Andrew and Sean talk about the high robotics team Andrew mentors, choosing to delay the MetaMonster launch, hiring updates from Miscreants, and muse on the pros and cons of adding a custom prompts feature to MetaMonster.

Links:

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

Transcript:
00:00:00.85

Sean

Yo, yo.

00:00:02.06

Andrew

Yo!

00:00:03.42

Sean

How was...

00:00:03.58

Andrew

How are you?

00:00:04.54

Sean

I'm good. I'm tired. I'm good. I'm gonna hop in on a flight to New Orleans. New Orleans.

00:00:13.79

Andrew

Nolens?

00:00:13.76

Sean

Next. New Orleans in 48 hours.

00:00:17.25

Andrew

Sick. I am so fucking jealous you're going to microconf before me, but I'll make it there one of these days. It'll happen.

00:00:24.46

Sean

Next year.

00:00:24.73

Andrew

Although, I guess if I keep doing robotics and it's always during robotics season, then maybe I won't ever make it.

00:00:29.86

Sean

Yeah, that's...

00:00:30.05

Andrew

I could go to the Europe one.

00:00:31.77

Sean

Yeah. yeah

00:00:32.79

Andrew

That would make no sense. That would be so much more expensive.

00:00:35.73

Sean

and But don't they have like... Oh, so Rob Walling. Rob Walling. Rob Walling. But...

00:00:41.58

Andrew

Hey, Rob.

00:00:42.78

Sean

Hey, Rob. Don't they have a... Do they only have a Europe and Americas one? For some reason I felt like they

00:00:50.68

Andrew

Yeah, I think they just have two.

00:00:51.62

Sean

Okay. Okay. Cool. Cool. Cool.

00:00:54.88

Andrew

Two per year.

00:00:56.38

Sean

I'm excited. and

00:01:00.18

Sean

I don't know what to expect. I feel like this is the only time I've been to a conference that wasn't a security conference.

00:01:01.88

Andrew

Mm-hmm.

00:01:05.52

Sean

So I'm like a little bit nervous about it.

00:01:06.35

Andrew

Whoa, yeah.

00:01:08.86

Sean

Yeah.

00:01:09.38

Andrew

You're not going trying to get work. You're just trying to meet people and learn and have a good time.

00:01:12.04

Sean

Yeah.

00:01:16.25

Sean

I know. That feels so wrong. Yeah,

00:01:18.70

Andrew

Yeah, they've that's like not Sean at all. how Sean going somewhere and not trying to make money?

00:01:22.02

Sean

hu

00:01:24.87

Andrew

What? What?

00:01:27.47

Sean

yeah I don't know what I'm supposed to do. Make friends is crazy.

00:01:35.75

Sean

yeah and Yeah, no, I, and then, and then, and then I'm going to RSA.

00:01:41.10

Sean

in well one month right after that so i'll be back to my normal self afterwards

00:01:47.39

Andrew

Gotcha. How many conferences are you going to be at this year?

00:01:54.19

Sean

like three just rsa black hat i mean you have to if if like i count rsa and b-sides as one conference and oh shit yeah you're right oh no what going i think i think i might

00:01:51.71

Andrew

Total. By the end of the year, you think?

00:01:56.20

Andrew

No.

00:02:00.94

Andrew

Weren't you just at ShmooCon and like... Did you go DistrictCon?

00:02:10.09

Sean

try to sneak into ChinaCon.

00:02:12.74

Andrew

Oh.

00:02:13.42

Sean

Small conference for like to talk about like like Chinese APTs and all that stuff.

00:02:13.82

Andrew

What's ChinaCon?

00:02:20.27

Andrew

Hmm, neat.

00:02:20.38

Sean

Anyway, yeah.

00:02:23.04

Andrew

Are you gonna go to any of the, we have like multiple clients who are throwing their own conferences this year. Are you gonna go to any of those?

00:02:28.75

Sean

Oh, yeah, yeah, yeah, yeah. You are reminding me of all the travel that is... You know, I don't think about these things until, like... I didn't book my travel... I didn't book my plane ticket for MicroConf until two nights ago.

00:02:41.70

Sean

Because...

00:02:43.10

Andrew

It sounds like someone who has disposable income.

00:02:43.84

Sean

Yeah,

00:02:47.87

Andrew

Someone who has has a business card they can put things on.

00:02:48.36

Sean

that's weird.

00:02:51.51

Andrew

and

00:02:51.75

Sean

That's true. That's true. The $100 difference doesn't make a that make good significant difference, I guess. Well, after taxes are involved.

00:02:57.64

Andrew

Yeah.

00:02:59.44

Sean

Speaking of which, my taxes are horrendous this year. Oh, yeah.

00:03:02.01

Andrew

Oh, fuck. Mine are going to be terrible, too, because I didn't pay any like estimated taxes last year.

00:03:06.33

Sean

Yeah. yeah

00:03:09.78

Andrew

Do you have a good accountant? and

00:03:12.42

Sean

we just got a new accountant or we stopped working with our previous one because like well he stopped working with us i think we outgrew sort of all the services because you know they're very much like here to do it for like very small solopreneur slash mostly w2 employees slash like like very small local main street businesses and i think

00:03:34.57

Andrew

Would they work for me? a very small solopreneur, but have like weird...

00:03:37.56

Sean

there therere They don't know Michigan tax law there in New York.

00:03:43.11

Andrew

yeah I'm also like, my LLCs are all Delaware-based.

00:03:46.71

Sean

you know That's not something you know.

00:03:48.86

Andrew

I don't know if that...

00:03:50.15

Sean

yeah He's like an LLC S Corp type of guy.

00:03:50.98

Andrew

Damn. Okay. I got to find... it I want to try to find somebody because I really don't want to do it myself.

00:03:54.12

Sean

Yeah.

00:03:57.25

Andrew

i Well, there's a decent chance I don't find someone in time, and then I end up just doing it myself. But the person that we used for crit...

00:04:06.76

Sean

yeah

00:04:08.87

Andrew

They're a little expensive for personal and they're just like not really, i don't and don't feel like they're super well set up to do personal. Also at this point, it's probably too late for me to find someone new and expect them to do my personal taxes by April.

00:04:20.48

Andrew

But I could, so they'd probably have to file an extension, I'm realizing. But yeah. but

00:04:26.28

Sean

Dude, I just found on extensions, like, every day you don't pay your taxes or or finish it. Like, it just accrues interest.

00:04:33.18

Andrew

Yeah, that's why like you should even if you file an extension, you should still like submit an estimated payment or something, I think.

00:04:40.45

Sean

Yeah. Yeah.

00:04:41.68

Andrew

Because otherwise, you end up just racking up penalties, which sucks.

00:04:45.27

Sean

Yeah.

00:04:48.77

Sean

It's fucked up.

00:04:49.00

Andrew

I mean, I think I owe i will owe interest. I think you're technically supposed to, as a solopreneur, make quarterly estimated payments. And so I think I will owe interest from not having made those already.

00:05:00.50

Sean

Same. Me too.

00:05:02.32

Andrew

which will suck yeah we'll see i've i've been putting money away for taxes so i've got a good chunk saved that should cover it all i hope to god but uh yeah figuring it all out is going to be a little bit of a pain and then like yeah i i just i want to talk to someone because like

00:05:03.49

Sean

Me too. Yeah. Yeah. Right.

00:05:13.27

Sean

Hmm.

00:05:26.52

Andrew

We finally closed CRIT last year, like officially like wound down the company.

00:05:29.08

Sean

Hmm.

00:05:29.98

Andrew

And so like I got a disbursement check from that. And I'm like, what the fuck do I count that as on taxes? I have no idea.

00:05:36.67

Sean

Right.

00:05:39.04

Andrew

So yeah, I need to get on that. I'll probably do that next week, try to find someone. I tried to use Sam's List.

00:05:47.08

Sean

Me too. No one responded.

00:05:48.39

Andrew

and every company on there is like we work with you if once you have five million dollars in revenue or raise 20 million dollars in vc and i'm like cool this is not fucking helpful yeah

00:06:00.90

Sean

Yeah. You can make Andrew's list. Baby Sam's list. Sam's baby list. Yeah. Every time this comes around, always feel like this should not be like the information asymmetry.

00:06:14.76

Andrew

like it should be stupid

00:06:16.88

Sean

No, just the it's information asymmetry should not be like this significant. You know, like I feel like.

00:06:21.26

Andrew

Oh, yeah, like the government should just send you a bill and be like, here's exactly what you owe, just pay it.

00:06:26.60

Sean

Sure, not even not even.

00:06:27.88

Andrew

I mean, that's how it works in like most countries. Most countries, yeah, you don't have to like do this all whole like tax filing bullshit.

00:06:30.20

Sean

Oh, really?

00:06:34.83

Andrew

the The government knows how much money you made because like the companies various companies reported it.

00:06:35.20

Sean

and

00:06:40.22

Andrew

I guess it might be different for solopreneurs. so But in most countries, like the government knows how much money you made. They just send you a bill and you pay it.

00:06:49.12

Sean

yeah but you don't get any tax loopholes in america for the for the people who you know really need those tax cuts uh I just mean like someone make me like a comic book that explains this stuff or just like a very no bullshit guy.

00:07:08.77

Sean

Like, like I don't want like to watch multiple YouTube videos to figure all this stuff out. I just want very exacting.

00:07:16.48

Andrew

I mean, that's that's literally like what Tax Act and hr Block and you know TurboTax all are.

00:07:18.54

Sean

no,

00:07:23.20

Sean

no, no. See, those are.

00:07:24.18

Andrew

They're all just mad libs for your taxes.

00:07:27.39

Sean

Yeah, but that's. that's not what I'm saying. I don't want, okay, sorry, sorry. I don't want, it's not that I want to make, and don't need my tax filing to be easier. i want my ability to understand taxes and like the, like S corps and K ones and LLCs and solo printer stuff and paying estimated quarterly taxes and all these sorts of things like to just be like, I just want like a two page checklist.

00:07:52.10

Sean

You know, if you are

00:07:52.54

Andrew

But you literally can't have that because of all the fucking bullshit tax loopholes in this country.

00:07:57.75

Sean

That's a good point.

00:07:57.86

Andrew

And yeah, you can't have both.

00:07:58.24

Sean

That's a good point. That's a good point. You're right. You're right. You're right. You're right. You're right.

00:08:00.56

Andrew

You either get a simple tax code where the government can just send you a fucking bill or you get our bullshit where the tax...

00:08:00.62

Sean

You're right. You're right. right You're right.

00:08:07.75

Andrew

law is intentionally kept as complicated as possible to make it easier for wealthy people to avoid taxes and make it easier for companies to charge you because taxes are so stressful to do and if you fuck them up it's like feels life altering and so it's all like the system is designed this way on purpose it is not it is not an accident

00:08:07.100

Sean

Right. Right. Right.

00:08:20.86

Sean

right

00:08:26.71

Sean

yeah That's a good point. That's a good point. Yeah.

00:08:32.95

Sean

Okay. I'm like more exciting and not

00:08:38.61

Sean

sad about the world news. How's robotics going?

00:08:44.14

Andrew

Ooh, it's good. It's really good and wild and and stressful and I wish we had more time. But it is like, you know, i think one I think the best way to describe it is it just feels kind of all-consuming right now.

00:08:51.45

Sean

I see.

00:09:00.45

Sean

mean Okay.

00:09:01.72

Andrew

I spent...

00:09:04.32

Andrew

21 22 hours this weekend working on robots with students i was just from nine to like seven or eight saturday and sunday i was there working on robots

00:09:10.88

Sean

wow

00:09:21.01

Andrew

so we our first competition is this weekend if i haven't said it before or anyone's listening that's new i mentor a high school robotics team in southwest detroit our students are predominantly hispanic and uh we are a super tiny underfunded team going up against like some of the best teams in the country uh because michigan has some like really, really crazy good teams that have been doing this for years and years and years and have like 50 kids and

00:09:51.41

Andrew

yeah tens of thousands of dollars in funding, if not more. And a ton of parental support and like you know all their parents are engineers at various places. And you know we're we're a tiny team of like, and we actually grew a bunch this year. So this year we're like 15 kids and like yeah half a dozen mentors.

00:10:14.37

Andrew

And yeah, so we are very much... little scrappy trying to punch above our weight class uh team and so yeah we have our first competition this weekend the robot is built there are a million little things i would like to tweak and tune and test that we will probably not have time to tweak and tune and test

00:10:33.25

Sean

well

00:10:41.55

Andrew

The game works. There's like 15 seconds of autonomous period and then like two and a half, three minutes of teleop, so human driver controlled. And our robot is primarily going to take these little, these, you know, 12 inch long chunks of PVC pipe,

00:11:01.75

Andrew

from a ramp and then try to move and place them onto metal rungs on these posts in the center of the field and then hopefully end the match by latching onto a cage that's dangling from a chain in the center of the field and pulling a robot bot up off the ground and uh so far our elevator to move the pieces up and down is like functional, but still a little glitchy.

00:11:36.02

Andrew

Our vision auto alignment is functional, but still a little glitchy. Our autonomous path planning is functional, but still a little glitchy.

00:11:45.87

Sean

Thanks.

00:11:46.59

Andrew

and our climber is functional, but still a little glitchy. So everything is like, In a best case scenario, we can do everything we need to do We'll definitely be able to move pretty quickly and like score a bunch.

00:12:01.26

Andrew

But like you know we have had very small amount of time to really test the robot and like work out all the kinks.

00:12:11.02

Andrew

And so we'll be doing that live at the competition.

00:12:16.83

Sean

Nice.

00:12:17.97

Andrew

Yeah.

00:12:19.88

Sean

Aren't you so excited for like these kids that grow up and do like very cool things with their lives later on?

00:12:27.34

Andrew

I'm stoked. I am. We have two of our kids who have been doing this the longest now are seniors. And so they're both graduating and going to Wayne State, the local college.

00:12:38.33

Sean

Okay, cool.

00:12:39.51

Andrew

And one's going to go into engineering and one wants to go into physics.

00:12:42.54

Sean

Nice.

00:12:43.33

Andrew

And they're both super smart, super capable kids.

00:12:43.69

Sean

Nice.

00:12:47.41

Andrew

And so I'm really stoked to see like where they are in four years. i hope

00:12:51.35

Sean

yeah

00:12:52.11

Andrew

My hope is I can stay in touch with them and just like see see what they do.

00:12:54.74

Sean

for sure for sure nice yeah my one of my one of my good friends in new york has like been a part of like or got started like has been a part of like the big brother program and is like a big brother for this like 10 year old or something like that yeah anyway like this is this is this is what mentoring the youth is is like it's cool it makes me it makes me want to go do it it feels like a better

00:13:06.51

Andrew

Oh, cool.

00:13:15.69

Andrew

Yeah.

00:13:20.24

Sean

use of my time and a commitment to go do that, you know.

00:13:24.76

Andrew

It is the perfect antidote for all the bullshit right now because it's like it's this fun, very tangible thing that's very local that is making a small impact on a small number of people's lives.

00:13:27.50

Sean

Yeah.

00:13:36.22

Andrew

And it's like it's so so focused.

00:13:36.20

Sean

Yeah.

00:13:38.89

Andrew

And so like when everything feels hopeless, it's like this is my thing that I have chosen to do right now to like try to make. the world slightly better and happier.

00:13:45.48

Sean

Right. Right.

00:13:48.48

Andrew

Although I do need to remind myself that because sometimes i get caught so caught up in like the competition and getting shit done that I like it's easy to lose sight of that and be like, I want the robot to work. I want it to do this thing.

00:13:58.70

Andrew

And i do need to remind myself sometimes like you're here for the kids.

00:13:59.32

Sean

right yeah and

00:14:02.66

Andrew

You don't it's OK if we don't win the competition. We you're here for the kids. Just help the kids have fun and learn and like and everything.

00:14:11.35

Sean

yeah yeah that's cool that's sweet nice well good luck good luck to them how many you say list to the 15 the 15 that's 15 you

00:14:16.13

Andrew

Yeah. Yeah. Thanks.

00:14:22.79

Andrew

What's that?

00:14:23.81

Sean

it's fifteen kids he said

00:14:25.38

Andrew

Yeah, we've got like 15 or 16 kids. We've got, you know, probably a core group of six who are actually in high school who are like, you know, the most consistent and and the most involved.

00:14:31.06

Sean

Nice.

00:14:33.76

Sean

Got

00:14:35.78

Andrew

And then we've got a bunch of eighth graders who are like part of the team temporarily because they're they're like right on the cusp.

00:14:41.83

Sean

it, got it, got it, got it.

00:14:43.26

Andrew

And so they're allowed to like age up. And so, know, we've got, you know, seven or eight eighth graders who come when they can.

00:14:52.03

Sean

Nice.

00:14:52.89

Andrew

Yeah, yeah.

00:14:52.86

Sean

Cool, man. Yeah.

00:14:54.48

Andrew

So with all this going on we made the decision this morning to push back the Metamonster launch again, which I have mixed feelings about.

00:15:02.40

Sean

Mm-hmm.

00:15:04.58

Andrew

It was like, so we decided to basically just wait until competition season is over and launch early April. Not putting an exact date on it yet.

00:15:15.82

Andrew

We have an internal date, but we're not publicly saying that date. But, you know, we made this decision because it just kind of makes sense. Like two weeks here or there isn't really going to change anything drastically about like the fate of the company or like what we learn or any of that.

00:15:32.25

Andrew

And it'll just be better to have be able to have my like full attention when we're launching. And honestly, we still have like a few things we want to button up. Still haven't started the WordPress integration. so this will give us a couple extra weeks to to hopefully make some progress on that.

00:15:46.16

Andrew

yeah Get the last few like little SaaS details in there. like yeah Build a settings page so you can add team members and configure your billing and all that. So I think it's the right decision.

00:16:01.43

Andrew

There's part of me that is that feels very much like we're supposed to be moving fast. We're not moving fast enough. We can't delay another four weeks. like That's not moving fast. we've got to like We're lazy now. become lazy

00:16:12.87

Sean

You guys aren't being, you guys aren't founder voting. I'm sorry.

00:16:15.91

Andrew

yeah we're like i'm like that's the thing that's the narrative that that builds up in my head and then but i the flip side of that is like the whole reason we're bootstrapping the whole reason we want to build our own company is so that we have the optionality to say hey i want to help the robotics team right now more than i want to launch a product

00:16:17.96

Sean

Uh-huh.

00:16:22.31

Sean

Yeah. Right. Mm-hmm.

00:16:29.42

Sean

right

00:16:40.30

Andrew

And that's okay.

00:16:40.74

Sean

i

00:16:41.38

Andrew

And so we're just going to do that. And so I'm trying to just own that and lean into it. And so I actually like, posted on LinkedIn about like, Hey, we're delaying the launch again, but we're doing it for a reason that I think is really good.

00:16:54.38

Andrew

And I just sent out an email to everyone on the waiting list saying like, Hey, we're getting close to picking a date for the launch here you know truthfully i was hoping we would have launched by now but here's why we aren't it's because i do this other thing this robotics thing and i just want to be honest with you because i think that shows the kind of company we're building so trying to like own it and like turn it into a positive

00:17:12.55

Sean

Yeah.

00:17:19.12

Sean

Nice, nice, nice.

00:17:21.44

Sean

Yeah, I always... Because, like...

00:17:25.81

Sean

i think that's interesting to see this

00:17:32.58

Sean

like it's interesting to see like the cognitive dissonance happen in front of me i don't know if cognitive dissonance is necessarily the word but like i always feel that way about like like concepts like county reports like slow productivity or like calm businesses all and and all that sort of stuff but then like my brain i feel like my brain when it comes to these things is is like what we should be we should be launched like yesterday you know i feel like

00:18:00.71

Sean

Anyway, I don't know where I'm getting at with this. I think i think I'm just trying to... Like, I don't think I've fully unraveled that in my brain of, like, having...

00:18:16.40

Sean

of not just like burning the midnight candle and and then burning myself out by wanting to do more than I should be doing. But yeah, it's good. It's good. It's good. to It's good. You know, it's cool.

00:18:29.06

Sean

It's cool to see. it's It's cool to see you work through that. So, yeah.

00:18:34.100

Andrew

where for a second I thought you were going to say, i think you're being dumb.

00:18:37.70

Sean

Uh-huh.

00:18:38.71

Andrew

You should launch. Like, I thought that's what you were saying is like, oh, if I were in your shoes, I would just launch it now.

00:18:41.08

Sean

Oh.

00:18:44.45

Sean

Oh, no, no, no, no. I think you should delay it, I think.

00:18:51.79

Sean

Yeah, no, I think it is... it is

00:18:57.94

Sean

These are the moments where I am sad that we don't edit our podcast at all, by the way, because it's just me struggling for like multiple minutes.

00:19:11.55

Sean

I think to me, you've always been like the, like you've always upheld thoughts around like calm businesses the important yeah yeah like bootstrapped and all those things and like those are like to me like when you like i would have given you the same advice back of like like when you were saying that like you were freaking not freaking out when you were contemplating like why you weren't reaching out yeah like i think my reaction would have been like but you're always the you're always the like you're bootstrapping for a reason don't want vcs hounding you get more decision

00:19:20.98

Andrew

The importance of work and yeah.

00:19:36.68

Andrew

Yeah, I was freaking out a little.

00:19:50.29

Sean

decision making and and freedom but I guess really what I'm trying to say it's good to see that you have cracks too Andrew that it falls apart for for like five minutes no it's it's it's good yeah i think it's I think it's cool I think it's it's like

00:20:04.50

Andrew

Cool.

00:20:08.49

Sean

Yeah, I think I'm just internal. I think I'm vicariously processing your emotions at the moment and like internalizing what what it would mean for me to like send out an email like that or like how it would feel about those things versus.

00:20:24.74

Sean

here's here's what this is just therapy now. Okay, what I'm getting at what I'm getting at is I think I am am empathizing with what you are what you're feeling and then thinking about if like and and the fact that you sent out an email saying hey we're delaying the launch like because I'm doing this thing is not a route my brain i think would have come up with as an answer.

00:20:33.35

Andrew

yeah

00:20:49.35

Andrew

yeah

00:20:51.48

Sean

And I think it's, that's like an interesting reminder that we can't just do that. Yeah, sweet.

00:20:55.98

Andrew

cool

00:20:56.91

Sean

sweet

00:20:57.30

Andrew

Yeah, I am. I'm trying to both practice what I preach and be transparent about it being hard to practice what I preach, like be transparent about the fact that like I am freaking out a little bit about not launching and delaying the launch. Like I really wanted this thing to be launched in January. And so it's like it's agonizing to me that we're now talking about April.

00:21:19.17

Andrew

Yeah.

00:21:19.94

Sean

For sure.

00:21:20.86

Andrew

But I'm trying to also practice what I preach and go, no, we are building this in a way that we can make these choices. And so I'm going to make this choice.

00:21:30.22

Sean

for sure

00:21:30.37

Andrew

I did grapple also with like, should I send this email? Like, is this too much me focus? Because like the best marketing is focused on you, the reader, not me, the person building the business.

00:21:43.41

Andrew

But I think this is one of the examples where you can kind of break the rules a little bit and and be like, I'm going to This is me trying to show you, like, be transparent with you and, like, and, know, real... Yeah. really

00:21:58.72

Sean

yeah I think you're setting the tone of what of like the type of company Metamonster is going to be, you know, which I think is valuable and is important.

00:22:00.56

Andrew

Yeah.

00:22:05.39

Andrew

Yeah.

00:22:08.41

Sean

Yeah, I think. Okay, I think I'm just like having feelings around like the amount of disappointment I felt during the entirety of like stack wise taking centuries and then dying. yeah.

00:22:22.03

Andrew

Sure.

00:22:22.73

Sean

yeah

00:22:22.82

Andrew

Yeah. Yeah. And I mean, at some point we do need to hold ourselves accountable and like hit a deadline.

00:22:29.99

Sean

For sure.

00:22:30.28

Andrew

and And so we have a date now internally and I don't see any reason we would need to move that date.

00:22:30.25

Sean

Nice. I see.

00:22:38.63

Andrew

Like the product is ready to go. We could launch it. I could flip the switch today if I really wanted to. There's more things I'd like to add, but I would be fine flipping the switch today.

00:22:48.72

Sean

i see

00:22:49.44

Andrew

And so I think that's a good place too, is we can, there's no reason to delay at this point besides just like, we don't think it's the right timing for personal stuff.

00:23:01.33

Sean

I mean, i think you are I think if you are busy with robotics and you launch the thing, it's go like and like you're going to be in this stage of, like are you going to do customer support or are you going to do robotics?

00:23:11.69

Andrew

Exactly.

00:23:12.34

Sean

you know And I think that's a bad.

00:23:13.28

Andrew

And I want to be super, super present, especially that first week, if we get handful of new people signing up and trying it out so that you know I can provide the best possible customer support.

00:23:17.01

Sean

Right.

00:23:20.53

Sean

Right.

00:23:24.98

Sean

Yeah.

00:23:25.30

Andrew

And truthfully, like because I've been so busy with robotics the past month, I also have let the list go a little cold. And so this now gives me a chance to spend the next few weeks rewarming the list and reminding people that we're here and what we're doing and and all that.

00:23:35.93

Sean

Cool.

00:23:38.86

Andrew

so i think it's the right call, but it was not... It was not a super easy one to make. It was just we had to grapple with it a little bit. So what's going on in Miscreant's world?

00:23:48.59

Sean

if

00:23:51.97

Sean

uh what is going on in streams world we are entering that pre-rsa crunch which is super exciting not at all like we are busy like i hate i hate turning down uh new business it is like every time we turn down new business like it feels like it's just be like like because it's a capacity problem right but like as i feel like as an agency

00:23:58.88

Andrew

Okay.

00:24:08.45

Andrew

Mm-hmm.

00:24:16.83

Andrew

Mm-hmm. Like, it's really a good thing to turn down new business. It means you are doing something right, and you have, you have like, sold all of your time.

00:24:28.37

Sean

Yeah, yeah, yeah, yeah. but But like, it also means that we didn't plan capacity correctly in a way where, like, the okay, perfect world is that everyone is always at 80% capacity. Every time we flex up to 90, we hire, and we you know, we have we have an active recruiting pipeline, so we're able to kind of continue to Aaron Shawl- To take on an infinite amount of world work and take over like the cybersecurity world because everyone just comes to us in my brain that's like what I wanted to be right.

00:24:58.26

Sean

Aaron Shawl- That is unfortunately not like, and for like, you know, it only took us four years to build a really, really steady inbound pipeline and reputation.

00:24:58.51

Andrew

Yeah.

00:25:08.85

Sean

And ive now we are like finally really working on the recruiting side and like thinking through all those things.

00:25:16.47

Andrew

And by the way, don't be shocked if like at some point inbound slows down a little bit or you or you overscale and like, but this is just, I'm saying like, just don't like you guys have awesome inbound.

00:25:21.04

Sean

Don't say that. Don't don'ttomp put that on whack. Yeah. For sure.

00:25:30.41

Andrew

I'm not, I'm not trying to will this into the world.

00:25:30.90

Sean

For sure.

00:25:32.75

Andrew

I'm just saying like, don't,

00:25:33.10

Sean

sure.

00:25:35.83

Andrew

like Don't get to a point where like things shift a little bit and you're like, don't start telling yourself the world is falling out from under you if if that happens.

00:25:44.85

Sean

Totally. Totally.

00:25:45.14

Andrew

because Because that's just the reality of like agency life is these things kind of like ebb and flow on both sides, both the recruiting side and the inbound side.

00:25:49.91

Sean

totally

00:25:54.86

Andrew

And it's really hard to thread that needle and keep them in a good steady balance all the time. It's it's super hard.

00:26:02.33

Sean

For sure. we We are quite literally at the whims of VCs in the market. you know like like the wheat like My inbound is not entirely because I am... you know like like we have not We have not done attributable marketing work for ourselves to justify, to say like we we did X ad and therefore we got something. like are It is at the whims of you know VCs are investing and...

00:26:31.10

Sean

people are raising and have budget and are doing things so yeah on another note we are recruiting which is great last time i think we spoke i told you we hired a recruiter

00:26:44.59

Andrew

Mm-hmm.

00:26:44.85

Sean

that uh that's continuing to pay dividends we have them on helping us try to find a project manager which was interesting i think i think it sounds like jj found someone that she really likes think i interview them on thursday will be cool to

00:26:51.39

Andrew

Cool.

00:27:03.40

Andrew

Really good project manager is like, in my opinion, might be the most important position in an agency.

00:27:10.36

Sean

For sure. For sure. Don't tell him that. But for sure. hundred percent 100%. 100% it is. Yeah. yeah You know, you can't scale without form is what a previous boss always told me.

00:27:21.40

Sean

And I think this is sort of...

00:27:22.58

Andrew

You can't what and say that again? Hmm.

00:27:24.01

Sean

You can't scale without form. And I think that's always sort of resonated with me.

00:27:27.38

Andrew

OK.

00:27:31.94

Sean

Good. Yeah. Otherwise, I mean, good. I still, okay. So I hate turning down work. That sucks. And we keep getting so busy. Like we're, we're busy.

00:27:44.91

Sean

I'm not the most busy, which is good in a way.

00:27:47.30

Andrew

Mm-hmm.

00:27:48.05

Sean

Like I'm i'm still busy, but I still have nights where I don't have, so a lot of things to do i do i have i did like dump out all the things i have to do for clients which is like not insignificant but there was space on the page for me to think what are the things i should be doing if it's not any of these like these are all in the business things right that's like this client wants this thing i can't i can't the problem is i

00:28:07.69

Andrew

That's great. That's awesome. I would love to get to the point where those are like not at night that it's like you have no night. Are you doing any work? And during the day are you are going.

00:28:18.04

Sean

I can't promise you that. Anyway, the thing that I'm getting to, though, is that the box of like what I should be working on, i think I'm struggling a little bit to figure out what exactly I should be putting in there.

00:28:18.86

Andrew

Yeah.

00:28:32.29

Sean

Yeah.

00:28:32.80

Andrew

yeah It's tough. The agency leader, CEO is like kind of a weird, hard to define role. And so it can be tough to figure out like, what are the highest leverage things you can be doing?

00:28:48.06

Sean

Yeah. Like, nothing. and I just shouldn't even work here.

00:28:56.47

Sean

No, I think there's obviously some things, but... Like taxes, for example. And like deciding... Yeah. But it has made me... has It has filled up my schedule where i haven't had as much time to push on the SaaS stuff.

00:29:15.76

Sean

Which is you know slowly eating away at me. So going back to the whole... like like he's just fucking sending out an email it's like hey i'm doing robotics fuck you guys is a is like an interesting no

00:29:27.98

Andrew

yeah

00:29:31.79

Andrew

interesting strategy for making yourself feel better

00:29:35.13

Sean

no i just think it's i think it's just like a good foil for it you know uh okay i think i think the reason it it like came as a shock to me is and why it's a good foil is like

00:29:38.47

Andrew

yeah

00:29:48.52

Sean

Like, it hasn't occurred to me that, like, you can just choose to do one thing. Like, you can just choose to prioritize one thing over the other rather than, like, yeah.

00:29:59.70

Sean

I know that sounds really insane.

00:30:01.43

Andrew

No, it does.

00:30:02.64

Andrew

Like, it's hilarious to hear you say it out loud. And, like, I can't help but laugh because it does seem so obvious.

00:30:06.98

Sean

Right.

00:30:08.54

Andrew

But, I mean, you're 100% right when you're in the thick of it. It does feel like the emotional feeling is I need to be figure out how to do everything.

00:30:17.10

Sean

For sure.

00:30:17.23

Andrew

And, like, it is so easy to forget that all of these are choices.

00:30:17.30

Sean

Yeah. Yeah.

00:30:22.31

Andrew

Yeah.

00:30:22.99

Sean

christian

00:30:24.63

Andrew

And like the beauty of the beauty and the curse of being founder, being the one calling the shots is that you truly do have choice on everything.

00:30:38.17

Andrew

Almost everything versus like, you know, a lot of times you're an employee, you want to keep your job. You don't really get whole lot of choice. You need to do certain things to like, you you need to check certain boxes.

00:30:49.42

Sean

Yeah. Yeah. Yeah, totally.

00:30:52.44

Andrew

Same is true to a certain degree. You know, as a founder, you build up obligations for yourself and you have to continue to fill those obligations in order to you know, keep things running. But you just have so much more, I think, choice.

00:31:05.24

Andrew

And, but it's easy to forget that.

00:31:08.26

Sean

totally Yeah, i've been pushed I haven't pushed a lot on the SaaS stuff. I owe a dev an email back on some things. What else?

00:31:19.80

Andrew

By the way, this is why i think it would be really interesting to see you try to hire someone whose sole job is to push on the SaaS stuff.

00:31:26.24

Sean

Oh, totally. That is my, like, one... That is my, like, my one goal at MicroCoff, actually. Or, like, one possible mission is to make it a high.

00:31:32.48

Andrew

Oh, interesting. You're trying to recruit a little bit.

00:31:36.01

Sean

Yeah. if i If I bum it in someone, you know, it's like serendipity.

00:31:38.73

Andrew

Yeah.

00:31:39.43

Sean

It's not. We'll see. has to be, like... i think it has to be the right person.

00:31:43.64

Andrew

Mm-hmm.

00:31:44.79

Sean

and And I think, you know, they have to pass, like, more than just me. It's like... I do realize that I'm incredibly...

00:31:56.47

Sean

bad at recruiting good professional people versus like my my instinct is like, yeah, you seem cool. Like, don't want to come work here.

00:32:06.53

Andrew

Vibes and talent.

00:32:07.67

Sean

Yeah, yeah, yeah.

00:32:08.20

Andrew

you're You're good at identifying vibes and talent, but easily overlook the like ability to get shit done and stay organized.

00:32:15.36

Sean

yeah yeah the like yeah yeah the organization the communication all that sort of stuff yeah i love a i love a cracked out you know designer of a wild card wild cannon yeah loose cannon type of guy yeah but i think i think that'll be my one very loose goal at microcom but i'm also trying not to go to it with like uh

00:32:16.96

Andrew

Yeah.

00:32:31.48

Andrew

Yeah.

00:32:41.00

Andrew

Too much of like, I need to get something out of this.

00:32:43.98

Sean

Yeah, yeah.

00:32:45.51

Andrew

Yeah.

00:32:45.51

Sean

i am I am relishing the fact that it's like one of the first times I'm going to a conference where I'm not working on the other side. And it feels and feels very weird.

00:32:51.54

Andrew

Yeah.

00:32:53.72

Sean

It feels incredibly weird.

00:32:54.29

Andrew

Yeah.

00:32:55.35

Sean

Yeah. Yeah.

00:32:58.06

Andrew

yeah i kind of wish that in retrospect that i had gone to some microcomps when i had the crit like budget to work with but it was always so hard to justify because it did feel closely enough aligned with like the work that we did that it felt like if i was going i needed to be going for like sales purposes and i didn't really want to go for sales purposes because it didn't feel like the right vibe like microcomp doesn't feel like a place you want to be selling really

00:33:07.67

Sean

Mm-hmm.

00:33:20.96

Sean

Yeah.

00:33:26.23

Sean

Yeah.

00:33:29.60

Sean

For sure.

00:33:30.71

Andrew

and so I didn't really want to go for that.

00:33:30.68

Sean

For sure.

00:33:32.39

Andrew

And like, I didn't know if I fit in. Rob has reached out to us and said like, yeah, you fit in at all or welcome, which I think is super cool, but

00:33:34.82

Sean

Yeah.

00:33:39.63

Sean

I still don't believe him. won't believe him. see it.

00:33:44.55

Andrew

I believe him. I think most of the people in that community are like pretty, pretty chill and like interested in other things.

00:33:52.70

Sean

Yeah. Totally. Yeah. Mm-hmm.

00:33:53.49

Andrew

what other people are doing and like open to a wide range of like businesses and stuff like that i mean like uh my friend brian castle is has been in that community for a long time has even you know i think raised from tiny seed and yet is doing building a service business and building

00:34:09.70

Sean

and so

00:34:15.31

Andrew

you know, product businesses and like kind of info product businesses and he's doing a whole range of things. So yeah, I think people are pretty open. and

00:34:24.23

Sean

Yeah, I'm excited, I think. Yeah, I mean, I'm excited to have the episode recording with you after the entire event.

00:34:33.09

Andrew

Yeah, I'm excited for that too.

00:34:33.35

Sean

and how you Yeah, we have to we have to push next Wednesday.

00:34:35.79

Andrew

I want to just like

00:34:37.68

Sean

I've got hop into an airplane.

00:34:39.67

Andrew

Cool. OK.

00:34:40.24

Sean

Okay.

00:34:41.25

Andrew

Yeah. Yeah, let me know when you want to do that, because I definitely want to do that one as soon after you get back as possible so that we can like I can get the fresh takes.

00:34:49.90

Sean

Totally. Totally.

00:34:50.72

Andrew

Cool.

00:34:50.86

Sean

Totally.

00:34:52.24

Andrew

Cool.

00:34:53.20

Sean

Yeah. I could call you to from the airport. me just...

00:34:56.18

Andrew

Sure, if you want. I think more and more I'm being becoming convinced that like quality of our podcast does not matter at all. like like Sound quality is nice to have when we can have it.

00:35:04.79

Sean

Oh, really? Okay.

00:35:08.40

Andrew

But you know I listened to an episode of the Mostly Technical where Aaron Francis was just like on AirPods in a hotel room.

00:35:15.41

Sean

Yeah.

00:35:19.39

Andrew

And I was like, this is fine. It's not bothering me. I don't care. I think there's a there's a limit. It gets to a certain point like I don't really like listening to live recordings like when people when podcasts will go on tours and do like live episodes with like audience reaction and stuff.

00:35:25.51

Sean

sure

00:35:36.19

Andrew

I just don't really like that.

00:35:38.03

Sean

I think it throws I think it's like like are you saying because like like pause to the audience and they kind of break the flow is that what you mean yeah it makes sense makes sense

00:35:46.03

Andrew

Yeah, yeah, yeah. It just feels kind of awkward and like it's just different than what I'm usually coming to a podcast to listen to. Yeah.

00:35:54.49

Sean

Yeah, fair. Fair, fair, fair.

00:35:58.40

Andrew

So I want to talk to you a little bit about this, some ideas that are ruminating in my head that are that are still pretty half formed for how I think we could...

00:36:09.50

Sean

Grow meta monster.

00:36:11.37

Andrew

it's It's not even so much about growth. It's more on the product side of, like, how do we unlock a lot of capability of Metamonster?

00:36:14.82

Sean

Okay.

00:36:18.99

Andrew

And I'm concerned that we might be thinking about building AI features like we have always thought about building traditional features and that that might be the wrong approach for AI features.

00:36:27.64

Sean

Okay. Say more.

00:36:32.82

Andrew

Okay.

00:36:31.49

Sean

okay say more

00:36:35.00

Andrew

So...

00:36:36.68

Andrew

We've been building, adding ai into Metamonster. We've been kind of taking the position of like, we're going to hide away all the prompting and the models and like all of the the details.

00:36:48.87

Andrew

And we're just going to like build AI into Metamonster as just like, any standard feature. So you can generate a title, you can generate a meta description.

00:37:00.04

Andrew

And if we want to give you something else to generate, we give you a new button to press and where all of that stuff is obfuscated away and hidden from you. And that makes it very easy to understand what Metamonster does and to to be able to use it.

00:37:17.42

Andrew

I kind of want to flip that on its head a little bit and explore building a more open-ended AI feature into Metamonster.

00:37:28.70

Andrew

And some of this is coming from looking at another tool that is similar to us called AeroPS. Aerobs is like, they're trying to be something much different from Metamonster.

00:37:39.90

Andrew

They're more of like, you know, one of these AI workflow builders, like, you know, a, yeah, like a Zapier make like a workflow orchestration thing, but for SEO specifically.

00:37:46.80

Sean

Zapier make Gummy Loop.

00:37:55.25

Sean

yeah

00:37:56.24

Andrew

And so they're trying to build tooling where you can basically come in and do anything you could possibly want with AI. related to SEO and they've just got the tools and kind of the infrastructure to organize everything.

00:38:10.75

Andrew

They're on like the far end of the spectrum where it's like totally open-ended, very like kind of much more data science-y approach to to using AI where it's less about like a use case, a specific use case and a thing and more about like giving you sort of the tooling to use AI however you wanna use it.

00:38:33.75

Andrew

And I don't want to go that far. But yeah know i was thinking about like if we gave you the ability to, say, add a field to our your pages table that shows you all of your pages with your titles and meta descriptions and stuff.

00:38:54.27

Andrew

So add a field to that and run custom prompts in that field and start to open up some prompting stuff for you. Then people could suddenly, like the power of AI is, it's not that it's accurate because it's not, right? It's a it's a fucking slot machine.

00:39:14.97

Andrew

The power is that it's flexible, that it can adapt and do lots of things, tackle a very broad range of tasks. And the way that we're building it into Metamonster right now, we're, we're like reining in all of that flexibility and like using it more like a, you know, a laser, an arrow, a like a single thing, a lever.

00:39:36.17

Sean

Yeah, there's a reroll button for what one thing.

00:39:38.77

Andrew

Yeah. Yeah.

00:39:40.10

Sean

So.

00:39:41.95

Andrew

But like, imagine if, if you like the beauty of Metamonster is we've built a crawler that integrates with AI. And so we can like, we have your site con we have all of your content and your data which like screaming frog you have to manipulate pretty heavily to actually get the content into a workable form but like we have all the content from your site and then we have like ways to tie that content to ai and so it's like imagine if you could modify and adapt prompts inside of metamonster then like we could release a single feature custom prompts and with that feature you could suddenly generate h1s generate h2s you could

00:40:29.71

Andrew

generate structured schema and like you could generate anything as long as it just needed the context of that one single page.

00:40:39.90

Andrew

And then i could start to use Metamonster to develop new features for Metamonster because I could build prompts inside of Metamonster, test them, run them, get them to where I like them, and then we could release them as like templates or something like that or like standard prompts or move them to a button, you know, move them to a more traditional feature. Yeah.

00:41:03.68

Andrew

And then you could eventually expand it to go beyond like, you know, I think we'd start with just giving people, here's the context for a single page, what can you do with the context from a single page, but then eventually we could give people tools to pull in context for multiple pages and pull in, know, reference their search console data and reference their Simrush data and do all of these other things and look at Metamonster as the way to tie all of these things together. And I think this would always be more of a power user thing. like I think we'd always, I want to keep the first run flow as simple as possible. And so this would all be like kind of hidden away.

00:41:43.33

Andrew

But it's there when you're ready to dive deeper.

00:41:48.74

Sean

I think it's interesting. I have two thoughts. One is that my security like security brain kind of like raises a lot of concerns and alarms about about this.

00:42:00.57

Andrew

Oh, well, yeah.

00:42:02.31

Sean

Because you're leaving an open open prompt field and I can only see prompt injection happening. But I don't know how much you care about that necessarily. so there's that.

00:42:15.60

Sean

The other thing is

00:42:19.62

Sean

Alex Bottschillian-O'B.com

00:42:24.64

Sean

generate things based on page content that in itself is cool valuable and i see a lot of things that can build off of it and think about in the context of meta monsterer that's where I get a little bit lost. Like, like here's like, here's like things can see what you're saying, right? Like, uh, suggest new meta titles, suggest better page content.

00:42:50.07

Sean

you know, suggest blog posts based on what you're saying. Like, is that what you mean? Or,

00:42:55.90

Andrew

No, not

00:42:56.82

Sean

okay

00:42:57.74

Andrew

no not really

00:42:58.42

Sean

What are like realistic things you would build with a custom prompt then?

00:43:02.09

Andrew

So the same things we've been talking about building ourselves. So like the thing that's one of the things that's been driving me crazy about how slowly we've been moving, yeah we've had to do all of this like stuff to tune and improve our crawler and add all the boring SaaS features.

00:43:18.52

Sean

Mm-hmm. Mm-hmm. Mm-hmm.

00:43:20.75

Andrew

And I want to add more capability to Metamonster. I want it to be able to do more, be more powerful, provide more value. And so there's a million things we've talked about adding. Structured schema.

00:43:31.02

Andrew

You want to generate structured schema for a page? All you need to know is, like, you need an AI, you know, you need to create a prompt, and you need the context from that page. uh og descriptions and titles and potentially even og images and stuff like that you you know oge images gets a little hairier so like you probably don't want to do that h1s h2s like optimizing h1s h2s checking to see if they have the primary keyword in you know rewriting them to use the primary keyword rewriting them to be

00:43:48.80

Sean

Sure, sure, sure. Sure.

00:43:53.34

Sean

sure sure

00:44:07.26

Andrew

in a better tone of voice or a better more descriptive or more action oriented or things like that you know these are the things that people have been asking us for in addition to like you know some of this stuff gets a little again gets a little hairier open graph images gets a little hairier internal links is like hard to do just with the context from the one page you need to know all of the other pages that exist and which pages are the most important and

00:44:26.65

Sean

Mm-hmm. Mm-hmm.

00:44:32.71

Andrew

So that that gets a little hairier. But a lot of the like basic on-page SEO optimization stuff you can do with just the context from the page.

00:44:44.02

Andrew

Yeah, just the context of from the page. And so it's like, how do we... you know We could build each one of those things one by one, and we could be like, all right, now you have now we surface page H1s for every page, and we give you a button to optimize your H1s.

00:44:59.20

Sean

Mm-hmm. Mm-hmm. Mm-hmm.

00:45:05.63

Andrew

And then, OK, now we do structured schema, and now we do OG stuff, and now we do yada, yada, yada. Or we could create a single feature once that lets you build custom prompts and create prompt templates and then just sort of release it like that.

00:45:27.48

Andrew

And then, so like that's, that's,

00:45:28.16

Sean

Yeah, just let yeah you anyone

00:45:31.35

Andrew

Yeah, so that's where my focus is right now, although I do think you could go broader with it. You know, you could find ways to so access the context from the whole site for and doing things like internal links,

00:45:47.42

Andrew

access information about the links on a page so that you could... you know surface things, you you could end up building like ah ah you know a rules engine to be able to let people customize how they surface issues i don't know if we want to go that far and then like i guess kind of what you were saying is like you could actually and austin had this thought too is you could go the route of like let me chat with my website let me talk to my website and like and start going almost the cursor direction where it's like i'm

00:46:08.42

Sean

Mm-hmm. Mm-hmm. Mm-hmm.

00:46:27.18

Andrew

you know, I can load context from, I can pick pages with context to load and I can ask questions about those or like try to generate things or I can like work within an agent sort of model and like make changes and and stuff. So, I mean, you could end up going, taking it that far potentially, but, but yeah, for right now, I'm really thinking like,

00:46:51.94

Andrew

I think starting point is just like, how do we enable all of these things use the same infrastructure. So how do we enable people to suddenly do any on-page SEO optimizations they wanna do by building one feature instead of building 20?

00:47:09.00

Sean

I mean, I think i think that fundamentally like is good. right the I think that's smart. Because i mean, I think every feature is very much just like like I imagine the back end wise, it's like the same type of thing with a different prompt, right?

00:47:29.21

Sean

Like the generate, like a generate button and it's like the same type of, you know, i I only imagine the meta title generator and the meta description generator are very similar outside of just the prompt and the text area size.

00:47:40.15

Andrew

Yeah. yeah

00:47:42.13

Sean

So similarly, like, so if you, yeah, if you just build like a custom prompt generator, I think it's cool. I think it makes it very flexible. And then I think there's some like,

00:47:54.80

Sean

cool extensible things i just like I just would expect that it to go out of it just being metamonster and it just being like web monster like web monster like one of my one of the first things that I would think of is like a QA bot that was like tell me if there's typos on the page and uh lorem ifsum on it because it happens like uh it happens i literally was on a client's not a client like a prospect site like oh that's a typo right and then i don't know how like deep it can go into later on but you know like like if it's not just markdown of the site but if it's like page code itself then i would you know

00:48:26.22

Andrew

Yeah.

00:48:40.69

Sean

Maybe i'll I would say like in in the page code, tell me, and don't know, like weird div layouts or weird whatever. I think that is like fundamentally cool.

00:48:56.52

Sean

I just would also expect it to also go like, look at this page, give me, you know, FAQs that to the bottom, for example, or other sort of things.

00:49:07.10

Sean

There's a, but

00:49:08.75

Andrew

That's an interesting one too, is like, we've been taking the fundamental hypothesis that people aren't, that there's going to be a group of people who don't want to use AI for like content generation.

00:49:10.41

Sean

Mm-hmm. Mm-hmm.

00:49:19.82

Andrew

They want to use it for your like optimization. And I think that like, I'm seeing stuff in the market happening that's challenging that, that hypothesis.

00:49:31.03

Andrew

And the other thing is like, if we give people something more flexible, then, we let people show us what they want.

00:49:41.25

Sean

Yeah.

00:49:41.39

Andrew

They, they can, we give them the opportunity to show us what they want in what prompts do they create?

00:49:41.44

Sean

Mm-hmm.

00:49:47.93

Andrew

What, did what, what do they hack this thing to do? It, the risk, I think the biggest risk is just that like, we make it more complex and like harder to use, harder to understand, like,

00:50:01.82

Andrew

If we go too broad, it becomes harder to like wrap your head around like what Metamonster is. But yeah. yeah And so that's why like none of this is like a decision that has been made. It's just like, these are just ideas that I have.

00:50:18.76

Andrew

But it feels like there's this fundamental idea of like, the power of ai is its flexibility and right now we're using it in a way that is very inflexible and so like how do we open up some of the flexibility in order to allow us to move faster and learn faster

00:50:35.96

Sean

I think it's...

00:50:39.27

Sean

I think it's cool because I don't think you'll ever, don't know if you'll regret building that feature because it sounds like you can just then people can like, sounds like it's just a, like a framework for people to build for you to also then build the, like QA bot or a, like whatever other, like a semantic, uh, a, like an H1,

00:51:04.35

Sean

h one like like there there are many apps right that's sort of I'm thinking about it so like I don't think there's a problem in you building basically like a like a app language like I don't know what phrasing is but like like your own like Metamonster SDK but not an SDK yeah

00:51:07.97

Andrew

Mm-hmm.

00:51:29.89

Andrew

Yeah, I mean, I want it to be way, way, way simpler than like an SDK.

00:51:30.48

Sean

you yeah

00:51:33.21

Andrew

I don't want people to have to to start.

00:51:35.42

Sean

I know. Uh-huh.

00:51:36.65

Andrew

And OK, I hear you.

00:51:37.14

Sean

like

00:51:37.93

Andrew

youre that You're just a technical person. And so this is like the language that your brain uses, right?

00:51:43.21

Sean

Yeah, i don't I don't know what the, to like a tool maker.

00:51:43.97

Andrew

Yeah.

00:51:45.69

Sean

Like, I don't think there's a problem with you building like an app tool maker because you will just be able to make like speed up development later down the line because you can now generate an H1 thing and all these things.

00:51:57.12

Sean

And like, there's no reason that someone in first run should have all those things, but they should be able to click on like additional, like a plus button that's like additional, additional optimizations or additional monsters and they can then scroll through all the,

00:52:10.21

Sean

things and choose the ones that they want or make their own.

00:52:10.35

Andrew

Yeah.

00:52:13.00

Sean

So think it's I think you won't regret building it. I did just send you this thing I found.

00:52:17.69

Andrew

I think the.

00:52:18.76

Sean

For sure.

00:52:19.67

Andrew

Where we could regret building it is, like again, it adds complexity. And so figuring out like what the right level of complexity is for us.

00:52:30.88

Andrew

And like you know the best marketing has an opinion. I think the best product often has an opinion, too.

00:52:36.86

Sean

For sure.

00:52:37.42

Andrew

And so if you try to make something that's too flexible, you end up like kind of reinventing Screaming Frog, which is what we don't want to do. Like, we want to have more of an opinion. we But at the same time, I think often the software that succeeds long term is relatively flexible.

00:52:53.37

Andrew

Yeah, I don't know. i guess i I guess I need to think more on that.

00:52:55.81

Sean

well

00:52:56.48

Andrew

I might, I might, I'm not sure if I do feel strongly that the best software has an opinion.

00:52:56.83

Sean

right

00:53:00.91

Andrew

Because, like, the best software is, like, often pretty damn flexible.

00:53:07.83

Andrew

Hmm.

00:53:08.89

Sean

I just like brain dumped a blog post about this last night actually. think that

00:53:16.72

Sean

i think that uh i think that you are probably right that the best products can be ultra flexible but i also think that a lot of product like there's a long period of it being flexible to it being the best product for example like i like

00:53:37.38

Sean

for i don't think notion is the best product for anything

00:53:40.70

Andrew

Mm-hmm.

00:53:41.76

Sean

besides ah ah being notion i think that's because like i would i don't want to use the notion of project management because notion project management and notion requires everybody to follow a singular like happy path and they don't have enough like guardrails and things put in place for someone like

00:53:59.23

Andrew

Mm-hmm.

00:54:00.54

Sean

Like, you know, back when we when we tried Notion for project management, a designer could just like accidentally delete a table or accidentally delete a card and and now it's now it's fucked, right?

00:54:02.98

Andrew

Mm-hmm.

00:54:10.80

Andrew

Mm-hmm.

00:54:12.00

Sean

I know that they have data table locking now, but I don't think the guardrails are as granular.

00:54:16.19

Andrew

Mm-hmm.

00:54:17.73

Sean

um like But like ultimate flexibility.

00:54:19.96

Andrew

Notion is, like by the way, the exact thing that popped into my head. It's so interesting like because it is one of the most flexible pieces of software that's been built in like the last 30 years.

00:54:23.88

Sean

Yeah. don't know.

00:54:29.13

Sean

Right. But I think it is like, I think it is like, at some point you are building like a no code language, right?

00:54:30.77

Andrew

It's wild.

00:54:40.15

Sean

Like it gets closer and closer to like, yeah, to me it gets closer and closer to like trying to build like an ultimately flexible programming language where you have functions and these.

00:54:41.46

Andrew

Maybe. Yeah.

00:54:52.28

Andrew

Yeah, I think in my head, it's like the danger is you try to copy

00:54:53.67

Sean

Mm-hmm.

00:54:57.74

Andrew

I don't see it being going quite that far, although i i see your point and why you're thinking of it that way.

00:55:01.64

Sean

For sure. Right.

00:55:02.92

Andrew

i think the danger is that you end up building another generic AI workflow builder like Aerops or, you

00:55:07.86

Sean

but

00:55:11.02

Andrew

And I have friends working on, they probably wouldn't use the word generic, but AI workflow builders. And they very much want to go as broad as humanly possible because they've all raised VC money and they're trying to grow super fast and they're that's the thing they're doing.

00:55:25.36

Sean

right

00:55:26.88

Andrew

And like so we don't want to go that broad. But is there a happy medium where we can be like where we can have, again, this like focused first run path with like a small feature set that is like kind of built in but then have some extensibility options that make it possible to just do a lot of things with Metamonster that were hard to do without it.

00:55:55.52

Sean

Well, I think going back to the point, but point that I

00:56:01.75

Andrew

Yeah.

00:56:05.29

Sean

it sounds like it's too like

00:56:09.20

Sean

Okay, one, would you be okay if, let's say you had this custom thing inside of Metamonster, would you be okay if the custom prompt was then used for things not meta-related slash not SEO-related, right? If it got into content, if it got into QA stuff.

00:56:25.82

Sean

And if yes, it sounds more like, then it shouldn't be meta title-related. It should just be that you've built WebMonster, the thing that

00:56:35.02

Andrew

Well, I never intended Metamonster to stay meta titles forever. Meta titles are fucking stupid. I don't want to do meta titles forever. Like page titles, page titles matter.

00:56:44.87

Sean

Yeah, I know. and

00:56:47.71

Andrew

Page descriptions basically don't matter.

00:56:50.68

Sean

Mm-hmm.

00:56:51.16

Andrew

And like, this was always just a starting point. Like my, like the whole thing is Metamonster is just a name and like screaming, like my vision this whole time was like screaming frog that can fix things for you.

00:57:01.21

Sean

Mm-hmm.

00:57:05.70

Sean

yeah okay my my my point is that it it like screaming frog that can fix things for you stays in like the realm of like seo so take take meta replace meta with like seo monster like i guess my question is like how are you going like because ai is infinitely flexible how are you going to limit my usage of its extensibility because the

00:57:26.93

Andrew

i don't think i have to limit your usage right like you you can sign up and make there are people who like use uh like i'm trying to think of a good example like

00:57:37.37

Sean

yeah sure

00:57:43.34

Andrew

It's like, okay, you're an agency, right? And you know someone comes along who's not a cybersecurity company, but you think they're fucking cool and you want to work with them.

00:57:46.28

Sean

yeah

00:57:52.48

Andrew

You can work with them. You're not going to put them in your portfolio.

00:57:55.63

Sean

sure

00:57:56.72

Andrew

I think you can do the same. I think the same thing is true of like software products. If someone comes along and wants to pay me a couple hundred bucks a month to use Metamonster, and they're using it for some weird use case that I don't totally agree with...

00:58:08.87

Sean

Sure.

00:58:09.22

Andrew

fine cool use it i'm not going to build features for you i'm not going to like design the product for you i'm not going to like intentionally shape it for you but if it happens to work for you and you want to use it fine i'm not going to put you in my marketing i'm not going to like make you my focus but if you use it for an unintended use case i don't care

00:58:29.04

Sean

sure

00:58:35.62

Sean

Agreed. Well, I think the point that I'm trying to ask, or the thing I'm trying to ask is like, I

00:58:46.13

Sean

and lost the point that I'm trying to ask.

00:58:47.93

Andrew

Like there is a risk that you could get distracted, I think. there is a risk that you could get distracted and you could dilute your positioning, you could dilute your user experience.

00:58:56.15

Sean

Mm-hmm.

00:58:56.65

Andrew

Like you could end up getting pulled in a direction that is too broad, too unopinionated, too kind of bland and uninteresting.

00:59:04.11

Sean

Mm-hmm. Mm-hmm.

00:59:06.38

Andrew

And I think that that is the biggest risk here. But i also think, you know, it just, it feels like we're not leveraging the technology to its full potential if we,

00:59:19.26

Andrew

put if we try to harness it in too much.

00:59:23.24

Sean

Yeah, i think I think my question, yeah, I think the thing that's that's bringing my brain is like, I don't know how you stop it from becoming like a generic builder. Maybe there are like other guardrails and features that keeps it locked into SEO specifically. And you know, I think there's a lot of SEO apps to build on top of it.

00:59:42.25

Sean

Like I sent you like a semantic SEO topic list that I found yesterday. It has like hundreds of things someone could be doing, right? and each individually could be like a custom prompt or a custom meta monster seo monster app i'm with you there i think my like like how do you keep hmm

00:59:49.19

Andrew

Cool.

01:00:00.32

Andrew

i think you I think you shape it, like you prevent it from going off the rails with like product discipline, right?

01:00:05.64

Sean

hmm

01:00:08.59

Andrew

product and marketing discipline. You stay focused on your ICP and you keep the, like, yeah, it is, it's like a million tiny decisions. It's like, you're asking like how to, I feel like you're asking the same question as like, how do you keep your culture strong?

01:00:25.77

Andrew

It's a million tiny decisions. It's, it's, we're going to do this and not that.

01:00:27.77

Sean

Sure.

01:00:29.66

Andrew

We're going to fire this person and hire this person. We're going you know, we're going to tolerate this behavior and not this behavior. It's a million tiny decisions that turn, into like a company culture. And I think the same is true of like product and marketing is it's like a million tiny decisions that shape what, how people see your product,

01:00:48.81

Andrew

that, you know, marketing decisions, copy decisions on your website, you know, what features to hide and highlight and, and hide and, and all of these things.

01:00:51.44

Sean

yeah

01:00:58.28

Sean

Thank you.

01:00:59.50

Andrew

And I totally think it's a risk going more broad, but I think it's also a risk staying super narrow.

01:01:07.71

Sean

Yeah, I mean, just for the record, like my vote is for building the custom prompt thing.

01:01:12.12

Andrew

yeah

01:01:12.18

Sean

just like I'm not arguing against you not building the custom prompt thing.

01:01:15.65

Andrew

Yeah, yeah, yeah. I know.

01:01:16.44

Sean

just I just think that you have touched on the more important part of like you also have built a cool thing, which is an AI scraper for...

01:01:27.09

Sean

for an ai like like like an ai scraper for a website like a website and to me that sounds like there's a bunch of things you can extend off of it and i guess like my question my question isn't like my question is like is it actually worth keeping it just locked into seo and then building down this path or is it just like

01:01:32.74

Andrew

Okay.

01:01:51.86

Sean

like you have built like a cool core of a product and adding custom prompts means that it could just be all these other things.

01:02:01.74

Sean

Does make sense?

01:02:02.04

Andrew

yeah okay interesting yeah that's a good question uh yeah and i think and i wasn't necessarily trying to argue against you that like we have to do things a certain way i was just i was trying to sort of figure out how to verbalize like some of this stuff

01:02:03.34

Sean

Hmm.

01:02:23.26

Sean

yeah for sure sure i think you have a good point i think like if you wanted to stay in seo you could definitely make a bunch of decisions to stay in seo i think yeah i think there's a point where i'm like wow why

01:02:38.25

Andrew

Yeah. And I don't know that I, like, the typo thing is a good example of, like, that's not really an SEO thing. And it's something that just feels so obvious to me that, like, it just, it feels like it fits within MetaMonster.

01:02:45.32

Sean

yeah so totally totally

01:02:50.96

Andrew

Like, typo or lorem ipsum detection or whatever. Like, that, those things just feel like they fit. And so, but I also, like, I think that, like, staying in SEO for now makes sense just because it helps make our marketing clearer.

01:03:10.91

Andrew

And then, and I think that, like, over time, you know, you could morph outside of SEO.

01:03:09.79

Sean

Yeah. Yeah, yeah.

01:03:17.03

Andrew

But I think also, like,

01:03:19.08

Andrew

the kinds of people who use SEO tools are often... They're not shy. They're not like afraid of those tools because they're called SEO tools. They'll use them for whatever they need them for.

01:03:32.20

Andrew

Like maybe, I don't know.

01:03:34.36

Sean

yeah i i do think that like your point about like the fact that the typo thing isn't totally an seo thing but it feels like it fits i think that it is possible that you will run into a lot of those things where it like it kind of feels like it fits in seo because it's web related but actually it's just because it's web like i think of the thing you build is like web monster

01:03:34.53

Andrew

I don't know.

01:03:47.18

Andrew

Mm-hmm.

01:03:50.93

Andrew

Mm-hmm.

01:03:56.49

Andrew

Yeah.

01:03:56.95

Sean

And especially you had custom prompts to it, like, like, you can just skin it as like, I want to do SEO things and then versus other things versus like, I want to do content things. I do think there's product decisions with like, because like, the thing that it's missing is the thing that would be missing is like integrations to pull other like,

01:04:18.30

Sean

like I wouldn't make it my like content writing suggestion platform because I can't feed it more data, right? I can't give it context about something. I can only give it context about a website.

01:04:29.94

Andrew

Yeah.

01:04:30.97

Andrew

And I still don't really, i don't want it to be that yet.

01:04:34.58

Sean

Right, right, right.

01:04:36.18

Andrew

Yeah.

01:04:36.22

Sean

But, you know, if if you were to do custom prompts, I would just start scraping competitor sites and build a custom prompt of like give me an like, you know, and give me like steel man arguments against these claims.

01:04:51.64

Sean

and then use that as copy for my actual website, for example, is like, yeah.

01:04:51.07

Andrew

Hmm.

01:04:56.66

Andrew

That's interesting.

01:04:58.41

Andrew

that's wild i my mind would have never gone there but that's that's the other thing is i'm like if we do this again we have these risks that we have to watch out for but we also have the chance to learn so much faster because it's like we can learn what people want to do by just watching what they do yeah uh i mean i guess great noise this is why andrew originally built gnql is he just wanted to watch like

01:05:07.53

Sean

Hmm.

01:05:14.26

Sean

For sure. For sure. Yeah.

01:05:23.42

Sean

Hmm.

01:05:23.50

Andrew

figure out what people wanted to search for. And then the challenge is like, it's really hard to reign that in and to now start building like productized offerings around those things and to,

01:05:25.72

Sean

Hmm.

01:05:33.93

Sean

Yeah, yeah, yeah. Data science is actually a really hard thing to do.

01:05:36.82

Andrew

yeah.

01:05:37.42

Sean

Correctly?

01:05:37.62

Andrew

Yeah.

01:05:37.62

Sean

Yeah, yeah, yeah.

01:05:38.10

Andrew

And also just like, it's kind of hard to like look at the G and QL queries and figure out what the trends and patterns are.

01:05:46.09

Sean

Yeah.

01:05:47.32

Andrew

But yeah.

01:05:47.30

Sean

Because data science is hard. Data science is very hard.

01:05:49.26

Andrew

Yeah.

01:05:50.14

Sean

This is why, yeah, Open Door or whatever, like, died as a company.

01:05:54.26

Andrew

Yeah.

01:05:54.49

Andrew

Cool.

01:05:56.88

Sean

Anyway, I'm all for the custom prompting.

01:05:56.98

Andrew

cool

01:05:58.59

Sean

I think the custom prompt adds a lot of creativity. i i think it'll just push you farther and farther away from SEO monster which I totally think is fine like I think I know it's it's not in just your vision to do the meta monster thing just think you are

01:06:16.30

Andrew

By the way, there's a little seedling planting in my head that like our ICP might end up not being SEOs. it might be

01:06:23.56

Sean

hmm hmm

01:06:24.84

Andrew

it might end up being all the people who have to do SEO-related stuff that aren't really SEOs themselves, like you marketing agencies that do bunch of have to do a bunch of content but aren't really...

01:06:36.30

Sean

Yeah.

01:06:41.32

Andrew

ah ah seo isn't their bread and butter. And so...

01:06:43.00

Sean

yeah

01:06:43.26

Andrew

like it might be that that's actually our ICP and our value prop is like, Hey, we make SEO so much easier.

01:06:51.88

Andrew

Yeah. So that's, that's another inkling that's like rumbling around in my head.

01:06:57.44

Sean

i I think if an example of a really good, like flexible product with opinions is like, honestly, is like an air table.

01:07:11.12

Sean

because I think it has significantly more guardrails than Notion and has things that you are allowed to do and not allowed to do.

01:07:16.92

Andrew

Mm-hmm.

01:07:18.34

Sean

And then it also has extra power thing, power user things with scripting and all this sort of stuff. And I'm getting here because like the nice thing about Airtable is that it does, it is able to like, you know, you can do like all these, like it can be a project management tool if you really want it to be.

01:07:35.57

Andrew

Mm-hmm. Mm-hmm.

01:07:37.13

Sean

It can be all these other things. And at the very least, Airtable as a project management tool is arguably better than Notion as an Airtable tool, as as a project management tool. It does require a lot of building and and and stuff like that.

01:07:50.78

Sean

But it's also just an Airtable stand in general. yeah anyway that's my two cents you know build a custom prompting who cares just do it like you can just do things you can yeah just do do robotics and then build a custom prompting and then let me build my comp and tell tool off of it it it does make it does make the the pricing thing really interesting by the way because now like

01:07:56.65

Andrew

Yeah.

01:08:00.86

Andrew

Yeah.

01:08:10.38

Andrew

Sounds good.

01:08:15.27

Andrew

Yeah.

01:08:18.13

Sean

I don't know. I feel like it adds an extra layer, which does make it more viable for it being like a multi-million dollar ARR product, which is cool.

01:08:28.07

Andrew

Yeah. Yeah, my thought, one of my thoughts there, by the way, is that like if we start to expose prompts and then we could also expose models and let you pick your model and we could have like included models, which are basically the cheap ones.

01:08:41.84

Andrew

And then if you want to bring your own model, you can bring your own API key to leverage more expensive models.

01:08:41.87

Sean

Yeah.

01:08:46.31

Sean

Yeah.

01:08:48.87

Andrew

And that so that way we like keep our costs manageable while also not like restricting your usage and stuff.

01:08:56.75

Sean

I think it's a good idea.

01:08:58.90

Andrew

Yeah.

01:08:59.77

Sean

Oh, cool. I got to run.

01:09:01.13

Andrew

right, man.

01:09:01.61

Sean

That was fun.

01:09:01.70

Andrew

This was a good conversation. Thanks for bouncing ideas off with me.

01:09:03.33

Sean

problem. That was good. That was cool.