Andrew is in Atlanta visiting friends and continuing to think about how to best integrate more AI features into MetaMonster to make the product stickier. Sean had his first demo of Margins and... everything broke! But that's okay!

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

Transcript:
00:00.93
Sean
This is a new background. Where are you?

00:07.44
Andrew
Oh, you can tell how well my brain is working today. Good Lord.

00:11.24
Sean
Yeah.

00:12.58
Andrew
i am staying with one of my college roommates in Atlanta. And after four years away from the South, my body has apparently completely forgotten how to function in a land of like pollen and pollinators. Every time I come to the South now, it's like the first three days I'm here.

00:30.46
Andrew
My nose is just like, fuck you, dude. So... Yeah, that's where I am. Check out my sweete my suite setup. I've got my suitcase with a couple books on it.

00:38.71
Sean
Amazing.

00:41.60
Sean
Amazing.

00:42.00
Andrew
Little, know, who needs a fancy stand-up desk when you've got a suitcase desk?

00:46.59
Sean
Yeah.

00:49.28
Sean
You can just record things. You can just do things.

00:52.52
Andrew
You know, don't have the fancy mic today.

00:52.93
Sean
Yeah.

00:55.88
Sean
Yep.

00:56.28
Andrew
know, the MacBook microphone is shockingly decent, maybe.

01:03.40
Sean
Yeah, your audio level is better than mine for what it's worth. So maybe should just swap. Maybe I should just... Yeah.

01:11.52
Andrew
Dude, it's all it's all an illusion. It's all just, like, stuff.

01:14.81
Sean
Exactly.

01:16.03
Andrew
And you don't need the stuff to do things. You can just do the things without the stuff.

01:21.45
Sean
Yeah. You just record it your vlog and on an iPhone and put on the internet. You might get a couple hundred thousand subscribers that way.

01:28.08
Andrew
Was it...

01:29.24
Sean
That's the Jackie Cho way.

01:30.69
Andrew
I just heard the other day that... Dude, do you remember Call Me Maybe?

01:36.79
Andrew
The song? The, like, pop song?

01:38.33
Sean
Yeah. Yeah. Yeah. All right.

01:40.12
Andrew
That was recorded on a MacBook.

01:43.03
Sean
Whoa.

01:44.15
Andrew
But yeah, that was recorded on a MacBook, like on iMovie.

01:44.89
Sean
That's sick. That's I didn't know that. but That's crazy.

01:48.59
Andrew
It was just like them holding a MacBook.

01:51.86
Sean
I think have you heard Here With Me by David?

01:58.63
Andrew
No.

02:00.87
Andrew
Uh-huh.

02:02.46
Sean
I think that guy did it on his like iPhone. With like an iPhone like digital audio workstation and just cut it up.

02:10.33
Andrew
That's sick.

02:11.29
Sean
It's pretty crazy.

02:12.11
Andrew
That's wild.

02:12.22
Sean
I don't know. Kids these days, man.

02:14.27
Andrew
There really are no excuses. Yeah. yeah I think we might have a little bit of lag.

02:18.70
Sean
Yeah.

02:20.30
Andrew
It feels like we're lagging a little bit.

02:22.62
Sean
Oh, definitely. We definitely are lagging a little bit. Yeah.

02:25.65
Andrew
Okay.

02:26.21
Sean
Oh, well, all right, if you're listening, we're sorry.

02:26.53
Andrew
Well.

02:29.21
Sean
Well, we'll do better.

02:29.98
Andrew
Yeah.

02:30.48
Sean
not Not this week or next week. I'm traveling next week, so I will also have, I'll be on hotel Wi-Fi.

02:34.68
Andrew
Yeah. Also...

02:39.24
Andrew
Yeah. Zincaster's pretty good, though, at, like, cutting out pauses, so maybe there will be no lag to the user.

02:40.21
Sean
Yeah.

02:45.18
Andrew
It's just lag to us.

02:47.08
Sean
Maybe, yeah.

02:48.11
Andrew
User, listener, what call this? Yeah. Okay.

02:51.26
Sean
I don't know, people, friends.

02:51.45
Andrew
Alright. What's going on with you? what's What's going on in Miscreants land?

02:56.98
Sean
I don't know, man. Giant tax bill.

03:00.15
Andrew
Oh, yeah.

03:01.44
Sean
That's...

03:01.85
Andrew
Yeah. Sucks making money, huh?

03:03.93
Sean
I know. It's the worst. i

03:06.12
Andrew
the

03:06.64
Sean
Life is so much easier.

03:06.85
Andrew
It's so hard to be profitable.

03:08.97
Sean
Yeah. yeah Especially if all of our profits just went into last year's taxes. No, it's it's it's like...

03:14.28
Andrew
not all of your ta Not all of your profits went into last year's taxes.

03:17.34
Sean
All all of our profit this year so far are paying off the taxes from last year.

03:24.24
Andrew
Okay, that's very different. This year so far is a quarter of the year.

03:26.02
Sean
For sure.

03:28.38
Sean
For sure. For sure. For sure. For sure. Yeah. and it's it's It's good. It's fine. We will just have more taxes to pay next year. And it'll just be this.

03:36.53
Andrew
Yeah.

03:37.43
Sean
Yeah.

03:38.44
Andrew
Okay, wait.

03:39.24
Andrew
Controversial topic.

03:40.69
Sean
Yeah.

03:41.17
Andrew
Do you pay estimated taxes?

03:43.81
Sean
I would like to. hi

03:46.39
Andrew
So, no.

03:46.59
Sean
Yeah. I would like to. We don't.

03:49.77
Andrew
I don't pay estimated taxes, and I don't currently have plans to start paying estimated taxes.

03:55.40
Sean
why don't you have like a don't you get fined i'm not a cpa i don't okay yeah yeah

03:56.97
Andrew
Because... Yeah, but... Yeah, yeah, yeah. So, you pay a penalty if you don't pay estimated taxes, which kind of bullshit.

04:08.09
Andrew
Like... Okay, government, I'm not gonna go there. I don't wanna be the anti-tax guy. I actually think it's good to pay your taxes. I want the government to have money and exist, and I want social services to exist and national parks to exist.

04:21.30
Andrew
So I don't wanna be the anti-tax guy.

04:22.12
Sean
agreed

04:24.51
Andrew
But, you know, at the same time, as much as I want all of the national, you know, public services to exist, I also don't, like, love, like, giving the government an interest-free loan if I don't have to. Yeah.

04:39.67
Sean
Right, or getting fined for it if you don't. For sure.

04:42.19
Andrew
Yeah, well, I'm fine with getting fined for it if I don't. that's i'm I mean, yeah I don't know. but But yeah, so basically my thought is the penalty that I pay is relatively small.

04:54.75
Sean
Yeah.

04:55.56
Andrew
And I don't know what my income is going to be quarter to quarter right now.

05:01.26
Sean
Yeah.

05:01.52
Andrew
And so I would rather not tie up money in an estimated tax payment. And then like, cause if I make, have a good quarter and then the next quarter i make nothing that those two things will like kind of balance each other out and I will end up overpaying on my estimated taxes.

05:12.82
Sean
Fair.

05:15.40
Sean
yeah

05:18.22
Andrew
And I'd rather just pay the penalty and have access to the cash when I need it.

05:24.31
Sean
fair

05:24.42
Andrew
So I'm kind of just like willing to eat that penalty right now until my income is more predictable.

05:31.93
Sean
That's fair. I mean, that's also been our case through the through the years where like our our cash flow is a lot less lumpy than it used to be when i sent invoices late.

05:43.08
Sean
And you know we also didn't have a steady stream of clients and all that stuff. So I'm with you. I also just found out that like because we did well last year, I now owe New york City like metropolitan business tax as well.

05:56.12
Sean
Because we are now like a New York City nexus because we operate out of here.

05:56.15
Andrew
oh Oh,

06:00.45
Sean
Yeah.

06:00.49
Andrew
yeah, yeah.

06:02.38
Sean
I don't know. I'm like one more frustration away from just going to take and learn a and be a CPA at this point. Or at least or like take like a giant tax course.

06:12.98
Andrew
God, no, you could never, you couldn't pay me enough money to do that. Absolutely not.

06:17.39
Sean
I and don't know if you know how spite-driven I am right now.

06:18.42
Andrew
Sounds horrible.

06:21.63
Sean
i There's not a literal target, but there's it's written on my whiteboard in front of me of like things I hate and taxes slash the internal revenue services currently on that list.

06:33.41
Sean
So,

06:34.75
Andrew
Yeah.

06:35.51
Andrew
It doesn't have to be this complicated, but it is.

06:37.30
Sean
Yeah, yeah, speaking of Speaking of, like, governments and money and all that stuff, like, I'm a little bit worried about, it it didn't occur to me until, i don't know, was driving, and then was like, oh, shit.

06:50.79
Sean
Tariffs are might fuck the e-com start side of miscreants, and I don't know.

06:56.83
Andrew
him

06:57.95
Sean
We are, we're, like, trying to figure out, like, should we just buy a bunch of blanks? We're talking to our printer to see we can just buy a bunch of blanks right now,

07:04.60
Andrew
Yeah.

07:05.05
Sean
Well, they're not, you know, I mean, that you know, they're not, they're not like, they're not made in China. So it's not a crazy tariff, but I also would not like to pay anywhere between 15 to 30 or whoever's whim.

07:18.68
Andrew
Where are they made?

07:20.87
Sean
I don't know. It's kind of different. Like I think Nicaragua for one of them, maybe Portugal for another, they're all, they're all different.

07:28.77
Andrew
Hmm.

07:29.70
Sean
everything. Yeah. Everything kind of comes from everywhere. I think maybe the hats are technically you pong. So they're, they are a Chinese brand. So God, if I have to double the price of a cap, I'm going to lose it.

07:44.24
Sean
yeah.

07:44.35
Andrew
Yeah, that's going to be tough. I don't know

07:46.11
Sean
Yeah.

07:46.77
Andrew
Yeah, I don't. Do you think people would absorb that? Like people would pay double?

07:51.31
Sean
I don't know. Like, maybe because we are, like, Shreeware is all about brand equity, right? And, and like, as you build that, you get to charge more, right? Like, Rue charges, like, Braindead charges 80 bucks for a t-shirt, for example. We charge 28.

08:05.76
Sean
twenty eight So kind of maybe, but i also don't want, like, I just think it's wrong. I just think it's wrong to charge, like, 60 bucks for, like, a shitty little, not shitty, but, like, a dad cap.

08:17.19
Sean
Like it just is against my principles as a person.

08:17.39
Andrew
Yeah.

08:20.90
Andrew
Yeah. I don't want to pay $60 for a dad cap.

08:23.75
Sean
Yeah. Yeah. yeah that's another That's another thing that I'm anxiety slowly, you know, just building internally, bubbling. But yeah.

08:34.19
Andrew
Yeah. That sucks. I hope. I mean, I guess it is. right Prices are going to go up, right? Like there's.

08:41.83
Sean
Yeah.

08:42.68
Andrew
your costs are going to up for sure.

08:44.90
Sean
Yeah.

08:45.03
Andrew
That's just going to happen at this point unless, I mean, who knows what the orange man will decide tomorrow, but, yeah.

08:50.78
Sean
Yeah. Yeah. Yeah. I mean, there's definitely going to be like the the company that we use for the majority of our blanks are like a lot of them are Canadian companies. So there's at least that amount added and that's going to be whatever is in that supply chain. So we'll see. We'll see. Yeah.

09:13.06
Sean
yeah What's going on with you? How's MetaMonster? You got a demo after this?

09:16.80
Andrew
I do have a demo after this. Yeah, so like the super positive thing is like I've been able to get in front of people and talk to people. I've had some really good conversations this week. Talked to a couple of old friends, which has been cool. Folks who run web design agencies.

09:34.29
Andrew
Talking to them.

09:34.57
Sean
Cool.

09:36.20
Andrew
Consensus from both of them was like, hey, I think there's something interesting here. i wouldn't pay a monthly subscription for it, but I would maybe pay as you go. Which... is interesting.

09:47.14
Andrew
i take that with a grain of salt because anytime someone says, I would pay if, you kind of have to go, eh, I don't, eh, would you?

09:57.22
Sean
that's like That's like someone saying, I would give you $20 because you're my friend, and I would buy you dinner.

10:01.37
Andrew
Yeah.

10:02.18
Sean
you know That's like the cost of once a one-time run the product. you know It changes if you're like, what if it costs $1,000 to run or to run once?

10:10.21
Andrew
Well,

10:15.45
Sean
it's

10:15.75
Andrew
Yeah, but it's also just like all the product advice I've ever read is like, don't believe what people say they would do.

10:22.75
Sean
yeah

10:23.22
Andrew
Focus on what people do and have done in the past.

10:26.76
Sean
Mm-hmm.

10:28.06
Andrew
So, but it's interesting, right? And we're we're seeing this not just from them, but we're also seeing, you know, from in, like, user behavior, people will sign up, they'll kind of use the tool, and then they'll cancel.

10:43.45
Andrew
And, like, you know, we probably our best, our best lead right now, we have a user who's been on a trial.

10:52.89
Andrew
They actually just converted their trial.

10:54.99
Sean
Nice.

10:55.07
Andrew
I'm checking with them to make sure they're cool with that and that it wasn't just like, know, they didn't just lose track of time. But they they said yeah, i see myself primarily using this for audits right now and I won't have another audit for like a month or two.

11:12.85
Andrew
And so, you know, that's someone who has actually put a credit card down and is using the tool and maybe even willing to pay for the tool, but they're still saying, and this isn't like shocking, right? Like if you just use logic right now, the tool is for,

11:26.90
Andrew
page titles and meta descriptions. And like, depending on the volume that your agency is doing, the volume of of new clients that your agency is working with, you just may not have a need for that one feature.

11:39.15
Andrew
I was just catching up with my friend Asia, who runs Demand Maven here in... Atlanta. And, you know, she was, she put it really succinctly. She's a big fan of the jobs to be done framework. And she's like, Oh yeah, you do one job right now.

11:53.42
Andrew
And in order to justify like a subscription cost, you probably need to do more jobs for people you need.

11:59.08
Sean
Yep.

12:00.69
Andrew
Which, yeah, I think that's kind of the boat that we're in. And so we're just having some internal strategy conversations right now to figure out. yeah We're still trying to figure out who our ideal customer is.

12:14.07
Andrew
Still think it's probably small SEO agencies or folks with like an SEO focus. But still trying to figure that out. yo i think for right now, we don't want to explore the pay-as-you-go option just because like I want to build a SaaS product more than I want to build a pay-as-you-go usage-based product.

12:36.58
Andrew
I'm not opposed to it if we need to go that route. like if If we tried some more things and can't add enough value to make the subscription feel worth it, then I'm i'm willing to to try that.

12:50.49
Andrew
But yeah, I think right now we're having like strategic discussions of like, okay, what do we do to like, how do we want to tackle building more jobs into the product, being able to build the product in in a way that it can handle more jobs and deliver more value.

13:09.74
Andrew
And yeah. Yeah, we're also talking about pricing and like playing around with that a little bit. but But I think at the core, we need to figure this out. And we've always wanted the product to evolve beyond page titles and meta descriptions. So that's not really a concern.

13:23.65
Andrew
The question is just how do we want to do that?

13:26.51
Sean
Mm-hmm.

13:27.81
Andrew
And so the two paths that we're thinking about are one, just staying the path that we were on, which is just like continue to make incremental improvements, layer in new features. Like we were working on planning to work on, you know, a couple of nice to haves, know, like ignoring pages and ignoring pages and, and then like,

13:52.62
Andrew
looking for missing keywords in page titles and meta descriptions. And then also going from there into, all right, now let's start looking at H1s, H2s, image alt text, internal links, things like this.

14:07.47
Andrew
And so we can just build those in kind of one at a time, keeping the product structure much the same as it is now where, you know, okay, cool, we roll out headers as a feature.

14:17.86
Andrew
Headers are now in... the issues table and they're now an option to to generate on each page and an option to to publish to your CMS.

14:33.27
Andrew
Oh, by the way, we did get our WordPress plugin approved.

14:36.28
Sean
I saw that.

14:36.33
Andrew
So our WordPress plugin is live, which is pretty sick.

14:41.33
Andrew
So yeah, that's that's like the the sort of, one path the question we're asking ourselves is like is that going to move the needle enough fast enough to convince people that like this is worth paying a subscription for and so then the other path that we're exploring is taking a bit of a bigger swing and this goes back to some of the conversations we've had on the podcast in the past about like leveraging the flexibility of ai

15:15.59
Andrew
And so it's like, do we just add a prompt, an open-ended prompt interface, chat interface into the product with like preset prompts or something like that so that you can suddenly go from, you can suddenly go from, right, the tool does two things, page titles and meta descriptions, to now the tool does, you know, kind of an infinite number of things. Like, whatever you can dream up with, you know, the context of a website and the power of an LLM.

15:55.87
Andrew
And, know, I've been testing this messaging with a couple people and, like, You know, one of my web design friends was immediately, like, he uses another tool that has that kind of functionality, and he was like, yeah, that would make it really interesting to me.

16:11.47
Andrew
But a lot of the people I run this by are kind of like, eh, what would I do with that? Like, what what would I, like, what's what's the use case you have in mind?

16:24.15
Andrew
And so I think that, like, That's not necessarily telling me we shouldn't do this, but I don't think we should talk about it like I'm talking about it right now.

16:35.20
Andrew
So I don't think we should present the feature as chat with your website if we go this route. I think it should be more like, like just this becomes how Metamonster works is like you youve run, and maybe we don't even call them prompts, like you run optimizations or something like that.

16:54.88
Andrew
And you can like,

16:54.98
Sean
Mm-hmm.

16:58.12
Andrew
you know, use natural language to create a new optimization or something. And we just make sure we have several built in by default.

17:10.42
Andrew
So yeah, this is... Yeah, i'm just I'm just thinking through all of this right now. And... and trying to figure out where where we go from here. But yeah, we're this is the thing we're debating. Like, do we just keep kind of layering things in one at a time?

17:28.68
Andrew
Or do we try to build a more flexible system that would allow us, you know, take us a little longer to build, but then allow us to drop, you know, eight new features at once where it's like, all right, cool.

17:42.20
Andrew
Now Metamonster will generate a voice and tone guide for you. Now Metamonster will, analyze content gaps and suggest, uh, new content for you to write. Now Metamonster will look for typos on your pages or broken links on your pages. Now Metamonster will, uh, suggest internal linking opportunities. Like, like the internal linking thing,

18:05.43
Andrew
think an open-ended prompt interface could do an okay job of that. If it has all the context because we've crawled it and it's all there and we can feed all of that to it, it could scan your entire website and look for internal link opportunities.

18:12.77
Sean
Yep. Yeah.

18:21.50
Andrew
I don't know how good it'll be, but I mean, that's kind of how we would build that feature under the hood anyway.

18:25.100
Sean
yeah

18:29.45
Sean
I mean, I think going back to the like lack of user creativity when given like a flexible platform, I think how everyone tends to how many folks tend to solve it is just through content, right?

18:41.26
Sean
and it And it means you get a lot of content. ideas off of it, right? How to generate internal links, or how to how to find internal link opportunities with Metamonster. How to do XYZ with Metamonster.

18:51.90
Sean
Like, Notion was like, here's a f flexible flexible workspace, but then the only way people like really wrap their heads around it is like, oh, here are templates. Here are, like, Notion for agencies, Notion for freelancers, Notion for software teams, etc., etc.

19:07.14
Sean
Yeah, I think it's one way to go about it. i what What is the delta between, like, Like, how long does it actually take to build that, do you think?

19:15.56
Andrew
Not sure yet. Austin's doing research this week, uh, to try to think through, we got to think through the user experience and like, I'm sure there's ways to stagger it so that we can like kind of, you know, there's an MVP version of this and then there's more advanced versions that we can layer in over time.

19:16.84
Sean
Okay.

19:33.77
Sean
Yeah.

19:37.39
Andrew
but, i Yeah, I'm not sure yet. i i would guess like even an MVP p version would be at least a few weeks of work, if not yeah a month.

19:47.70
Sean
Mm-hmm. Makes sense. Yeah. Okay.

19:53.45
Andrew
like an MVP that's actually usable and, and useful.

19:53.81
Sean
Yeah, I'm curious.

19:57.15
Sean
Yeah, not like a prototype.

19:57.88
Andrew
because I, yeah, cause I think like we could throw together an MVP that's just like literally just a chat window with all of the context of your website.

20:08.78
Andrew
Like we could probably throw that together in like a week or something, but that's not going to be useful.

20:12.13
Sean
Mm-hmm.

20:14.06
Andrew
Like people are going to be like, what the fuck do I do with this? yeah.

20:17.22
Sean
yeah yeah yeah and also it's not like more useful than it's not more useful than how we're currently doing it which is use metamonster pull down the markdown put in one markdown file drop into claude and then chat with claude and more so like with claude you know the way we're using it now is pull it down to a markdown but then put competitor mark like

20:19.76
Andrew
And then,

20:41.22
Sean
competitor scrapes all on the same project so that we can sort of look for like messaging gaps and things like that. Yeah. Yeah.

20:50.22
Andrew
Yeah. The other thing that this, that worries me about this is like, well, one, does this just open us up to, we have to build a bunch more shit that we don't necessarily want to build.

20:59.94
Andrew
Like, do we now have to build a word doc kind of thing in order to have like good, a good place to output the stuff from, from these prompts.

21:05.54
Sean
Right.

21:08.81
Sean
It's very hard, by the way.

21:12.11
Andrew
yeah, I don't, I don't want to do it if I can avoid it.

21:11.49
Sean
Pain the ass.

21:15.17
Sean
Yeah.

21:16.77
Andrew
And then it's like, know, I've talked before about like, you know, if we know what each prompt is and we've got it like all kind of structured, it's easier for us to sync it to your CMS.

21:29.75
Andrew
It's easier for us to put it in our interface and like, you know display it in a way that's really useful versus like the more flexibility, the more challenge there is there.

21:35.44
Sean
Yeah.

21:39.84
Andrew
And then, like, you know, you're talking about Notion and, like, you know educating people on how you could use this stuff. We're... tiny bootstrapped company.

21:50.14
Andrew
Like we don't have the time and the resources to crank out like, you know, notion level content on how to use our tool and like take time to convince people that it's valuable and that it, you know it fills a need for them.

21:59.68
Sean
Right.

22:04.99
Andrew
We need something that people will immediately crock and go, cool, this fills a need.

22:09.37
Sean
yeah

22:10.65
Andrew
I have, I would pay for this right now.

22:13.40
Sean
Yeah.

22:14.53
Sean
Yeah. Tough. Tough, tough, tough.

22:19.59
Sean
I guess that's the fun part about having constraints, right? ah fun like The The fun, miserable part of like you know doing it this way.

22:22.61
Andrew
Yeah.

22:28.65
Sean
yeah. it I mean, it also sounds like you just need to talk to more people at this point.

22:34.42
Andrew
Yeah.

22:34.91
Sean
like

22:35.43
Andrew
Yeah, I think so.

22:37.08
Sean
like you might You might land on a cold.

22:37.09
Andrew
well We're going to start. Yeah, we're going to start like doing some design work, like thinking through, doing the research into what it would take to build things like this.

22:46.90
Sean
Yeah.

22:46.89
Andrew
And I think like one of the things thinking about today after my most recent conversation is like how much of this, like building things to be more flexible would help us roll out more features faster.

22:47.00
Sean
Hmm.

23:02.17
Andrew
And like, we don't, we can choose how much of that flexibility to expose to users. And so like, maybe it's as much about building like internal tooling for us to, you know, test new prompts and then roll out new prompts.

23:15.50
Sean
no

23:17.17
Andrew
Like maybe we can, there's a way to like sort of leverage the flexibility of ai without like overwhelming users with that flexibility and like,

23:30.63
Andrew
Maybe not. Maybe that's the wrong approach. But...

23:35.24
Sean
I think it's a good middle ground.

23:35.33
Andrew
Like, I'm...

23:36.10
Sean
I think it's like...

23:36.88
Andrew
There's something... Yeah, rolling around in my head that I... I haven't fully formed yet. That's like, what if... We just look at MetaMonster as how do we build like a more flexible like rules engine grid system with like flexible fields and things.

23:55.19
Andrew
And then we can choose whether or not to expose that to users, but we can start using that to like quickly iterate and test and roll stuff out.

24:03.40
Sean
Yeah. I mean, I think it, i like I feel like the missing piece now is just literally how long does it take and how feasible is it?

24:13.48
Andrew
Yeah.

24:14.47
Sean
i what i What I like about that though is if you are dogfooding your own flexible building system or architecture or whatever, you don't have to expose it to users until you fine tune that by building more and more features on using that.

24:23.55
Andrew
Yeah.

24:29.77
Sean
And that way when you actually launch it, it's not like, you know, it's not like crappy, but there's a bunch of use cases already prebuilt. And every time you launch a, you know, a feature through that, that in itself is a, you know, yes, you don't have time like notion to go and write a bunch of these pieces, but you're going to promote each feature anyway. So you have some level of product marketing kind of built into that until, you know, until you, and then, you know, I think that way you can also, you you iterate faster it makes clients happier you know that uh not that we're unhappy but you know i i think one of the reasons actually like as users of metamonster one of the things that we enjoy is the fact that as we make feature requests we get them responded to pretty quickly from you and austin and this would only let that be faster right like

25:14.90
Andrew
Sure.

25:17.85
Sean
especially like it's not like we ask for it like give us internal linking or alt text or anything like that but as you get more of those the requests like product velocity is is like a nice thing for i'm sure is really nice for client churn yeah yeah just depends on how long it's gonna take it could be yeah

25:37.79
Andrew
Yeah. And maybe the thing to do is to like switch to pay-as-you-go pricing for now and like make metamonster relatively cheap and focused be like all right metamonster for right now is like a and metadata tool and like it's going to be relatively cheap and just here it is get in and and use it and then we start you working towards this bigger vision of like building a more flexible, like kind of system that we can, we can roll out features in.

26:09.17
Sean
Mm-hmm.

26:10.23
Andrew
And then, you know we can update our pricing as we start to roll some of these things out. and don't but I but I'm worried that that would get people kind of, it would either attract the wrong users or get people kind of frame MetaMonster in the wrong way to people.

26:27.93
Sean
Yeah, i think, like, at that point, yeah, I know you're going to hate this idea, but, like, at that point, I think you're better off just selling a lifetime subscription with capped features, well like a capped feature set.

26:39.57
Sean
like if Like, if someone's going to do pay-as-you-go, yeah, i I feel like I would rather just say give me this forever for a lump sum of money.

26:51.16
Sean
I knew you were going hate it.

26:51.35
Andrew
yeah Yeah, maybe.

26:54.42
Sean
knew you were going hate the idea, but

26:55.98
Andrew
I mean, plenty of AI tools right now where you buy a credit pack and then you just like you use those credits and then when you need more, you top it off.

27:04.94
Sean
Yeah, yeah, yeah. I mean, I guess my point is that, like, you can do that with a lifetime subscription as well. Like, JawPortly does this. You get a lifetime subscription.

27:15.00
Andrew
Yeah, but with a lifetime subscription, you don't ever get someone to pay you money again versus with credit tax, you do.

27:15.55
Sean
It's like, oh.

27:20.36
Sean
Yeah, you do. no you're you're yeah So that's my point. With Job Boardly, like you you pay the lifetime subscription so you don't pay for the whole SaaS monthly billing thing. So you have it.

27:29.24
Andrew
Uh-huh.

27:29.72
Sean
But then using Job Boardly, like there is a usage-based pricing locked into it like you get With the lifetime subscription, you get 750 credit back. Or this is what they used to do. They get 750 credits.

27:41.27
Sean
That's 750 jobs that it backfills. But if you want other things or more features or just more generation, like backfills for them, for jobs then uh you buy other credit packs to kind of keep it going but like i guess my point is my the nuance in my head is that lifetime subscription still frames it as like this is a sas product you just are getting it on the ground floor early rather than like your use case for your use case for meta monster starts as just the pay as you go sort of thing don't know i feel like you guys are not the same like

28:10.87
Andrew
yeah

28:18.94
Sean
like just the utility API type of tool, like webhook as a service or a stripe even or anything like that.

28:23.66
Andrew
yeah

28:27.16
Sean
So it doesn't make sense to me. Also I don't wanna pay a credits per generation as I go, right?

28:33.02
Andrew
yeah

28:35.72
Sean
But, hmm.

28:37.78
Sean
Yeah, lots to think about. but's

28:40.13
Andrew
That's the thing about... The other thing is, like, if we go this route, do we need to... switch to optimization credits or like AI credits of some sort anyway in the future.

28:51.67
Sean
Yeah,

28:53.64
Andrew
if it's a subscription that comes with a certain number of credits or whatever, cause like if we start doing a lot more AI shit, are we going to need to do some sort of credit back credit system?

29:00.92
Sean
yeah that's true.

29:03.90
Andrew
but at the same, I really want to talk to, i want to see if I can talk to the, Nathan from Lex, I know they're a little different. Like, they've raised some money, I think, and then they have, like, the every subscription model, like, to fund them.

29:19.04
Sean
Right, right.

29:19.71
Andrew
but But I'm curious, like, how they're making their pro models, like, how they're thinking about their pro models. or Because, like, right now, they're, like, Cursor, where for $20 a month, you can run an infinite number of...

29:35.14
Sean
Right.

29:35.13
Andrew
of like queries against like you all the frontier models. And so it's like like, how are they making the economics, unit economics work on that? Or do they just are they just choosing not to care right now?

29:48.51
Sean
Maybe they're self-hosting. Because you can you can self-host like like an open AI model and technically it's cheaper at scale, right?

29:57.09
Andrew
Can you? I don't think, don't think, you're not, OpenAI models aren't open source, aren't fully open.

30:00.09
Sean
I think so. I think with like Azure you can

30:04.28
Sean
Okay, I could be wrong.

30:06.71
Andrew
Maybe.

30:06.90
Sean
Couldn't tell you.

30:07.45
Andrew
Maybe you're right.

30:07.87
Sean
i havent i thought I thought there was like an Azure instance. that let you like you It's not open source. You you still you know pay for the usage of that instance, and it's like closed that way. But it might be cheaper than it having to go through like open yeah like a chat gpt api key i don't know i i've also just never looked into this past the two conversations people have told me so

30:29.97
Andrew
Maybe.

30:33.04
Andrew
yeah I know they have as they have a separation between like free models, which are largely the cheaper ones, and then or and then the the paid models. But still, $20 a month for those paid models is like, how are you doing that?

30:43.57
Sean
yeah nothing

30:49.91
Andrew
Yeah.

30:49.95
Sean
Yeah, I mean, also just VC welfare, you know, like, yeah.

30:53.83
Andrew
Maybe. I don't feel like they raised a ton of money, but they might have raised some.

30:59.43
Sean
Oh, I meant i'm mean cursor, but yeah, I don't know about Lex at all.

30:59.88
Andrew
So... Cursor for sure. Cursor's raised a fuck ton of money. They don't care. They're scorched earth.

31:05.35
Sean
Yeah.

31:05.90
Andrew
Like, yeah.

31:07.49
Sean
Yeah.

31:07.86
Andrew
Yeah. So that's what we're trying to figure out right now as well as just continuing to think about pricing. Anyway, what else is is happening in Miscreants World?

31:15.28
Sean
Cool.

31:19.33
Andrew
Or in Margins?

31:19.63
Sean
Well, you told me to talk to people and show people margins. So I did.

31:22.98
Andrew
Yeah, how'd go? Classic.

31:24.75
Sean
oh awful. if I fucking broke. yeah i'm i'm miserable at the moment i'm miserable i'm glad i'm glad you know i'm happy my dev is going it it's taking vacation and enjoying time with his family but like i'm sitting here for uh so okay update from last time one we couldn't submit to the aflo at marketplace because we need to have a landing page and all this sort of stuff set up so i gotta do that which is

31:28.30
Andrew
Classic.

31:47.76
Andrew
Sure. Are you all even going to be an app, really? Or are you just going to be an integration?

31:52.81
Sean
it it It's not a Webflow app, but we have to submit to the internal app marketplace so that we can auth through Webflow. Right now it's just like a developer account that we auth and we can only get access to like miscreants at the moment.

32:03.21
Andrew
Okay.

32:06.47
Sean
But you know, that doesn't fly for our clients.

32:06.27
Andrew
Okay. Okay.

32:09.57
Sean
i played it around with it i mean you know we paid for a prototype we got a prototype to make sure that like certain things work and everything tables do not work the way i want them to yet slash don't really work and that's you know that's a non-starter at the moment for us to so we have to figure that out and then

32:20.71
Andrew
Yeah.

32:25.71
Andrew
Yeah.

32:32.08
Sean
There's some like little tiny bugs that i've that I've caught as I've been working on it. like It kind of kind of dies once we want to load in like like an entire workspace of sites in at a time. That means that the miscreants workspace is like 400 sites.

32:46.20
Sean
so

32:46.85
Andrew
Oof.

32:47.88
Andrew
wow

32:48.48
Sean
I've made a lot of tests and cloned a lot of things over the years. I've had that Webflow account since 2018 something like that, or 2019.

32:52.38
Andrew
Yeah. Wild.

32:56.55
Sean
So, yeah.

32:57.51
Andrew
wild

32:58.76
Sean
like so many So many clones of things. And so many backups of like, like you know, in case like a client's thing like gets messed up or something. But yeah, still a lot to do feeling a little Feeling a little under the gun at the moment.

33:15.89
Sean
You know, I think like there's a ticking time. there's There's, you know, when Webflow when web flow kills the editor like uh i would really like to be at least in some level of like alpha or beta i don't know feasibly if that'll happen at this point so i'm so coming to terms with that i think it's okay it's not like anyone else is gonna have like i i just don't see anyone else building this thing at the moment

33:35.78
Andrew
Yeah. Yeah. Yeah.

33:50.90
Sean
i think we constantly still get like a lot of client validation that they really want this tool mainly i just got off a call with the client they're like yeah i would love to be able to give multiple people on my team like a place to write and then i can publish their stuff so that's good what else don't know i i feel like

33:50.07
Andrew
yeah

34:14.86
Sean
I mean, this is, I feel like I'm in I feel like you, maybe three, four episodes ago where you were doing, like you were antsy about, i mean, you were, you were doing robotics. I think Austin was, I think doing some life things as well. I forget, but.

34:31.86
Sean
Anyway, like, you're both busy, so, like, you had a week of a lull. So I'm, like, sitting here antsy about it.

34:35.77
Andrew
Yeah. Yeah.

34:38.49
Sean
That being said, there's a lot of things i should be doing at the moment. Like, there's I'm not blocked on sending a landing page or anything like that. Yeah. I don't know.

34:49.27
Sean
I just... Yeah. I feel like there's just so much...

34:51.60
Andrew
yeah

34:54.15
Sean
i feel like I've signed myself up for so much fucking shit to do at this point.

34:59.14
Andrew
Yeah, turns out getting a product off the ground that's like actually marketable is a lot of fucking work and way different than just a prototype.

34:59.58
Sean
Yeah.

35:04.03
Sean
Yeah.

35:06.27
Sean
Yeah. yeah Yeah.

35:09.38
Andrew
Yeah.

35:09.56
Sean
We're still definitely a ways away from like me being willing to let a client use it just because of like it's just missing.

35:14.38
Andrew
Sure.

35:17.11
Sean
We don't have to forget password flow, for example, at the moment. So.

35:21.65
Andrew
To be fair, we didn't have a forgot password flow for our first five users.

35:26.90
Sean
That's true. That's fair. i think that i think I think some of the other like game-breaking things is bothers me more than the forgot password flow.

35:29.47
Andrew
Yeah.

35:34.13
Andrew
Sure.

35:35.41
Sean
but Yeah, that's it.

35:37.67
Andrew
You just got to get early customers with good memories who won't forget their password and they won't know there's no forgot password flow.

35:43.88
Sean
That's true. um's true I think our ICP is often folks who will forget their password. Yeah.

35:50.85
Andrew
Yeah.

35:52.92
Sean
I mean, it's supposed to be, you know, it's supposed to be like you can type in a Google Docs basically, right? So yeah, lots to do, lots of new learnings. The demo was sad, but I think it was, don't know, it was good to do it.

36:07.12
Sean
You know, it was good to finally go and show someone and then break it on a call, I guess.

36:07.55
Andrew
Yeah.

36:12.27
Andrew
Yeah.

36:13.68
Sean
But yeah.

36:13.81
Andrew
good It's a good experience.

36:16.31
Sean
Yeah.

36:16.51
Andrew
You now know what breaks it and you've got ideas, I'm sure, for like what you want to be done differently.

36:19.98
Sean
Yeah. Yeah, definitely, definitely. But it turns out building a editor is fucking hard.

36:24.93
Andrew
Cool.

36:27.51
Sean
So yeah. I think we got to get out of here.

36:29.84
Andrew
Yeah.

36:31.29
Sean
You got to do the demo.

36:31.59
Andrew
Yeah, man. this This was a short one, but good to talk to you.

36:34.19
Sean
Cool. I'll see you next week from RSA. Peace.

36:37.83
Andrew
All right.

36:38.75
Sean
Later.

36:39.15
Andrew
See ya.