we talked about in the last podcast and got into some new stuff in today's episode. 

Sean is back from his busiest RSA yet and growing Miscreants like crazy! In fact, they're growing so fast that Andrew is hiring a full-time product designer to join the Miscreants team. Meanwhile Andrew is also rebuilding MetaMonster from the ground up (well, Austin is) and thinking about how to build products with AI at the core instead of just tacking it on. The guys talk about the Vercel AI SDK, the value of time in market when growing an agency, Sean's progress on his new Margins product, and more. 

Links:

For more information about the podcast, check out https://www.smalleffortspod.com/.

Transcript:
00:00:01.81
Andrew
Yo, so funny reason that I was late to this podcast recording. I was distracted trying to write a job description for the miscreants product designer role that we're trying to fill.

00:00:13.82
Sean
yeah.

00:00:17.66
Sean
That's true. You want to write the other ones, by the way? trying to hire like three more roles at this point. Yeah.

00:00:24.94
Andrew
That's nuts, man. Hey, actually, i was going to put in here.

00:00:26.23
Sean
I'm

00:00:28.46
Andrew
How many people have you hired in the last year?

00:00:31.76
Sean
I don't fucking know. I don't know, man. I can't.

00:00:34.90
Andrew
If you had to guess, like, give me give me a number.

00:00:37.42
Sean
In the past year,

00:00:41.08
Sean
five, six.

00:00:42.96
Andrew
I guess six. So I'm just going to leave it as six.

00:00:44.50
Sean
Yeah. Okay. to fire Yeah, we've hired so many full-time.

00:00:50.71
Andrew
Or should I say, like, we've grown the team by 50% or something like that?

00:00:55.02
Sean
Sure. Yeah, we've grown the team by 50%. It sounds way better.

00:00:58.97
Andrew
Yeah.

00:00:58.92
Sean
We've doubled in size. we're about to double the fucking team again at this point i think i like said on a call that jj heard where i was like yeah like we're trying to be like you know 25 30 people by the end of the year and she so yeah that was her reaction don't know

00:01:10.57
Andrew
Yeah.

00:01:15.03
Andrew
I'm sorry. What?

00:01:19.88
Andrew
I wonder why. i wonder why your COO would be is shocked to hear you want to triple it in size by the end of the year.

00:01:27.88
Sean
Yeah, yeah, but this was, you know, it's not triple in size. It's six times in size because just kind of said it

00:01:36.66
Sean
last year. ah home or or like the beginning. I think I said in the beginning of the year. i mean, I don't know. Inbound has been good. Inbound has continued to be really good.

00:01:47.89
Sean
RSA was really good. So it just has kind of kept growing. and It was i was kind of said not in jest, but in hyperbole.

00:01:59.66
Sean
It was hyperbolic. I feel a little less hyperbolic at the moment.

00:02:00.98
Andrew
Yeah. I will say...

00:02:06.94
Andrew
I will say always thought that like 20 to 30 people would be kind of the perfect size for an agency because at least for us, it was big enough that we could have enough clients to that.

00:02:14.82
Sean
Mm-hmm.

00:02:22.91
Andrew
Like if we lost a client or two, it wouldn't be like 20% of our revenue, 30% of our revenue.

00:02:27.42
Sean
Yep.

00:02:28.04
Andrew
Like it would be a smaller chunk. It would be,

00:02:30.23
Sean
Yeah.

00:02:30.70
Andrew
You know, we could have a good manageable number of clients. It was still small enough that you wouldn't have a whole lot of process or red tape. It would still be pretty flexible, pretty collaborative.

00:02:40.12
Sean
Yeah.

00:02:41.74
Andrew
and And also big enough that, like, if you lose an employee, like if ah if an employee takes a new job, you weren't losing, like, you know, when we sold Crit, we were six full-time people, 10 including part-time.

00:02:47.89
Sean
Mm-hmm.

00:02:53.71
Andrew
And so if we lost one person, like one full-time person, that was, you know,

00:02:54.21
Sean
Mm-hmm.

00:02:58.47
Sean
It hurts.

00:02:59.71
Andrew
20% of our full-time workforce.

00:03:01.87
Sean
Yeah.

00:03:03.66
Andrew
That's brutal.

00:03:05.33
Sean
Yeah.

00:03:06.87
Andrew
So I do think that's a good size. Now, would you want to stop there? Because you are more ambitious than I am, probably.

00:03:14.82
Sean
I am i am greed personified. Yeah. I mean, when when the years started, i like one of the exercises I did with JJ was that we sort of mapped out what a 20-5 to 30-person agency would look like.

00:03:30.21
Andrew
Yeah, I remember this.

00:03:30.53
Sean
you know with yeah yeah and Yeah, I think maybe I showed you as well that org chart.

00:03:34.35
Andrew
Yeah.

00:03:36.49
Sean
I mean, I think agency-wise, I'd probably stop there. I'm saying probably, i don't know, give it like six more months, I'll let you know.

00:03:46.79
Sean
I can't, i think i i I think I can't fathom like more.

00:03:51.38
Andrew
Running a 50-person agency or something.

00:03:53.54
Sean
Yeah, yeah, like that, like I can, I think at that point it's just, like, I

00:03:54.31
Andrew
Yeah.

00:04:01.50
Sean
i think, I mean the market is huge for sure.

00:04:06.89
Sean
I think I would just, I think my shiny objectpreneurism would just want to do other, like even in that 30, right? There was still like two people were doing merch. other There was like a developer on a product and I think those things would grow.

00:04:19.97
Sean
I think mysteries as a whole, I can see becoming larger and larger, but like pure client services, I don't know. I think it's, I think quality of work still matters and I still worry about getting too large in that way.

00:04:31.97
Andrew
It's definitely like possible to scale while maintaining quality, but it is fucking hard.

00:04:36.95
Sean
Yeah, I mean, we've won deals because people are unhappy with killer agencies because ultimately, you're large, even now, can't even now we can't like even now when a creator comes to us and they're like hey i only have like this much money to do it like like i have three thousand dollars to do something right it's like dude we can't like we can't take that but like at this point we'd rather just do it for you as a like here's a free creator program and like you know ultimately like you have to know that you're not going to get like

00:04:55.63
Andrew
Mm-hmm.

00:05:11.23
Sean
creative director bends time on every single part of it, you're gonna get, it's gonna be a training project for ah junior. It's still gonna be good, it's just not gonna be, you know, it like, the, the, cause, cause the, ah the value dispatch is different, right? For them, like the $3,000 is a lot.

00:05:26.32
Sean
For us, it doesn't sustain anyone for time. So, but like, we don't want to give, like, we don't want you to have to feel like you paid for like crappy work. That was all a long timeline. But if you get it all for free, it's kind of evens out that way a little bit that feels i don't know that feels so like narcissistic or egotistical to say sometimes

00:05:50.51
Andrew
How's that?

00:05:51.98
Sean
it's like

00:05:53.45
Sean
Because it's like, oh, you're not paying enough for us to care about you. you know and that's like Like if I was a freelancer, I'd take it. you know and And I have done that on the side for some friendlies who, like like long-time industry friends who needed something.

00:06:06.58
Sean
Yeah.

00:06:07.30
Andrew
Yeah, but like you just said, it's the realities of like you have bills to pay and people to sustain.

00:06:11.40
Sean
Yeah.

00:06:13.26
Andrew
And and like, you that's a drop in the bucket for, you know, an agency level operation.

00:06:16.13
Sean
yeah For sure.

00:06:19.95
Andrew
It, you know can be very different for a freelancer.

00:06:20.21
Sean
For sure.

00:06:22.33
Andrew
And that's why, know, sometimes if you have a small budget, it's better to go to a freelancer, someone to, you that is a bigger impact for.

00:06:27.65
Sean
For sure.

00:06:30.85
Sean
For sure.

00:06:32.30
Andrew
Yeah.

00:06:33.14
Sean
Yeah.

00:06:33.82
Andrew
Yeah.

00:06:34.05
Sean
Anyway, the the point I'm getting at earlier was just like, you know, i've we've gotten deals because people felt scorned by their agency because they went to the large trusted agency, but like didn't pay the large agency budget and wanted the agency quality.

00:06:46.07
Andrew
Yeah.

00:06:46.06
Sean
But like ultimately they got, you know, juniors, mid-levels working on their things and then And then they're like, we weren't super happy. So I don't know. I guess that's the cool thing about ProServe, right? It's like you could always start something because the market is just so professional services.

00:06:59.55
Andrew
What is proserve? Oh, OK. I've never actually heard that term.

00:07:04.07
Sean
like the I think i think an agency in like creative agency spaces, it's not used so much as much like law firms and whatever.

00:07:09.08
Andrew
Mm-hmm.

00:07:10.53
Sean
But I think it it's everything under that NAICS code or

00:07:10.84
Andrew
Mm-hmm.

00:07:13.73
Andrew
Yeah.

00:07:15.28
Sean
Anyway, that i like the cool thing about ProServe is that It doesn't matter how many people are in the room, like you can always start another one because there's like there's just so many buyers at the and end of the day.

00:07:28.21
Sean
But I don't know.

00:07:30.78
Andrew
Cool. So since you borked the last recording, give us give us like a 30 second recap of RSA this year.

00:07:36.46
Sean
Yeah.

00:07:40.85
Sean
Yeah, yeah. So, okay. So for context, we got like 23 minutes in before Zencast was floating me a message and then we just kind of risked it and it turns out it only recorded Andrew's voice the entire time.

00:07:47.98
Andrew
Yeah.

00:07:50.96
Sean
So,

00:07:51.69
Andrew
By the way, I published that episode to Transistor and then unpublished it. And I don't know how that works with like RSS feeds and podcast apps and stuff. And so there might be just a kind of borked episode out there for people or it might.

00:08:05.50
Sean
I mean, they can't even tell our voices apart, right? So, like, they just...

00:08:07.55
Andrew
Yeah, it doesn't matter.

00:08:09.05
Sean
this This podcast has just been one person talking to himself the entire time. It's just...

00:08:14.26
Andrew
All right, 30 second recap of RSA, go.

00:08:16.68
Sean
RSA was super busy. RSA was cool. Met a lot of clients. Met a lot of prospects. Brought Ben and our product marketer, a bond there. was good to out with them.

00:08:26.96
Sean
And we have been swamped ever since with Inbound.

00:08:34.15
Andrew
Hell yeah.

00:08:35.70
Sean
Yeah. Which is great. You know.

00:08:38.89
Andrew
Was there anything you did differently this year that led to the increase in inbound? Or do you think it's just like time? Because this is something I always felt with marketing agencies or that like there's a component to marketing agencies that is, you thought leadership and, you know,

00:08:57.22
Sean
Yeah.

00:08:57.67
Andrew
doing talks and networking like doing all of the marketing activities but there's a large portion of your success in marketing and agency is just like how long have you existed and how long have has there been time for people to like find out about you like marketing and agency feels much more like brand marketing to me where you just kind of need to get the word out slowly over time and like

00:09:18.86
Sean
Hmm.

00:09:24.38
Andrew
build trust that you're not going to just disappear with their money versus demand gin.

00:09:28.01
Sean
Yeah.

00:09:32.99
Sean
Yeah, yeah. i think I think when you are this specialized into a single industry, time in industry, like there there is just benefit of seeing the same people and then being like refreshed in their mind, right They're like, oh yeah, these guys exist and it's not the first time they're hearing about it.

00:09:53.01
Andrew
Yeah.

00:09:53.73
Sean
And then you can update them like, here's the stuff we've done since the last time we spoke and i think over time that build that that builds a lot of trust so i mean i i do think just time in seat has paid off dividends i think there's a lot of like smart things we've done i think a lot of these like the inbound now is just like i i keep saying to the team it's like goodwill compounded over just all the years of

00:10:18.12
Andrew
Mm-hmm.

00:10:19.52
Sean
you know, the times where we have done really cheap work, like really like where we've just bought the project basically, or even maybe it wasn't as profitable because it was early on or just like times where we've, you know, done something as a favor to someone, maybe not even agency related or something like that.

00:10:38.24
Sean
yeah i mean in terms of things we did differently this was the first ah rsa out of six rsas where i actually had like planned scheduled meetings on my calendar and it was meeting people not just like going to different people's happy hours like i got to walk the expo for once because was just in different meetings every single day so i think i think that

00:10:58.97
Andrew
But like my guess is you were able to schedule those meetings again, just kind of because you have now built relationships and built trust and built the miscreants name.

00:11:02.28
Sean
yeah

00:11:05.84
Sean
Yeah. Yeah.

00:11:08.89
Andrew
And so people wanted to take your meeting invite versus, you know, five, six years ago.

00:11:11.44
Sean
Yeah.

00:11:15.31
Sean
Exactly. Exactly. So, I don't know. It's been good.

00:11:20.04
Andrew
Cool.

00:11:21.48
Andrew
Yeah.

00:11:22.02
Sean
Kind of terrified at the amount of inbound, but I'm glad we have another project that we're working on with you.

00:11:26.20
Andrew
Yeah, yeah, yeah. I am too.

00:11:27.38
Sean
Yeah.

00:11:28.00
Andrew
kind of full circle. I don't know if you want to say who we're working for or or want to hold off on that right now.

00:11:33.60
Sean
uh we'll hold off on the name we'll redact it like well this person yeah

00:11:36.65
Andrew
okay. this person who we're about to start a a product project with that is, you know, the reason for the product hiring we're doing,

00:11:48.20
Andrew
They initially came to crit back in like, I don't even remember when, like 2018, 2019. twenty eighteen twenty nineteen No, it would have been earlier than that.

00:11:57.62
Sean
nineteen twenty twenty twenty no twenty one twenty one and twenty one

00:12:04.46
Andrew
Cause

00:12:05.45
Sean
Yeah, it was the end of 2020. It was the end of 2020.

00:12:08.05
Andrew
they came, they came to us when miscreants, when you were still working on miscreants part-time, I'm pretty sure.

00:12:08.82
Sean
or Oh, shit. That's right. They're the reason I went full time.

00:12:17.87
Andrew
Yeah, yeah.

00:12:19.37
Sean
Yeah.

00:12:19.44
Andrew
They came to to create initially we didn't have capacity to work with them. And so I was like, you and I were had just become friends.

00:12:27.38
Sean
Yeah.

00:12:27.61
Andrew
And I was like, Oh, you got to talk to my friend Sean. He's like, doing the same kind of work we're doing and like getting his agency off the ground. And so, yeah I think if I remember correctly, that project was the project that led to you going full time on Miscreants.

00:12:39.40
Sean
Yeah.

00:12:42.42
Andrew
And then fast forward five years later, I'm now working for you at Miscreants.

00:12:42.40
Sean
Yeah.

00:12:47.92
Andrew
Miscreants is twice the size that CritEver was. And we're getting to, they're coming back and we're working with them together again, which is just super dope.

00:12:54.75
Sean
Yeah. Yeah, those motherfuckers got acquired like three months out of launch, out of much my which which is a terrifying idea of like the thought of like, oh, they're going to pay us on retainer, which means I can quit my job and do this full time.

00:12:59.61
Andrew
Yeah, that was wild. Yeah.

00:13:09.89
Sean
And three months later, so we don't have any work for you because Mandy had bought us out.

00:13:17.13
Andrew
Yeah, but at least at least he invited us to hang out with him in his pool cabana at DEFCON that year.

00:13:18.00
Sean
Yeah.

00:13:22.20
Sean
Right. Right, right, right. His personal cabana. Yeah. No, he's been super good to us the entire time. And i think he was also excited to hear that you were on the project as well.

00:13:34.60
Andrew
Oh, sick.

00:13:35.28
Sean
and also help i mean That also helped build a lot of confidence too.

00:13:35.80
Andrew
Cool.

00:13:38.18
Sean
He's like, yeah, I remember those guys. So this is like a full circle. This is a cool full circle moment.

00:13:42.32
Andrew
This will be a fun one. This will be a really fun one.

00:13:43.66
Sean
Yeah. By the way, and bought a pool table.

00:13:49.56
Andrew
What? you Where are you going to put a pool table?

00:13:52.86
Sean
the me In the basement. In my parents' basement. It's for it's for my parents.

00:13:56.64
Andrew
Nice.

00:13:58.42
Sean
For me to go and play pool.

00:14:02.34
Andrew
I've always wanted a pool table. Pool is one of those things that I've always wanted to be better at.

00:14:04.22
Sean
Yeah.

00:14:06.74
Andrew
i grew up playing with my uncles when we would go visit them.

00:14:09.25
Sean
Nice.

00:14:10.79
Andrew
They had pool tables. And so when we'd go visit them like over Thanksgiving and stuff, they'd teach me how to play.

00:14:12.70
Sean
Nice.

00:14:15.76
Andrew
But I never played consistently enough.

00:14:16.08
Sean
Sweet.

00:14:17.40
Andrew
So I'm very like streaky. Like every now and then I'll hit a streak and I'll feel like a pool god. But the vast majority of the time I can barely string two shots together.

00:14:22.70
Sean
Yeah.

00:14:26.66
Sean
I'm with you. I'm with you. I took a semester of billiards in college for as a gym elective or some elective and still garbage at it.

00:14:37.49
Sean
Still get washed every time I go play at a pool bar by random bully people.

00:14:40.84
Andrew
Yeah. One of my friends in Detroit was talking about playing pool. And was like, oh, that's cool. You play pool frequently. i come to find out she's like in a pool league and has like her own pool cues that she keeps in her car.

00:14:52.55
Sean
Oh.

00:14:55.36
Andrew
And I was like, oh, shit, you're you're like playing pool, playing pool.

00:14:56.34
Sean
Sweet.

00:14:59.62
Andrew
You're not just.

00:14:59.84
Sean
Yeah. No, she's playing billiards.

00:15:02.12
Andrew
the Yeah, probably.

00:15:04.08
Sean
We're barely playing pool.

00:15:05.90
Andrew
I still don't understand the difference because they are different games, right?

00:15:06.42
Sean
know

00:15:09.39
Andrew
Like they're slightly different rules or something.

00:15:12.46
Sean
That's news to me. I thought they were just... I knew Snooker was different. any e snook

00:15:17.39
Andrew
yeah, yeah.

00:15:17.65
Sean
I don't understand the rules, but...

00:15:17.95
Andrew
yeah I also remember when I went over to the UK when I was living there in college, like they had different rules than we did, which was interesting. So I had to learn like UK rules.

00:15:28.98
Sean
interesting

00:15:29.59
Andrew
Maybe they played billiards. maybe they played billiards

00:15:31.97
Sean
maybe that's billiards yeah yeah i do know i think jump shot like when you pop the ball over are illegal in like uk competitive but not in usa competitive yeah yeah scooping jump shots are illegal everywhere yeah

00:15:45.85
Andrew
Interesting. OK.

00:15:55.72
Andrew
I am not good enough to make a jump shot anyway, so it doesn't really matter for my purposes. but

00:16:00.72
Sean
meaning yeah But yeah, I plan to become one of those uncles that are just like secretly good at pool so that when I'm

00:16:07.05
Andrew
Hell yeah.

00:16:10.78
Sean
retired when i'm 65 i'll like hustle my brother's kids or something that'd be great yeah cool uh okay do you want to recap what we missed the more important stuff that we missed out on the second episode i guess yeah but this was big this was huge so yeah

00:16:13.74
Andrew
Yeah.

00:16:20.12
Andrew
That sounds so

00:16:27.90
Andrew
dope. RSA was important. ah Yeah, so...

00:16:34.16
Andrew
Okay, so we are rebuilding Metamonster from the ground up. Kind of, kind of not.

00:16:39.74
Sean
Hell yeah. Mm-hmm.

00:16:41.70
Andrew
Kind of. So the big thing is like, we've been talking for a while on this podcast and offline about one, just this consistent feeling that we're not doing enough right now.

00:16:58.45
Andrew
You know, right now our tool, runs a crawl, and then you can press a button to generate page titles and meta descriptions. You can do those one at a time or in bulk.

00:17:08.78
Andrew
You can tweak them. You can publish them to WordPress, and you can export them to a CSV or export your page content to a CSV or Markdown files.

00:17:18.58
Andrew
Like, that's That's everything our tool does. And while people sort of say like, oh, that's cool, that's useful, I would use that, we're finding that people aren't really using it super consistently. It's just like, it's not enough of a need consistently enough to warrant paying a subscription, and certainly not a subscription that starts at like $50 month.

00:17:38.29
Andrew
yeah We have couple of customers, ah but you know largely not seeing the level of adoption that that we want to see.

00:17:49.37
Andrew
And so my friend Asia Mattos, when I went to hang out with her in Atlanta, put it pretty succinctly, which was like, oh, yeah, you do one job now and you need to do more jobs in order to be able to, do like, be valuable enough to to warrant a subscription. and And warrant, like, there's also the all the other costs of software that we don't talk about enough, which is, like,

00:18:11.97
Andrew
you adoption and you know training and risk, like the risk that you disappear and like you know all this stuff. and So all of those are like these sort of unseen costs that people are are factoring into.

00:18:26.72
Sean
Yeah. Yeah.

00:18:28.26
Andrew
and And we're just not overcoming the costs right now.

00:18:28.46
Sean
yeah

00:18:31.30
Andrew
And we could try to just like slash costs, but that's not really what we want to do. We don't want to build that kind of business. So we want to add more value. Couple this with we've had this long, we've had this feeling for a long time that we're not leveraging AI in the right way.

00:18:51.34
Andrew
And maybe the right way is the, isn't, maybe that's not, maybe the right way is the wrong way to put it. But like, we're we're not getting the most out of AI.

00:19:00.18
Sean
a more useful...

00:19:03.66
Andrew
Yeah. I see the value of AI as like flexibility first and foremost. like It's obviously got accuracy issues. It is not always all that fast.

00:19:15.32
Andrew
its power, and the power of LLMs in particular, is they're reasonably good at a huge swath of tasks. so that flexibility is what makes LLMs special.

00:19:29.42
Andrew
And we are basically kneecapping that flexibility because we are locking it down to this, like, one button in the UI and saying, like, all right, now we're taking the magic of LLMs and we're just going to do this one little thing with it.

00:19:43.34
Andrew
So for a long time, I've wanted to, like, open up the flexibility of LLMs. I've talked to you and talked to some other people floating this idea as, like, chat with your website, you know, custom prompts.

00:19:55.28
Andrew
And the feedback I get is generally, like,

00:19:59.81
Andrew
o cool, that sounds interesting, it sounds cool, but what do I do with it? And so the realization I had a couple weeks ago was one,

00:20:12.99
Andrew
i even though I see this as a major change to Metamonster, I don't need to talk about it like a major change. Metamonster is still the site audit tool that can fix issues for you.

00:20:25.91
Andrew
And when I talk about these new features, I don't need to talk about chat with your website. I don't need to talk about custom prompts and like AI. and just talk about what... That will enable.

00:20:37.08
Andrew
And what it will enable is for us to very quickly prototype out audit features and generation features. So can now... We'll be able to really quickly generate page titles and meta descriptions.

00:20:51.11
Andrew
Find alt text images with missing alt text and generate those. Analyze your H1s and H2s for missing keywords or duplicates. Analyze your site for internal link opportunities.

00:21:04.64
Andrew
Generate a voice and tone guide. generate a readability score for a page and offer suggestions for how you can enhance the readability of a page.

00:21:15.83
Andrew
So there's this huge swath of like use cases that we can enable and I think that's what we need to talk about. So essentially like We're rebuilding Metamonster from the ground up to make the AI way more flexible.

00:21:31.85
Andrew
We've also realized that we were building it like a traditional product. So we were building it just as like a traditional crawler or site audit tool with AI bolted on.

00:21:44.04
Andrew
And with this new focus, we're actually going to build it more like the crawler is just one of a set of tools that the AI can leverage. The export, the WordPress integration, the accounting tool, like we're building this collection of tools that the AI can leverage.

00:22:06.08
Andrew
and an interface to help people analyze the the output and and interact with that output in a scalable and well-organized manner.

00:22:20.03
Andrew
so

00:22:20.03
Sean
Yeah, it's there's more it's it's the flexibility that you can create workflows for yourself to replace whatever you currently do. Right. I think the difference is like workflow replacement versus like workflow automation.

00:22:34.88
Sean
that makes any sense. You're squinting.

00:22:38.64
Andrew
Yeah, I don't know if i workflow replacement and workflow automation sound like the same thing to me. So double click on that. Tell me tell me more

00:22:48.27
Sean
I mean, this is like my product marketing brain switching in, which is like you find a similar thing, but it's different enough. But so it makes someone go, what do you mean by that? In my head, workflow automate, if if I were to frame it, like workflow automation is,

00:23:03.31
Sean
structured, like, you know this is gonna come in and therefore these are the outputs that are gonna come out of it, right? Zapier is workflow automation. You you have a set of use case, like you have a specific set of structured and predicted inputs versus like, it can be anything on a website.

00:23:21.88
Sean
So you don't have to tie it to, you don't have to tie it to like specifically a webhook event or something like that. And then you can replace a workflow that you that you normally have.

00:23:33.79
Sean
and like ah ah Like you have you replace a specific job to be done in your in your like day-to-day work. And you sort of trust AI to sort of do that, if that makes any sense.

00:23:45.20
Andrew
Yeah, I'm still struggling because like automation feels like workflow replacement.

00:23:47.49
Sean
Mm-hmm.

00:23:50.48
Andrew
Like it's the reason you you set up automation is to automate a job to be done so that that job is handled by machines in some way. Right?

00:24:01.61
Sean
Let me workshop it. Let workshop it. this is this is like a This is like a thing I've just been slowly brewing on, which is like the whole, there's this, like I think I sent it to you, like this Harvard Business Review, like people don't buy products because it saves them time.

00:24:08.15
Andrew
Okay.

00:24:15.39
Sean
No one buys, because nobody buys products to spend more time doing their job.

00:24:19.06
Andrew
Yeah.

00:24:19.27
Sean
So how do you like create messaging that ladders down from that of like, it's not like metabonsor saves you time doing SEO is not valuable.

00:24:22.88
Andrew
yeah

00:24:28.83
Sean
Therefore something else has to be. So this is me finding parallels and trying to figure out.

00:24:33.71
Andrew
Yeah. And i I do love the emphasis on workflows.

00:24:34.98
Sean
Yeah.

00:24:37.73
Andrew
Like I just listened to a great episode of Startups for the Rest of Us where were talking about positioning and talking about positioning in terms of workflows instead of in terms of like, you uh pure benefits or or like you technical jargon and like product categories and stuff it's more like this person's process was like identify the workflows that you fit yeah yeah yeah yeah you they they spoke at microconf right

00:25:06.56
Sean
Oh, was that the guys at Fletch? Fletch P&M? Yeah, yeah, they're great. They're awesome. Yeah. Yeah, yeah, yeah. Yeah, they're super smart.

00:25:14.25
Andrew
Yeah.

00:25:15.18
Sean
Yeah.

00:25:15.37
Andrew
Yeah. Their process made a lot of sense to me. i I do think you're on something with like the workflow piece. I'm just not, to me, workflow automation and workflow replacement sound like synonyms for the same thing.

00:25:27.50
Sean
Yeah.

00:25:28.03
Andrew
So I'm not, I'm not, the ah difference difference there isn't clicking for me.

00:25:30.31
Sean
It's good feedback.

00:25:33.20
Sean
Yeah, that's good feedback. The point so to kind of to not pulse on a tangent for MetaMonster or to segue back to that. The thing I think that was cool was, think you're right, you guys were building a traditional SaaS product. There was one button that you clicked on, generated all these things.

00:25:50.04
Sean
do You this specific thing with it, I think. So after after the podcast, we also kind of cut it. I just kind of wanted to jump off so you could show me the demo.

00:26:00.25
Sean
I think the cool thing, do you want to talk about like the thinking behind the UI and how you were going to do it?

00:26:03.03
Andrew
Yeah, so we're...

00:26:06.55
Andrew
yeah so we have been pretty heavily inspired by a product called Clay, which is a lead generation tool with, um tool with, um like, an AI agent so built in.

00:26:07.25
Sean
Okay.

00:26:12.32
Sean
Yep.

00:26:21.24
Andrew
the way Clay works is you run, like, a traditional search for lead gen where they just like pull from a bunch of data sources you and run a bunch of filters. And so you get this like spreadsheet essentially of leads.

00:26:33.14
Andrew
And then you use their agent to their like agent builder to run workflows or automations against that data set. And to us, that feels like a very similar process to how SEOs work. Like SEOs generally like pull content from a website and that is their data set.

00:26:55.44
Andrew
And then they are running, they're doing various things with that data. So they're performing optimizations, they're doing analysis and they're, all of those things are, could be considered kind of enrichments on the data set and just like,

00:27:10.01
Andrew
little bits of information you're adding or editing with the data set, which also coincidentally is like how data scientists think about things and think about working with AI.

00:27:10.01
Sean
yeah. Yeah.

00:27:21.62
Andrew
It's like you've got a data set and then you're running that data set against a series of prompts and then you have evals to test like how good, how reliable and effective the output is.

00:27:32.88
Andrew
it makes a lot of sense in this like sort of new AI paradigm of product building and I think has us really excited. And so what I've realized is we're really building, like, an SEO agent builder.

00:27:48.89
Sean
Yeah, I mean, clay for ah ah SEO, I think is where we landed on at the end.

00:27:51.65
Andrew
Yeah. Clay for SEO, ah SEO agent builder, like, and it's going to feel kind of like, you know, a spreadsheet with, like,

00:27:53.64
Sean
Yeah.

00:28:02.75
Andrew
some fairly sophisticated tooling for running analysis and generating new fields on on that spreadsheet.

00:28:05.93
Sean
Mm-hmm.

00:28:12.91
Sean
Yeah. Yeah. I mean, I think it's sick. So that was last week after the demo.

00:28:14.95
Andrew
Yeah.

00:28:17.24
Andrew
Yeah.

00:28:19.98
Sean
Where are you now?

00:28:22.35
Andrew
We have a prototype.

00:28:23.82
Sean
Okay. Sweet.

00:28:24.58
Andrew
We have a... we are Austin is polishing up some of the details on a little proof of concept. So, can't generate rows in the in the spreadsheet yet, not doing any of that, but he has started working with...

00:28:41.66
Andrew
building out a little like AI chat widget inside of Metamonster. And like at first you hear ai chat widget widget and you're like, okay, that's not that big a deal. You're taking content and and throwing it against like an LLM.

00:28:55.74
Andrew
But we're building a lot of the little components that will ultimately power the more sophisticated agentic flows. And we're using this really cool new set of tools for this, Vercel's AI SDK, which makes it really, really simple to like validate output from an LLM.

00:29:17.00
Andrew
So like require output in a structured format and make sure it fits that structure. makes it really possible to define tools that the LLM can use. So there's... like all this talk about MCP servers right now in the AI space, right?

00:29:32.13
Andrew
MCP servers are like one way of providing tools to that an LLM can call where you can run some like traditional code for things that LLMs aren't that great at, like math and, you know, querying a database or, you know, whatever. Things that traditional code is good at that you don't need LLMs for and that LLMs are bad at.

00:29:54.86
Andrew
and MCP servers are like one way to expose tools to LLMs, but the AI SDK also lets you do like a version of that locally.

00:30:05.96
Andrew
And so we'll have, like, our web crawler will kind of be a tool. I don't know exactly what the workflow is going to be yet, but, like, our web crawler might be a tool.

00:30:16.75
Andrew
The, like, the WordPress integration might be a tool. We have a tool to, like, count things. So, yeah. Austin's been building out some of these tools.

00:30:28.51
Andrew
He's also like the other piece of this that's going to, that's pretty cool. And, and is another piece that doesn't exist in like your default chat GPT or whatever, like chat GPT can in theory, like with a deep search, like pull a page on your website and stuff like that.

00:30:49.65
Andrew
But, Because we have already pulled all the content from your website, we're also going to convert that to embeddings and store those embeddings in a vector database and be able to feed that to the AI so the AI can run a vector search.

00:31:07.56
Andrew
you like can run a vector search based on your prompt, find the most relevant content from your website, then take the text from that content, feed it to the LLM, and we'll create, that'll be like a key part of the agentic flow. And that is ah ah that is important for getting the right information to the LLM and also not like Even though context windows are getting large with websites, you can still blow past them pretty fast and large amounts of content are also slower to process and et cetera.

00:31:38.20
Sean
Yep.

00:31:43.09
Andrew
And so it also helps you like streamline everything. that like, That embedding process is something that, you know, like, Cursor does for you when you're when you're using Cursor to your codebase.

00:31:56.13
Andrew
I haven't seen people talking at least about doing that with, you know a website for marketers.

00:32:02.96
Sean
Yeah, i mean, the closest we have gotten is when you added the exporter's markdown. you know, I would scrape a site, pull it down, drop it into clod as an artifact, create clod project, and drop it in so that it creates its mini little, like, I think it's like a vector embedding.

00:32:19.78
Andrew
It might be doing that, yeah, yeah.

00:32:19.86
Sean
i don't know. Doing something with rag.

00:32:21.90
Andrew
It might be doing embeddings on like all of the content of a project.

00:32:22.07
Sean
Yeah.

00:32:25.91
Sean
yeah

00:32:27.17
Sean
Maybe. But yeah, that's the closest that we've gotten right in. That has its own limits of can't be this too large of a file, can only upload 25 of them, et cetera.

00:32:37.24
Sean
So I think, I mean, it's cool. I'm excited.

00:32:39.17
Andrew
Yeah. We're also creating embeddings of the raw HTML and the markdown so that we can do analysis on the HTML.

00:32:44.77
Sean
Cool.

00:32:48.92
Andrew
So we can do analysis on like links and tags and you know the technical details of your website and and pages on your site.

00:32:57.67
Sean
Yeah, yeah, especially for like SEO, you know, i' wanting to know which ones are, like if i'm missing an ARIA label or something doesn't have alt text, something like that.

00:33:07.10
Andrew
Yeah.

00:33:07.87
Sean
the cool The doubly cool thing is that this is, I think, well-timed for figma Figma's site thing.

00:33:16.77
Andrew
Oh, that's interesting.

00:33:18.12
Sean
because yeah Since last week, I sent half of my designers to config after we went to RSA, and they were there the conference at the conference when they announced Figma Make, Figma Draw, Figma Sites, all that sort of stuff.

00:33:32.50
Andrew
Yeah.

00:33:32.59
Sean
Anyway, it's a cool new... web builder that enables more people to build sites, which I think just kind of blows the market up a little bit more for you guys too.

00:33:44.97
Andrew
Yeah, I'm really excited. We should have it under a feature flag in the next day or two. And so I'll ah ah give you access, at least on your your personal account, you can start playing around with it.

00:33:57.08
Sean
Sweet.

00:33:58.38
Andrew
It's pretty buggy right now because yeah it's just a proof of concept.

00:34:01.23
Sean
Mm-hmm.

00:34:01.87
Andrew
but like Austin was was testing it and he was like running it against the small efforts website and he was like, tell me all the like episodes where Andrew or Sean mentions Rob Walling and it's coming back.

00:34:20.30
Andrew
So you can do like this really interesting querying and analysis sort of stuff with it that just isn't possible with like Screaming Frog.

00:34:26.26
Sean
Yeah.

00:34:29.79
Andrew
Yeah.

00:34:30.32
Sean
Yeah, definitely.

00:34:31.88
Andrew
So I think it's going be really important for differentiating.

00:34:32.06
Sean
Love it.

00:34:34.95
Andrew
i don't know I still don't know exactly how we're going to position it. I mean, I think i to start, it's just going to be like a site audit tool that fixes things for you.

00:34:44.86
Sean
Mm-hmm.

00:34:45.63
Andrew
And then just talk in our marketing about like some of the analysis that we can do that other tools can't do.

00:34:51.76
Sean
Mm-hmm.

00:34:53.09
Andrew
and I, like, we still feel like we want to stay in that analysis space space as much as possible. Like, Aerox is another one of our competitors and someone we've been inspired by and they have a similar, like, grid concept and workflows and, like,

00:35:05.95
Andrew
They're honestly like a more powerful, more complex version of what we're building.

00:35:11.19
Sean
Yeah.

00:35:13.12
Andrew
But I think the flip side is, one, I think there's value in being easier to set up, easier to get started with. But they're also like all of their positioning is focused around content generation at scale.

00:35:25.93
Andrew
And, know, I'm also seeing a shift in like the way marketing thought leaders are talking about marketing. And like people are kind of talking about, you know, a shift away from generating massive amounts of content and a shift towards, you know, brand based marketing, social based marketing, like, and, and so I think we're going to try to stay in more of the analysis space.

00:35:54.01
Sean
Mm-hmm.

00:35:54.20
Andrew
Like, we can ingest your website, maybe eventually we can ingest your social accounts and different things. And we help you run analysis and like, and, and

00:36:04.48
Sean
Yeah, I mean, I think it's the I think it's a better angle and to kind of touch on the point about like marketing leaders and whatnot. You know, I think because we're out of that Zerp era, because we're now in like expense management era land, i think like content marketing at scale is attractive at first because like you get ah ROI through to two places, right? You either make money or you save money.

00:36:27.55
Sean
and and and when you want to save money you choose like different thing places to kind of save and when you generate content marketing at scale you save money of hiring outsource for like writers to do but it's like to me it sounds like content slop at scale like content seo slop at scale which if that is the strategy that you know ah ah sure i guess but going off of what you're saying of like marketing leaders sort of talking more about like brand or demand gen or like high value traffic on the site like that's more towards towards that sort of um like make site better so it makes more money less make content easier like mass content easier to write and so i think i think the analysis part i think there yeah i mean i think there is something there to it it's yeah yeah cool sweet sweet

00:37:11.10
Andrew
Yeah.

00:37:20.94
Andrew
So I'm feeling pretty hyped. You know, long way to go. Long, long way to go.

00:37:26.00
Sean
Yeah.

00:37:26.65
Andrew
this is taking on, this an ambitious direction.

00:37:30.63
Sean
yeah

00:37:31.69
Andrew
But, yeah know, I also think that, like, you know, I just was listening to an interview ah ah podcast with Brian Castle and Justin Jackson, and Brian was arguing that, like, you know,

00:37:45.45
Andrew
AI enables Bootstrap founders to be a little more ambitious than we used to be. And I think there's some truth to that. So, yeah. It's going to interesting.

00:37:55.01
Andrew
We'll see how it goes.

00:37:57.33
Sean
Cool. Sweet. yeah I mean, I think the same feedback I gave you offline, but I think this is like multi-million dollar SaaS startup.

00:37:58.59
Andrew
Yeah.

00:38:06.98
Sean
Wow. Like, like.

00:38:08.36
Andrew
I hope, man. I hope we're on something. i am in the process of writing a blog post. I've i've got the rough draft written about like the validation process we followed for Metamonster.

00:38:19.00
Andrew
And like i the title, working title, is like, This Time Feels Different.

00:38:25.21
Sean
Nice. Nice.

00:38:26.32
Andrew
But at the same time, like as I was writing that, like one of our four customers canceled this week. And so I had a little bit of a moment of

00:38:33.07
Sean
I see.

00:38:36.86
Andrew
Does it feel different? or Or are we still fucked?

00:38:39.85
Sean
Uh-huh.

00:38:41.44
Andrew
are we still Are we still building something no one wants or cares about? But no, i think i think people have I think people have been engaging with us a lot more, even though we haven't just magically hit on the right product and the right feature set and everything out of the gate.

00:38:58.94
Sean
Hmm.

00:38:59.54
Andrew
I think people have been engaging with us more talking to us we have a couple of customers that's different than chart juice ever was so uh it does feel different even if the signals aren't all like blaring you're doing the right thing just yet yeah

00:39:08.52
Sean
Yeah. Yeah.

00:39:18.90
Sean
Yeah. Yeah. Makes sense. Cool. I'm excited for you to work on this product project, by the way, because I think given the fact that it's an AI project as well, there's going to be some, I think there's going be crossover and interesting learnings.

00:39:28.26
Andrew
mm-hmm

00:39:34.23
Sean
The client that we've spoken about has gone pretty deep on some other ideas as well.

00:39:35.33
Andrew
Yeah.

00:39:38.99
Sean
And I think, I think it'll be a good like extra project.

00:39:39.84
Andrew
That's cool.

00:39:42.69
Andrew
I honestly wish I could like redesign Dreadnode from the ground up, knowing what I know now.

00:39:49.62
Sean
don't say Don't say that. don but Everyone loves Dreadnode.

00:39:53.25
Andrew
Everyone does love Dreadnode. And i'm and like I think we did a solid job, but I also know so much more and understand the space so much better now. And I'm like, oh, but what if we could just start fresh?

00:40:04.87
Sean
You know, I saw your ex-roommate at the Dreadnode party.

00:40:09.07
Andrew
Oh, Morris was there?

00:40:10.00
Sean
And yeah, the first thing he said to me was, you guys really outdid yourselves with this one.

00:40:16.15
Andrew
Hell yeah.

00:40:16.85
Sean
And I was like, all injured. All injured. And he's like, it's like, makes sense. That checks out. So sweet.

00:40:24.41
Andrew
Cool, man. how's How's he doing?

00:40:25.67
Sean
and He's good. He's good. I think he's enjoying not being CEO.

00:40:30.57
Andrew
Yeah.

00:40:30.68
Sean
So yeah.

00:40:31.66
Andrew
Yeah.

00:40:32.50
Sean
Yeah.

00:40:32.69
Andrew
It's kind of a shitty job. Yeah.

00:40:34.94
Sean
Yeah, it's awful. I think of how shitty it is every day right now. No, no. its i think it's it's I think it's better without VC bosses.

00:40:48.90
Andrew
Yeah, 100%.

00:40:49.28
Sean
Yeah. yeah

00:40:52.18
Andrew
How are things going with your product? What's what's new on margins?

00:40:56.66
Sean
So, making progress.

00:40:57.68
Andrew
I almost said miscreants. The fact that you have two M name and we're also an M name, like we gotta we gotta do better next time.

00:41:00.99
Sean
I know. Yeah.

00:41:07.44
Andrew
No, I actually love the name margins.

00:41:09.41
Sean
Thanks. Thanks. Thanks. I think it's good. I think it's, we can all have the same logo. It should be eminent circle. margins is good. We dev, our dev came back from vacation.

00:41:19.53
Sean
He's been jamming on a little bit more, you know, ways to go. i think, uh,

00:41:24.74
Andrew
Oh, oh, the website resolves.

00:41:28.86
Sean
Oh, no.

00:41:29.40
Andrew
Margins.so resolves to an actual site now.

00:41:29.54
Sean
don't Don't tell the world that. Oh, my God. that's because i was That's because I was testing something else out. It's not even done. I was ah ah i was trying to test.

00:41:37.61
Andrew
You mean link one, link two, like link three, link four, link five is not the final footer copy?

00:41:37.79
Sean
try to

00:41:44.02
Sean
No, no, no, that is, that is, those are, those are secret Easter eggs. and People click on them.

00:41:49.23
Andrew
Nice.

00:41:50.06
Sean
Uh, no, i I published that cause I was testing out a thing I was vibe coding and I needed like a test site to mess around with. was like hosted somewhere and didn't want to do with the miscreants or, well, I want to do more than one site.

00:41:58.26
Andrew
Yeah.

00:42:02.87
Sean
um um uh margis is going well i think you know i'm anxious on timeline given the fact that uh webflow is deprecating their editor in about a month from now so thinking long and hard about that might have to just sit with the fact that it's not going to be fully ready for the world by then but maybe that's okay because i don't think anyone's going to solve the problem the way we are i think

00:42:04.03
Andrew
Mm-hmm.

00:42:30.60
Sean
we have a we have a i think similarly to you and air ops we have also you know some other incumbents in the space like story chief and contentful except the difference is that you know storage even contentful are meant to be like content marketing operating systems where you put your social media content, all this sort of stuff inside of this thing. And then like publishing to a CMS is a second class experience.

00:42:54.31
Sean
And like writing is okay, but it's not the same, right? Like where we're prioritizing the ability add tables or to do all these other things that, whereas for those, it's more like, you know, let's not add additional features to your to your CMS or additional whatever sort of things. So that's my parallel.

00:43:15.31
Sean
yeah i mean long ways to go internally we've kind of phase one is have to get it to a point where internally we can use it for our clients and publish things phase two is that our clients can start using it and phase three is it's ready for public consumption so long ways away from phase two at the moment it's still early early alpha stages but I mean, hopeful.

00:43:39.93
Sean
It feels better. I think everyone's reaction... I showed Austin other day on a call and for for a separate thing, but I think his reaction was that, you know, this like looks much better than Webflow CMS.

00:43:54.45
Andrew
Hell yeah.

00:43:55.51
Sean
I think the fact that Figma Cites is launching and going to launch its CMS soon, and having seen some mocks of that CMS, like...

00:44:07.09
Sean
ultimately this being able to publish the Figma sites, framer, and webflow is just like how we'll be able to grow and there's just a lot of opportunity.

00:44:16.39
Andrew
Yeah.

00:44:17.12
Sean
So excited for that, you know? Yeah, that's, let me think. That's about it.

00:44:26.21
Andrew
Makes sense.

00:44:26.26
Sean
I think, yeah, yeah. But yeah, there's like half a website that's live now. So, yeah.

00:44:33.68
Andrew
Can follow along with the process.

00:44:35.76
Sean
yeah Yeah, it really is building in public because I just published it.

00:44:39.03
Andrew
Are you going to throw up like a waiting list or something at some point or.

00:44:44.91
Sean
yeah, yeah, I think the CTA will probably be a waiting list. Yeah, again, it was like half built before I literally took it live because I was trying to build, I think I told you about this like version control thing for Webflow.

00:44:52.18
Andrew
hmm.

00:44:56.16
Andrew
Yeah, yeah.

00:44:56.69
Sean
thought it was like vibe coding. yeah once a code base gets large enough lovable and bolt really just start to struggle like i can feel them getting dumber you know they start like not understanding things i'm saying like things i referenced before but it makes sense to pull out cursor to help solve some other things

00:45:05.34
Andrew
Yeah.

00:45:13.78
Andrew
Mm-hmm.

00:45:16.57
Andrew
I was about to say, have you tried like taking a Lovable or Bolt project and pulling it into Cursor and seeing if it does

00:45:21.74
Sean
yeah Yeah, it it helps solve some of the problems better. I think at some point the the the context window becomes so large, I think, that Lovable gets into this like recursive...

00:45:33.94
Sean
like It can't solve the problem because it can't think out of the single file that caught... I ran into this problem, like this type instantiation problem, and Lovable could not fix it.

00:45:45.06
Andrew
hmm.

00:45:47.42
Sean
And then other problems, because I think... they have a tendency to like not really use React hooks and use effect very well.

00:45:57.28
Sean
And but you know I think going back to like context window, not large enough. So it starts only yeah it starts getting lazy and like looks at only a small segment of the code.

00:46:07.49
Sean
So it only thinks about that and it doesn't think about um everything and to pull it into cursor you can fix that problem and republish fix it but yeah it feels like two steps forward one's three steps back sometimes but it the the the coding with lovable takes so long now i've just been watching sopranos while um

00:46:31.88
Andrew
I mean, there is something nice about that workflow sometimes where you, like, if you give AI, like, a bigger task and then you just walk away and then come back to it and then, you know, provide it some critique and then, you know, go back and forth.

00:46:44.95
Sean
Yeah. Yeah.

00:46:48.02
Andrew
That's how, whatever happened to that, there's that thing that everyone was, like, freaking out about that was supposed to be, like, the developer killer.

00:46:48.23
Sean
Yeah.

00:46:55.57
Andrew
It was called, like, Devin.

00:46:57.21
Sean
Devin. Devin. Yeah.

00:46:58.83
Andrew
Devin.

00:46:59.05
Sean
I mean, yeah, it turns out that demo...

00:46:59.36
Andrew
Yeah, yeah, yeah. It, It worked that way, i think. Right? Like...

00:47:04.94
Sean
uh yeah yeah yeah you could walk away for a full day and come back later on yeah i mean it turns out that that demo was faked so that's that's exactly what happened to it

00:47:05.03
Andrew
Mm-hmm.

00:47:14.41
Andrew
Classic.

00:47:17.31
Andrew
Yeah. So funny. I remember everyone was, like, freaking out. Like, Devin is the end of developers. And then I haven't heard anything about it in six months. Yeah.

00:47:26.65
Sean
Yeah, yeah. I think a lot of the engineers also then like did a little whistleblowing and were like, yo, like it's it's a mess at the moment. I'm sure they'll improve it. You know, I think great video for what it's worth.

00:47:41.04
Sean
Like excellent marketing team. good Great job. Great demo. Yeah. I mean, that was like that was like pre 3.7 Sonnet or 3.5 even, right? Like I'm sure i'm sure it's gotten much better now, but Can't swing 500 bucks a month at the moment yet.

00:48:01.01
Andrew
No, that's nuts.

00:48:02.32
Sean
Yeah.

00:48:02.93
Andrew
I'm just trying to get people to put down a credit card for seven-day free trial for $50 a month.

00:48:08.83
Sean
Yeah.

00:48:09.13
Andrew
And no one wants to do that. So, yeah.

00:48:10.83
Sean
Yet, yet, yet.

00:48:12.98
Andrew
By the way, I dug into our Google Ads a little bit more. And I think part of the problem... So, bunch problems. But...

00:48:21.54
Sean
Sure.

00:48:22.07
Andrew
I do think one of the problems with our marketing right now is that the majority of the people hitting our website, even though we're getting this crazy high, did I tell you this? we're We're 27% of people who are coming in from Google ads are clicking the start free trial button.

00:48:39.48
Sean
Oh, yeah, you did you did show me via the postdoc analytics.

00:48:41.68
Andrew
which feels nuts to me that feels so high but then less than one percent like less than half a percent are getting through to actually like creating a trial account and uh so i think that's like largely a pricing mismatch i'm still a little surprised by how many people drop off at the like create account step unless my messaging is to

00:48:44.30
Sean
Yeah.

00:49:05.04
Sean
Well, you still ask for a credit card in first, right?

00:49:07.62
Andrew
So, no, it's create account, out then like pick your pricing plan, input your credit card, and then... and then Yeah, yeah, they're not even like creating the account to get to the pricing page.

00:49:15.95
Sean
I see. They're not creating accounts.

00:49:22.48
Andrew
Although I do try to, like in the microcopy there, make it clear that you're going to have to put a credit card in, so maybe that's part of it, maybe that's working. Yeah.

00:49:30.75
Sean
Maybe.

00:49:31.55
Andrew
But the other thing I just noticed was like the people who were creating accounts, even though they weren't putting credit cards in, they were all like personal email accounts. It was all like so it's such and such 1234 at gmail.com.

00:49:44.07
Sean
Hmm.

00:49:45.15
Andrew
And even the business accounts that we're getting through were mostly not agencies. They were, they seem to be more like little startups or small businesses or media companies or something like that.

00:49:57.13
Andrew
And so I think that like, while there's volume for the keywords that we're currently targeting with our ads, I don't think it's SEOs necessarily searching those keywords.

00:50:08.91
Andrew
Like, you know, we get a lot of volume from ai meta description generator or something like that. And if you think about it, like what SEO is searching AI meta description generator?

00:50:23.09
Sean
Yeah, yeah. I mean, yeah, I mean, that like, that's, ah SEOs and, like, agencies are more B2B buyers. They don't buy in those ways.

00:50:34.37
Sean
Mm-hmm.

00:50:35.07
Andrew
I think they still buy from, from Google search and from.

00:50:39.56
Sean
Oh, sorry. Those, like, that keyword, like, some of the more higher volume keywords like that.

00:50:42.05
Andrew
Yeah. Yeah. yeah

00:50:44.88
Sean
I think there's, like, more specific things they look at, you know?

00:50:47.17
Andrew
Yeah. Like my guess is screaming frog alternative is probably better. You know, that's another one that gets a decent bit of volume.

00:50:51.82
Sean
Mm-hmm.

00:50:54.99
Andrew
AI SEO tools maybe is probably a mix of like crap and, but maybe some, some SEOs.

00:51:02.08
Sean
There's maybe, I think if if I'm an ah SEO, I'd probably maybe would look at more like try to search like geo related things like generative AI SEO related stuff.

00:51:03.40
Andrew
It,

00:51:11.12
Andrew
Well, I mean, we're not a GEO tool, so that doesn't do any good for us.

00:51:17.08
Sean
All, all geo is just a SEO in my opinion.

00:51:19.61
Andrew
I agree with that. But then if you know that, why would if you're an SEO and you think all GEO is SEO, would you be searching GEO tools?

00:51:21.03
Sean
Yeah. Hmm.

00:51:28.99
Sean
I don't think it's necessarily that common yet, but yes, you're not a geo tool. I was just saying that like, that would be maybe something an SEO person researching. also don't think every ah ah SEO has thought about something like that yet.

00:51:44.76
Andrew
has made the connection that all GEO is just SEO.

00:51:47.91
Sean
Yeah. Yeah. Yeah. mean, I don't know.

00:51:50.77
Sean
I don't know that. I don't know that many ah SEOs.

00:51:52.74
Andrew
It might also be that we like have to get in with influencers and stuff in order to really get in front of professional SEOs.

00:52:01.30
Andrew
It might be that they're not searching for new tools on Google on a regular basis. So actually my next project project ah While I am doing interviews for Miscreants, I'm also going to be doing some user interviews for Metamonster to try to identify the buying process for most for like SEO agencies and learn more about that.

00:52:17.53
Sean
cool

00:52:23.29
Sean
sweet sweet yeah how much of your know we're almost at time here what is your data just out curiosity

00:52:24.21
Andrew
Yeah.

00:52:35.19
Sean
What is outside of miscreants, right?

00:52:37.18
Andrew
Yeah.

00:52:37.98
Sean
What do you do day to day? Yeah.

00:52:42.35
Andrew
but

00:52:47.77
Andrew
That's a great question. and don't always know the answer.

00:52:50.15
Sean
Yeah.

00:52:52.95
Sean
gosh

00:52:54.48
Andrew
It's been, yeah, the last like month and a half has been weird. So like normally my, I try to split my time when I have miscreants clients, I've been trying to split my time, like roughly half miscreants work and then half growth marketing work for Metamonster.

00:53:12.90
Sean
Mm-hmm.

00:53:13.40
Andrew
And so, that, you know that copywriting, it's managing our Google ads, managing our cold email outreach, ideally scheduling demos and running demos with people, talking about Metamonster on LinkedIn, you know, reaching out to friends to see if I can show them Metamonster, get their feedback, you know, trying to get introductions.

00:53:39.17
Andrew
You know, I've got a long list of things I want to do it like that, like, you know...

00:53:42.88
Sean
Mm-hmm.

00:53:44.67
Andrew
Tweaking our pricing. i did tweak our pricing a little bit one day, but like there's a lot more to be done there, although I don't think it's like super valuable to do that right now when we're building towards this new future that's going to be pretty different from what we offer right now. So I don't want to over-optimize our pricing for what we offer right now.

00:54:05.67
Andrew
you know There's also like building, you know I want to build some free tools, I'll eventually do a bunch of that kind of stuff. ah you Writing content both for my personal newsletter and for miscreants sorry the site.

00:54:17.30
Sean
Yeah. Yeah.

00:54:23.57
Andrew
Yeah, the M's are gonna fuck with me.

00:54:25.56
Sean
yeah

00:54:27.39
Andrew
So like, yeah, my ideally my focus is like growth marketing for growth and marketing. And I say growth because it's like, you know, it's not just traditional marketing. Like I will probably write some code. I'll probably do some sales. I'll probably, know, contribute on the product side from time to time.

00:54:52.12
Andrew
Certainly spend some time each week you know talking through product decisions with Austin and relaying feedback, what I'm hearing from people to him. And so, yeah like growth to me is a more, you know, encompasses a more holistic like marketing approach that includes like tweaking product stuff to try to impact your funnel and all that jazz.

00:55:17.59
Sean
Gotcha. Okay. Helpful.

00:55:19.69
Andrew
Yeah.

00:55:20.78
Sean
for Thanks.

00:55:21.70
Andrew
Yeah. Yeah.

00:55:23.04
Sean
Yeah.

00:55:24.05
Andrew
But in reality, I've also like took a four day weekend to go hang out with my mom and, uh, you know, have had some pretty lazy days where I did not get started until halfway through the day.

00:55:37.56
Andrew
So it's been, uh, yeah, it's been nice, but I also like,

00:55:39.82
Sean
mean That's pretty sweet. That's pretty, yeah.

00:55:45.70
Andrew
i don't want to get too used to, you getting my day started at noon. Or like there were some days where I just did like house tours all day, which Maddie very happy, but also probably not something I'll be able to do forever.

00:56:00.15
Sean
right Yeah.

00:56:03.23
Andrew
Or maybe it is. Maybe that's how i yeah I am trying to structure my life in a way that is like maximum flexibility.

00:56:11.98
Sean
Yeah, I just have... a growing sense of an existential crisis that's like i move from like zero to one founder or two like you know even even with even with uh with margins right like like i'm so used to the idea that if i'm starting something i have to be the one to build it and this is like you know thinking like an owner

00:56:16.97
Andrew
Yeah.

00:56:31.84
Andrew
Mm-hmm. Mm-hmm.

00:56:36.67
Sean
And yeah. And like, even with miscreants, like as more people take more work off my plate, like I literally asked JJ one of the calls a couple months ago, like I was like, after we get with those 30 people, like after we have enough, we have critical mass of people.

00:56:54.35
Sean
It's just like, what the fuck is my job? What am supposed to do? Cause like, I don't know. It's not like sales is the answer because at some point I would just have someone do sales.

00:57:05.59
Andrew
I mean, the goal for miscreants has always been to free up your time and give you the resources to pursue other creative projects like margins and the apparel and all of that, right?

00:57:18.16
Sean
sure yeah yeah definitely it's just like

00:57:22.98
Sean
i don't know it's just like a it's a different it's a different mindset and yeah

00:57:27.60
Andrew
Yeah. Yeah, it is for sure. And I feel like it's so hard for most founders to let go of the feeling that they have to be doing all the things.

00:57:39.72
Andrew
And if you don't let go of that, then you're never able to transition to the goal state that you had at the beginning of like having your time be freed up to do whatever you want. So...

00:57:51.06
Andrew
so Yeah, ideally too, you're like mentoring and like writing and you know providing you

00:58:01.10
Sean
Yeah.

00:58:01.51
Andrew
you know education and support, right?

00:58:01.69
Sean
Is is every...

00:58:05.62
Sean
Sure. I just hear writing. I hear writing a lot. I hear writing come up. I feel like writing comes up a lot with with folks who get to this sort of stage.

00:58:14.65
Andrew
Yeah, because I think at a certain point, like, you want to, if you want to set the company vision, if you want to set, if you want to like, we talked about in the episode, one of the things we talked about last week in the episode that didn't make it out was, you know, I've been wanting to try to slowly start to make the transition in my own writing from,

00:58:17.11
Sean
Mm-hmm.

00:58:28.08
Sean
Mm-hmm.

00:58:38.10
Andrew
like just sort of journaling almost like vlogging, blog, like personal firsthand experience to more like analysis and, you know, insight.

00:58:49.76
Sean
Yeah. Yeah. Yeah.

00:58:52.51
Andrew
And, you know, I, I want to be careful about doing that because I don't want to like act like I know more than I do. But I think part of, you know, if you look at some of the,

00:59:04.01
Andrew
and And a little bit, this is like the agent this is like selection bias because the agency owners who are public are going to be going to write more.

00:59:10.96
Sean
yeah

00:59:13.90
Andrew
So like Peter Kang, who's, fuck, why am I blanking on that guy's name right now?

00:59:23.60
Andrew
Who's the guy from the SEO guy? Yeah.

00:59:26.54
Sean
Oh, MP Digital? Neil?

00:59:29.19
Andrew
No.

00:59:29.82
Sean
Or Crazy Egg?

00:59:31.92
Andrew
Nope, nope, nope. The one in Philadelphia, there they're all over the country, but... shoot. Why am I spacing on his name right now?

00:59:43.10
Sean
The SIFT guys?

00:59:44.64
Andrew
No.

00:59:46.28
Sean
have no idea. don't know.

00:59:47.15
Andrew
there's There's so many SEO agencies. Dude, this guy is always in my... Here, Will Reynolds, Seer Interactive.

00:59:55.80
Sean
That's right.

00:59:56.61
Andrew
like Will is constantly writing and stuff. And, you know we talked about with you, like, you know, one of the bottlenecks at miscreants is, is transferring the industry knowledge that you've built up over the years to other people. And I think like the most effective way to do that when you have 30 people to do that with is not like through one-on-ones it's through writing.

01:00:21.37
Andrew
Even at Grey Noise, like Andrew was trying to do this where, you know,

01:00:24.46
Sean
Yeah, fair.

01:00:25.41
Andrew
you know, CEOs spend a lot of time, like, writing memos. They spend a lot of time talking to people and in meetings and, like, thinking and reading and, like, digesting information. And then usually the best way to, like, organize those thoughts into something useful for your team is through memo, a blog post, ah you know, some sort of written communication, I think.

01:00:50.59
Sean
yeah fair

01:00:50.89
Andrew
Or a presentation. You know, some people probably do it through, you know, presentations.

01:00:57.26
Sean
yeah

01:00:58.39
Andrew
Yeah.

01:01:01.03
Sean
yeah fair fair fair sounds fun sounds a good time

01:01:07.21
Andrew
But also, if you don't want to do that, you don't have to. Find someone else to do that part and go do whatever you want to do.

01:01:16.89
Sean
yeah um i mean you know could just go play pool with my new pool table coming on monday yeah it's guys gotta go disassemble it

01:01:23.05
Andrew
Sure.

01:01:27.75
Sean
ah well well maybe maybe by season five i'll let you know what uh i'll have more to say about this on the podcast

01:01:36.64
Andrew
Yeah. You could also follow the Peter King route and just start buying agencies and then then you've got more work to do.

01:01:45.11
Sean
yeah yeah

01:01:45.74
Andrew
But I think the like actual day-to-day work that Peter does is a lot of like meetings and writing and thinking.

01:01:53.90
Sean
i think so yeah i think so i think i'm sure he does thinking i'm sure he does all of that it's more the fact that like for the folks that do right and going back to the selection bias it's like the folks that are doing essentially content marketing right when they publish those thoughts like those ah ah those things

01:02:16.46
Sean
Like, I know. I got to think about this more. I got to, I got to.

01:02:19.84
Andrew
Writing is just also, though, a good way to organize your thoughts and into something coherent.

01:02:23.38
Sean
For sure.

01:02:24.81
Andrew
And so I think, like, a lot of great CEOs, even if they're not writing, like, there's a big difference between content marketing and just, like, writing.

01:02:35.80
Andrew
And so even if you're not, even if people aren't writing publicly, I think most effective CEOs are probably doing some amount of internal writing.

01:02:36.15
Sean
Mm-hmm.

01:02:47.02
Sean
Yeah, yeah.

01:02:47.25
Andrew
Like, writing processes, process documents, writing memos that like on here's the direction we're going. And again, like writing is just one form of communication. There are other forms you can record videos or you know you can give live presentations or yeah there's different ways to like communicate things.

01:03:02.35
Sean
yeah

01:03:08.85
Andrew
But I think writing is still like one of the most effective ways to communicate things to a large group of people and like get people and it's easier for people to refer back to and and everything.

01:03:20.04
Sean
Yeah.

01:03:21.11
Andrew
so

01:03:21.18
Sean
Yeah. No? Fair. and Fair, fair.

01:03:26.18
Sean
Okay, anyway. Over the hour. Anything else?

01:03:28.93
Andrew
Yeah. No, man. We covered a lot.

01:03:31.53
Sean
Headed out. Cool. Cool, cool, cool.

01:03:34.23
Andrew
Oh, hey.

01:03:34.35
Sean
I'm going to... Yeah. yeah

01:03:35.73
Andrew
when are When are we going get some miscreants hats?

01:03:38.66
Sean
We do have Mr. Intense.

01:03:40.12
Andrew
Oh.

01:03:41.09
Sean
Yeah, they're sold out, though.

01:03:42.45
Andrew
Damn it.

01:03:42.60
Sean
think I got... Yeah, think I have an extra. Okay.

01:03:46.47
Andrew
I want a miscreants hat.

01:03:48.18
Sean
If we have an extra, I'll send you one.

01:03:50.06
Andrew
Sick.

01:03:51.05
Sean
Yeah.

01:03:51.26
Andrew
right.

01:03:51.41
Sean
it's like It's a captcha.

01:03:53.98
Andrew
Ah, Ah, that's right. That's right. i do remember seeing these now. Yeah, okay.

01:03:57.56
Sean
Yeah.

01:03:57.91
Andrew
I see it. Sold out.

01:03:59.99
Sean
Yeah.

01:04:01.28
Andrew
Cool.

01:04:02.34
Sean
Cool. Peace.

01:04:04.07
Andrew
Later, dude.

01:04:05.16
Sean
Bye.