Sean survives on 1.5 hours of sleep while juggling Black Hat submissions and client launches! 😴 Andrew discovers a new MetaMonster content strategy and shows off the new grid UI that's generating page titles and scoring them for engagement. Plus: should you delegate your therapy homework to Claude (hint: no), auth provider decision paralysis, and why internal linking breaks the table concept.

Links:

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

Transcript:
00:00.81
Sean
Oh, you stopped.

00:03.40
Andrew
Yeah, that would be the most obnoxious sound in the world. I was doing the, like, little kid exasperated sound.

00:10.45
Sean
Yeah, yeah. What is that? What do you, what do you, how do you, like, what is the English phrase for that?

00:16.64
Andrew
have no fucking clue.

00:17.80
Sean
It's not like rolling your R's, but it's like blubbering.

00:21.54
Andrew
It's like flapping your lips. I don't know.

00:23.97
Sean
I guess.

00:23.99
Andrew
I was doing this. Pfft.

00:26.94
Sean
That's a thumbnail.

00:28.02
Andrew
Oh, God.

00:28.21
Sean
That's good.

00:28.89
Andrew
Jesus.

00:32.72
Sean
That's a good thumbnail. I'm going to have Jonah try that out. We'll see how it performs.

00:36.82
Andrew
Try what?

00:37.93
Sean
You just blubbered.

00:40.08
Andrew
Please no.

00:40.97
Sean
Okay.

00:43.70
Sean
How you doing? how's how's How's it going?

00:48.00
Andrew
I'm good. I just pushed out a YouTube video, LinkedIn video for Metamonster.

00:54.94
Sean
Nice.

00:55.68
Andrew
i I think this is maybe going to be a ah new tactic is I saw um like an SEO influencer type share something about how AI overviews really like structured schema.

01:12.35
Andrew
And so I was like, dope. I'm going to go write a prompt to create structured schema in the new MetaMonster UI, record myself doing it, tag this person in the post,

01:22.19
Sean
Cool.

01:24.43
Sean
Sick.

01:24.60
Andrew
and like shout them out in the video. And then did comment on my post. i don't yeah know They didn't share it or anything, but they commented on my post.

01:30.77
Sean
Sick.

01:33.39
Andrew
And so that was kind of cool. And I'm like, I think this is probably a good tactic to follow is just like take the things that people are talking about and just record short videos showing how you can use MetaMonster to execute on them.

01:38.71
Sean
Yeah, 100%.

01:47.17
Sean
Yeah, all all of that makes perfect sense. that like

01:49.42
Andrew
Yeah.

01:50.67
Sean
that That feels like exactly what to do to grow Metamonster.

01:54.27
Andrew
Yeah. Yeah.

01:55.81
Sean
In fact, I would not be surprised if that was the way grew your user base to your first 500 to 1,000 this point. Yet, yet, to to ah thousand at this point so

02:06.85
Andrew
Maybe. i mean, no one is watching, like, our videos aren't getting any, yeah, any views.

02:09.98
Sean
yes yet yeah

02:12.87
Andrew
And, like, my like LinkedIn, most of the people liking my LinkedIn posts about Metamonster have been, like, my friends. So I haven't gotten, yeah.

02:22.08
Sean
yet yeah i think you need you need one you know ah ah you need like

02:26.86
Andrew
Yeah, you need one to go viral-ish in the SEO world. And you need, yeah, I just need to stick with it and...

02:33.43
Sean
yeah plus i mean the content just gets reused so for the blog anyway or or the website anyway so yeah um i absolutely yeah but that sounds perfectly correct to me that's

02:45.76
Andrew
Yeah. I also, um i so I recorded this one today with Screen Studio, was fun to play around with. I like some parts of it a lot. I found myself really wishing, though, that it would generate a transcript like Loom does.

02:59.46
Andrew
can probably use, like, a transcript generator. I can just find something like that. But the reason is, i have gotten alex from like Lex like Lex.page set up pretty well.

03:12.97
Andrew
probably do it in Claude too. But I've got it generating pretty good marketing emails based off of video transcripts. And so it's been like a really nice way to like just, you know, reuse content as I record the video.

03:30.97
Andrew
And then I upload the transcript to Lex and I have Lex generate a, generate a marketing email, I edit that email, and then I copy it over to loops and send it out to the Metamonster list.

03:47.74
Andrew
So I've also been thinking about playing around with some of the automation tools to see if I could actually automate that whole flow.

03:50.05
Sean
Cool.

03:56.02
Andrew
And it would still be, like, human in the loop because I want to edit it before it goes into loops.

04:00.35
Sean
Yeah.

04:00.43
Andrew
But, like, even if it created it as a draft in loops, I could then go to loops and edit it there.

04:04.58
Sean
Yeah. 100%. 100.

04:08.89
Sean
hundred percent and

04:10.32
Andrew
So...

04:12.64
Sean
Yeah, 100%. um I always, well, okay. Sorry, 100%, you can do that. I always struggle with the idea of doing that because I feel like the time it takes to like do that is sometimes not worth the just doing it manually. I feel like I could be automating Okay, one of the things I could definitely be automating is sending you the link to this Zencaster thing. But every single every single time I do it, I'm like, yeah but it's just so easy to just make it.

04:42.92
Sean
And it's so much more work to go figure out how to do it in N8N and like increment the number by one.

04:48.73
Andrew
Super fair.

04:50.98
Sean
Which...

04:51.09
Andrew
This is the classic problem with delegation in general, right? Whether it's automation or delegation, it's just you've got to get over that hump of it's easier to do it myself.

04:57.56
Sean
Yeah.

05:00.78
Sean
Yeah. Yeah. Well... Okay, well, that's that's a different way to think about it. That's kind of like death by a thousand. I'll do it myself until you can't do it yourself anymore.

05:12.62
Sean
I'm not going automate it.

05:12.73
Andrew
Yeah.

05:13.46
Sean
It's not going to happen. I'm just going to send you.

05:14.86
Andrew
Yeah, I'm probably not going to automate this for a while either.

05:15.60
Sean
Yeah. yeah cool okay uh on to another productive topic that we didn't talk about before the show but now i want to know last time we had a call or two times ago we were talking about uh that you were writing down the story of why you would be pushing something off and procrastinating

05:22.29
Andrew
Okay.

05:26.80
Andrew
Yeah.

05:37.62
Andrew
Oh, shit. You just made me realize I haven't been doing my therapy homework.

05:42.10
Sean
Oh, fuck.

05:42.69
Andrew
but

05:42.78
Sean
Okay. Well, never mind. Well, I have. I've been doing the therapy homework.

05:47.74
Andrew
That's great.

05:48.67
Sean
I just, I just throw it at Claude and Claude just, you know, tells me I'm being stupid. And, uh, so, so anyway, uh, helpful, helpful thing to do for what it's worth.

06:02.19
Sean
yeah.

06:01.42
Andrew
Cool.

06:02.81
Sean
Yeah.

06:03.22
Andrew
What have you learned about yourself?

06:03.67
Sean
Uh,

06:05.58
Sean
oh, I just need someone to like, tell me that like, like, that That this is like I'm being stupid.

06:12.84
Andrew
Hmm.

06:13.27
Sean
Is basically, yeah. I just need i just need like someone to just give like basically a reality check me of like of like, you're not doing this because like, like i like I don't know. I would write things i write things down of like...

06:27.82
Sean
Because I feel like it's too easy to do and it's boring. So whatever. And and then Klaas is like, so you don't want to just like, you want to keep like stressing and and wasting brain space about this really stupid thing rather than just like getting it out of the way.

06:44.44
Sean
And then, and then I, yeah, just like, fuck. then I go do it. So. Okay.

06:49.18
Andrew
That's funny. You have outsourced the like part of the homework that is like I think the most important part and if it works for you great but like the whole idea is to train your own brain to inspect the stories and go like is this accurate is this helpful and if not let's replace it with a new story.

07:03.19
Sean
Oh.

07:08.29
Sean
Oh, I didn't.

07:09.47
Sean
Yeah, got yeah but's I forgot to do that. I forgot that was the part of the homework. I thought the homework was just to write the story down and someone else was supposed to do that for you.

07:18.73
Andrew
Now, the idea is to train yourself to do that.

07:19.83
Sean
I see.

07:20.93
Andrew
But, I mean, if Claude helps you, like, get there, i don't know if you're supposed to delegate your mindfulness practice.

07:20.98
Sean
less

07:25.07
Sean
I'm delegating. I'm delegating to Claude. It's, you know, that's.

07:33.60
Sean
Okay, next topic.

07:39.03
Andrew
Oh, Lord.

07:40.36
Sean
um um Well, okay. Anyway.

07:45.78
Sean
Pretend amazing segue. mewa is speaking speak okay Speaking of Claude, Claude's doing great things for you. Or Austin.

07:55.47
Andrew
it is actually yeah dude austin is flying on new meta monster features we're like also sorry for the sniffling my allergies are killing me once again i don't know why my body hates me it's not fun

07:55.74
Sean
Is I'm seeing.

08:01.85
Sean
Yeah,

08:12.93
Andrew
So, I don't know. How many weeks ago did we decide to rebuild Metamonster? Like, a little less than a month?

08:19.62
Sean
yeah like three and a half four weeks.

08:19.77
Andrew
Maybe a month?

08:22.07
Sean
I feel like your timeline was six months and sure doesn't feel like it's going to take six to nine months.

08:22.31
Andrew
Yeah.

08:27.94
Andrew
Well, my timeline was I wanted it done in a month.

08:31.63
Sean
I see. I see. Okay.

08:33.56
Andrew
But the crazy thing is we're actually kind of close to that. So... We started with like the chat UI, and we're playing with that a little bit. And then Austin has built out the flexible grid UI.

08:47.03
Andrew
So basically, like you can import a site crawl into essentially like a spreadsheet inside of MetaMonster. And then you can start just adding things to the spreadsheet with AI and then having it yeah running prompts page by page.

09:03.32
Andrew
And you can adjust the context. You can adjust the model. can adjust the prompt. And the really cool part to me is you can structure the output. so you can like more reliably and consistently like control the types of output you're getting.

09:20.98
Andrew
then you can like filter on that output. So I did this crazy, like a fun little experiment that I did the other day. ah ah had it generate page titles.

09:31.91
Andrew
And then I wrote a prompt and, to evaluate those page titles for how engaging they are, like from a click through perspective, like how, you know, clickable is this on a scale of one to 10.

09:43.66
Andrew
Then I filtered by everything under a five and then had it regenerate page titles just for those. And then score itself on how good the, it did.

09:58.67
Andrew
And so, Yeah, there's like all these cool little loops you can start to build and things you can start to do when you just have this like flexible UI that can suck in website data, run prompts on a per page basis, and then like filter and sort and everything off of that.

10:19.80
Sean
Cool. like I haven't had time to watch any of your videos yet or play with it at all for what it's worth. So cool.

10:26.29
Andrew
All good. yeah

10:27.70
Sean
How, how, how like riddle is the code that you guys are getting back from Claude code right now?

10:36.11
Andrew
I don't know. i think it's pretty solid. Like, like I am still of the belief that like you know these all of these programming tools are more effective if you have at least a basic knowledge of programming.

10:50.84
Andrew
So like Austin is taking the code from Cloud Code and modifying it and changing it. But it is... Like it's scaffolding things really fast for him and helping him like figure out things that he didn't know how to do. Like he was talking about the filters are pretty fucking complicated in terms of like all the different SQL queries because you're kind of like building SQL queries on the fly.

11:13.75
Andrew
And he was like, yeah, it would have taken me ages to do that. without Claude. And the cool thing about Claude code versus like cursor and taskmaster.ai is another one of these that does things in a similar way is they kind of build a plan first and then walk through the plan step by step.

11:34.62
Andrew
And then you can like evaluate each step of the way.

11:38.39
Sean
You can evaluate, like, do you have to, I've not touched clock code, so I don't know.

11:44.47
Andrew
Yeah.

11:44.95
Sean
Do you have to, like, is it, is it like, do you have to prompt it to do that? Or does it just do that? And then it just like truly collaborates with you.

11:53.01
Andrew
The way I believe that it works is like you tell it what you want to build. It generates a multi-step plan first. So like it doesn't start generating code first.

12:03.82
Sean
No, no, no, no, no.

12:03.89
Andrew
It starts generating a plan. And then you like approve the plan or tell it to adjust the plan. And then it goes step by step within the plan and will generate that chunk of code.

12:16.62
Andrew
and then you review it and then it'll generate the next chunk of code and then you review it and it'll and so it'll like kind of work through it so it's kind of the funny thing is it's kind of like you know inserting a product manager it's like hey things get better when you take the time to like figure out what you're building first and like think through the different like use cases and stuff and so it's basically like you know it's just a more organized structured approach

12:35.09
Sean
Yeah.

12:44.02
Sean
Cool.

12:45.18
Andrew
For what it's worth, Claire Liu, do you know who that is?

12:49.35
Sean
You mentioned her. Yeah, chat PRD.

12:52.30
Andrew
Yeah, yeah, yeah.

12:53.24
Sean
Yeah.

12:53.26
Andrew
So she's been doing this like with ChatPRD for a long time. Her product, ChatPRD, I don't remember if we talked about it last time, but ChatPRD generates a PRD, a product requirements document or whatever.

13:01.50
Sean
Mm-hmm.

13:07.81
Andrew
It's all about like generating the stuff or improving the stuff that product managers create, the artifacts that they create.

13:08.06
Sean
Mm-hmm.

13:14.59
Andrew
And then she's been taking those and feeding them into Cursor for ages. And she's been raving about it. And I just, I don't know, I was kind of skeptical, so I never tried it. And now it seems like that's the approach that people are using with Cloud Code and with Taskmaster, is just like build the plan first and then feed that into an agent to generate the code.

13:37.76
Sean
I wonder how different Claude code now is to because I have been doing like the PRD thing with just like Claude 3.7 and then now 4.0 but I wonder I wonder how different or better it is.

13:48.22
Andrew
Hmm.

13:50.79
Sean
Yeah. Okay. I'll give it a shot.

13:55.20
Andrew
I don't know, but Austin's been raving about it.

13:57.83
Sean
Cool.

13:58.04
Andrew
He just pushed up templates this week.

14:02.47
Sean
Cool.

14:02.86
Andrew
So it's, and it's, it's amazing. It's like the product feels light years better when you can actually like create a template and save a template in the product. Cause you don't have to keep copying and pasting prompts over and over again and like restructuring your outputs and all that.

14:19.29
Sean
Cool.

14:20.62
Andrew
So my job, as soon as I get off this call, is to spend the next you hour or two just cranking through templates. And yeah I've got, i spent last week working with Claude to create prompts for like 30 different things, maybe.

14:39.33
Sean
cool

14:39.41
Andrew
And I want to get all of those in the product as templates so that you and like my friend Tony, who's one of our other customers, can start to play with them.

14:48.99
Sean
You another customer?

14:50.36
Andrew
Yeah, we have two customers, two other customers.

14:51.88
Sean
Oh. Well, now I don't feel special.

14:53.82
Andrew
my

14:56.93
Sean
That's cool. That's good.

14:58.43
Andrew
my friend tony is 100 customer uh just to support us i was like i offered him a refund and he's like no no no i want to support like it's it's legit but i don't think he's been using the tool much i think it's he's just like supporting us and then we have one total stranger who

15:01.52
Sean
Nice.

15:13.52
Sean
Gotcha.

15:16.32
Andrew
used one of our discount codes and so is paying 25 bucks a month. I need to check on their usage. i don't remember how much they're using it, but yeah, they haven't canceled yet.

15:21.53
Sean
Cool.

15:25.53
Andrew
So we have three customers.

15:26.32
Sean
you three customers.

15:28.59
Andrew
Yeah, we had a fourth, but and I refunded them because they weren't using it.

15:32.01
Sean
and Okay. i feel um feel more special now that you have three customers. Because now I'm the first. Now I'm not.

15:36.81
Andrew
You are the first customer. You will always be the first customer.

15:37.79
Sean
Yeah. That's true. That's true. But it it feels less special when there's two.

15:42.84
Andrew
Oh, OK.

15:43.44
Sean
You know. Yeah. yeah

15:45.38
Andrew
I'm not.

15:45.62
Sean
Yeah. I'll talk to my clot therapist about it. Anyway. Okay.

15:52.03
Sean
Cool. Okay. Now, yeah, I will i will i will try it out Claude Code. I feel like I keep hearing about it. i'm just

16:00.02
Andrew
I mean, more importantly, try out Metamonster.

16:00.13
Sean
have a bad thing when

16:02.33
Andrew
But ah yeah, yeah, I guess Clog Coach or whatever.

16:05.74
Sean
I think the team uses Metamonster every now and then. Every once in a while.

16:08.71
Andrew
Yeah. i I'll have to ask Guillermo or Yarek if they've played with any of the new stuff. I'll send them a Loom video or something and see if if they can they have time to play with it.

16:19.62
Sean
Cool.

16:22.69
Sean
Cool. ah Cool, cool.

16:23.94
Andrew
I feel like Guillermo is my most likely candidate.

16:27.31
Sean
That's possible. don't know. I feel like Yarg would play with it too.

16:30.55
Andrew
OK, cool.

16:31.20
Sean
Yeah.

16:33.15
Andrew
I'll hit them up.

16:34.56
Sean
Cool.

16:35.36
Andrew
What's going on in your world?

16:37.98
Sean
So the black hat sort of spotlight competition submission was due today this morning.

16:44.01
Andrew
OK.

16:44.47
Sean
So my entire last five days, including the weekend have been helping clients prepare for that. Uh,

16:52.15
Andrew
This is a competition.

16:53.71
Sean
This is competition. So the RSA Startups Innovation Sandbox is like one of the more, was like the popular one and then Black Hat started doing one as well. Differences are, say, little bit more press, 10 finalists, sort of larger sort of thing.

17:08.64
Sean
Black Hat, you

17:09.67
Andrew
Is this a pitch competition?

17:11.48
Sean
Yeah, yeah, it's yeah, it's for it's for like early stage like seed startups. submit a recording. They have like different sort of angles for both for the two of them. And then you go pitch in front of like a panel of judges at the end and and they ask you some questions and someone wins.

17:29.22
Andrew
Is the pitch live in front of an audience?

17:29.40
Sean
I think. Yeah. Yeah.

17:32.69
Andrew
Cool.

17:32.86
Sean
The RSA one is much larger. and but mean Blackout one this year might be larger too. So it's pretty cool.

17:38.78
Andrew
Hmm.

17:40.79
Sean
mean, it's a good, you know, it's a nice little like feather in someone's cap when they get to be a finalist. It's validating.

17:49.94
Sean
But yeah. Editing their videos and and all that stuff until last night and

17:57.14
Andrew
Damn, you guys really are full stack if you're like submitting people's applications to like pitch competitions even.

18:00.01
Sean
yeah

18:06.02
Sean
yeah yeah we really really are so you should don't don't hire us we can't take on any work lines uh i got a sales call uh like an hour and a half after this yeah just got just got so much stuff yeah when it rains it pours i'm tired as fuck i'm like

18:28.16
Sean
I got like yeah one REM cycle last night. So that was exciting. But yes, it was an hour and half of sleep, Andrew.

18:34.47
Andrew
Isn't that like an hour and a half of sleep? Jesus Christ, Sean.

18:38.27
Sean
how

18:39.18
Andrew
A little?

18:39.36
Sean
It's fine. I'm kicking. It's all good. I got caffeine in me. and That's all I need. No, but but things are things are good. i think we are a little bit overextended in terms of projects.

18:49.23
Andrew
A little?

18:51.92
Sean
and just a little bit just a little bit so we gotta like restructure some of the ways we're handling i don't i i think i think we are at the maximum of work we can take on uh assuming we don't like none of these sales calls are like we need stuff right now not that we can say yes at this point but we'll see

19:13.95
Sean
we do need to like restructure a little bit about who's on what to kind of balance out some of it. I think we're like, some projects are very top heavy. Some projects don't like are, there's more like janitorial work basically. So, they're all going okay or well, I think, but stuff to do there.

19:35.37
Andrew
If you're a client and you're listening, they're going very well.

19:39.12
Sean
I think our clients are happy. I think our clients are happy.

19:40.71
Andrew
No, your clients are totally happy.

19:43.91
Sean
yeah, it's it's just like the backend of like, you know, someone working more hours than they should be to help make sure they're happy. So, but yeah, black hat keeps looming. Defcon keeps looming.

19:58.24
Sean
Merchandise keeps looming. All of that stuff. Which is, yeah, I don't know. It's good. I'm complaining about success, I guess. You know, it's a trap.

20:09.24
Sean
I do have a thing I want to pick your brain about. Because this feels like a one-way door.

20:12.34
Andrew
OK.

20:14.69
Sean
And...

20:17.43
Andrew
One-way door as in like an irreversible decision?

20:20.40
Sean
Yeah, or like a pain in the ass to reverse later on. you know

20:23.29
Andrew
OK.

20:25.02
Sean
Basically, we're trying to figure out what to do with auth for margins between Stitch, Clerk, AuthKit by WorkOS, Auth0, all that sort of stuff.

20:28.17
Andrew
Hmm.

20:38.16
Andrew
What's your tech stack again?

20:38.19
Sean
Mainly because Next.js and NeonDB.

20:41.51
Andrew
OK. OK, you're not using Supabase?

20:46.01
Andrew
What is NeonDB? don't know Neon.

20:47.86
Sean
There's also, like, basically the consideration is, like, we have clients that use Okta and are going to want to log in and, like, manage that sort of stuff with Okta. Yeah.

21:03.04
Sean
I don't think you can even get that with Supabase auth anyway. Neon's like a serverless Postgres sort of thing. I think it's a little cheaper. dev just had more experience with it, so he just ran with that.

21:12.81
Andrew
Okay.

21:13.22
Sean
like a I don't have a horse in the race. so

21:17.96
Andrew
Neon has auth, but you think it won't be able to do what you need it to do?

21:22.18
Sean
i think I think it's like new.

21:22.72
Andrew
Hmm.

21:23.97
Sean
I think it's think Neon Auth was rolled out after Superbase rolled it out as well.

21:28.48
Andrew
Okay.

21:29.93
Sean
there's like There's like nuances to like all of all of it, like whether or not it's like enterprise ready and all that sort of stuff.

21:37.77
Andrew
Is Work OS is... ah ah owned by the Laravel team, right?

21:44.15
Sean
Is it? i don't I don't know. I don't think so.

21:48.18
Andrew
I could be wrong.

21:49.39
Sean
maybe Maybe you're right.

21:51.48
Sean
I don't know.

21:53.04
Andrew
I swear, yeah.

21:57.17
Andrew
I swear I've heard them talk about WorkOS and AuthKit. good

22:08.84
Sean
I don't see it on on this thing.

22:11.81
Andrew
Maybe Laravel is just pushing people to use WorkOS.

22:17.44
Sean
I mean, I see a lot of like, like a Laravel starter kit has Work OS already in it.

22:22.56
Andrew
Maybe that's what it is. I've i've just heard ah ah i've heard the guys on Mostly Technical talking about WorkOS a lot and talking about it with the Laravel guy.

22:30.52
Sean
Hmm.

22:34.16
Andrew
And so i was wondering if...

22:37.35
Andrew
Yeah, I was wondering if it was like a... Yeah, I had thought it was owned by them. Okay, anyway, sorry.

22:45.17
Sean
Okay.

22:45.39
Andrew
Auth, auth, auth, auth.

22:47.19
Sean
Just like, I don't know how much I'm like over considering it. I don't know if it's like, you know, ah ah I don't know if, I don't know if like this is like an advice thing that I'm asking for as much as I'm just kind of complaining about it at the moment.

23:00.91
Andrew
Yeah.

23:03.18
Sean
I think we're down to work o or clerk like offuthkit WorkOS or Clerk, or Clerk. I just don't, I think it's that I don't have like a good mental model of like how to choose between these things. Cause it's not, it doesn't feel, it's not a preference thing, but I also don't know like the full requirements that we need. They all have, they all seem like they have very similar features and similar pricing. Yeah.

23:28.19
Andrew
Yeah, I feel like this is one of those ah ah decisions you have to make all the fucking time building a new product is you just, you know, you got to kind of evaluate the trade-offs and then move.

23:29.10
Sean
Yeah.

23:36.04
Sean
Yeah.

23:41.00
Andrew
And like the important thing is just not to get stuck waffling.

23:44.67
Sean
Yeah.

23:45.80
Andrew
Because the reality is, yes, it would be a pain to reverse it, but you can reverse it. You know, we use Supabase Auth because it is super fucking easy to get up and running with and like works should work just fine for us.

23:56.52
Sean
It is. Yeah.

23:59.47
Andrew
Like we can do single sign on, uh, with like Google and stuff. If we want to, we can add that pretty easily.

24:06.17
Sean
Yeah.

24:10.21
Andrew
So I see, again, like part of me feels like you're getting a little stuck in the future of like enterprise land of like, we have to be ready for our enterprise clients.

24:18.89
Sean
Mm-mm.

24:22.50
Andrew
Maybe that's true. but maybe again, you can build for the non-enterprise clients first and then add, you know, Okta support or whatever in the future.

24:35.35
Andrew
Flip a coin. Like, I don't have great advice on this one. i don't think there's a ah ah right or wrong answer. The providers are all probably pretty similar. And as long as, like, the pricing is roughly similar and they both have good API docs and, like, don't have a bad track record of cybersecurity breaches, I think it comes down to just deciding.

24:48.29
Sean
Yeah.

24:57.66
Sean
Sure. I mean, you know, and I, yeah, I mean, I did think, did realize I was like overly considering some enterprise features with like, you know, so some of our clients use Stitch, right?

25:11.90
Sean
Because they're B2B enterprise, they're actual B2B enterprise SaaS applications.

25:15.07
Andrew
What stitch.

25:16.68
Sean
S-T-Y-T-C-H, oh, so another like auth provider, very enterprise focused.

25:21.57
Andrew
Okay.

25:23.43
Sean
And I realized that like their compelling feature is like just-in-time provisioning, which is great when you're, you know, B2B, like a full-on B2B enterprise app and you have like, you know, you create an account and like their roles get created and all this sort of stuff. That's never gonna happen with margins. Like it's too small of a content marketing team for that to ever.

25:42.84
Andrew
Yeah.

25:42.93
Sean
be a thing i can see octa being being a being part of this like like a like a requirement for some of our clients because it just has been for some of our clients even with webflow like they'll pay for it like the enterprise license of webflow so that they can get like whatever skim or like special sso stuff they get with webflow enterprise it's like

26:05.82
Andrew
Yeah, but, like, I wonder if it wasn't there, if it would be enough to stop them from buying. Because that's the real question, right?

26:11.57
Sean
it

26:11.74
Andrew
It's...

26:12.04
Sean
It's happened. we've lost We've lost deals because of that, too. Yeah.

26:15.43
Andrew
That's wild. Because I, like, I think, you know, I think about all the products I've worked on, and a lot of them have not had, like, enterprise-level authentication stuff for a long time, and they get customers and people buy.

26:31.53
Andrew
Yeah.

26:34.84
Sean
Yeah.

26:36.50
Sean
Yeah, I don't know. I don't know if ah ah like how much of this is like because I'm in cybersecurity land all the time and like my influences are cybersecurity SaaS apps. Anyway, i mean for what it's worth, I think we're just going to roll the dice and go with WorkOS and just not worry about it.

26:49.82
Andrew
Cool.

26:51.22
Sean
And if it's an extra something a month, we'll just deal with it. Yeah.

26:57.26
Andrew
I think that's probably the right move right now. i

26:59.34
Sean
yeah

26:59.87
Andrew
you know I wouldn't worry too much about this decision. It's not, you know, I think one of the questions i keep trying to force myself to ask with metamonster and i asked this with like the marketing efforts i'm doing i asked this with the features we're debating and like will this move the needle like we do not have customers we don't have revenue we need customers we need revenue like will will this thing move the needle yes or no and like yeah auth is like a pain in the ass to reverse but like for the most part

27:23.81
Sean
Mm-hmm. Mm-hmm.

27:36.08
Andrew
you know, the value of your product is not off. It's like a checkbox thing that you may have to have for some clients, some customers. um And so if it doesn't move the needle, don't spend too much time fretting about it.

27:51.12
Andrew
And, you know, if you lose a couple of customers in the early days, everyone does, so you can always you know satisfy the enterprise requirements later and charge an enterprise-level price for it.

28:07.58
Sean
I was just going to charge the enterprise level price out the gate with some.

28:10.11
Andrew
Great.

28:10.83
Sean
Yeah. Yeah, fair, fair. All right, well, I'll let you know how it goes. Yeah, our our dev is also very fast.

28:20.40
Sean
this This came up because our dev is really fast and I just kind of like on a whim was like maybe Stitch and then and then he like showed me a demo of it and I was like, oh shit. i didn't know you were just gonna do it but i but i realized i think he was blocked on me figuring out some of the two-way syncing uh work that i owe him yeah yeah anyway that's that's margins margins go it's going like uh i think this was off the call uh andrew and i hopped on afterwards and

28:48.43
Andrew
tool

28:55.07
Sean
He looked at my linear and he was like, why do you have these things in here? None of these things are V1 or V2. Like, why why are these features in your beta for launch? And I was like, oh, shit.

29:08.83
Sean
And our...

29:09.01
Andrew
It's always easier to do with someone else's product than with with your own product. I mean, Austin and I were just doing this yesterday with, like, the new UI is, like, you know, really going through and asking ourselves for every feature.

29:12.86
Sean
Thanks.

29:20.69
Andrew
Like, does that really have to be there? And, like, truthfully, like, part of me feels like we should just ship the new UI as it is now. Like, it fucking works. So, like, we should just ship it.

29:31.94
Andrew
But where we're going to try to push... We're going to wait a little while try to improve the experience a little bit. But, like, you know, yeah, everyone struggles with this.

29:39.26
Sean
Nice.

29:44.52
Sean
Thanks. Thanks. Anyway, our launch, our launch list became like way tinier after that. and now it's like more basic features and more like basic SAS features that we just like forgot password flow, for example, like, like, um, uh,

29:57.04
Andrew
Hell yeah. margins launch win.

30:02.88
Sean
I don't know. let's let's Let's see if we can do it by June 27th. Uh-huh.

30:07.05
Andrew
June 27th. Damn. Nice.

30:09.77
Sean
Well, that's when the editor gets deprecated.

30:12.43
Andrew
Oh, right, right, right.

30:12.52
Sean
So we'll see.

30:15.69
Andrew
Cool.

30:16.40
Sean
Yeah.

30:17.03
Andrew
yeah we're

30:17.50
Sean
We'll see.

30:18.86
Andrew
We're kind of aiming for end of the month as well to have, like, most of the features done.

30:19.06
Sean
At least the...

30:24.33
Sean
Nice.

30:24.81
Andrew
We've been debating... like how if and how to integrate the chat UI into the grid.

30:34.51
Andrew
But I think we're going to keep it pretty simple, keep it as is, even if it's a little bit brittle, and just like you know stick it in there.

30:34.38
Sean
Nice. Nice.

30:39.79
Sean
nice

30:42.36
Andrew
And then we're debating like whether you're to take the time to build a credit system now.

30:42.46
Sean
nice

30:47.98
Andrew
And I think we'll probably do like end up doing a basic credit system where it's like... like we do the math to like calculate how many pages you've crawled, how many generations you've run, but we don't like cut you off if you go over unless you're on like the free tier.

31:05.21
Sean
Nice.

31:06.47
Andrew
And then we got to decide if we want to do a free tier, not like a free, me not like a true free tier, but like we give you a hundred credits for free or we give you 500 credits for free or something.

31:17.73
Andrew
So you can like, you know, it's a, a usage based free trial instead of a time based free trial.

31:19.28
Sean
Cool. Dude, I saw I was trying.

31:25.19
Sean
I was trying an AI icon tool the other day, like an AI motion icon tool.

31:28.31
Andrew
Yeah.

31:31.77
Sean
And it was a really interesting thing of like how to get free credits and you would have to like share this on Twitter and share this on LinkedIn and then stay on this gallery page for 30 seconds and you get like however many free credits.

31:44.35
Sean
but thought was kind of interesting as a concept.

31:46.71
Andrew
yeah

31:46.85
Sean
And I almost feel like gay growth hacky micro SAS version of that that you sell to other ai tools could be kind of interesting.

31:54.56
Andrew
Yeah, it's the kind of thing that I feel like matters more for low-ticket consumer products where you you need you know by a rule you need to create viral loops in order to survive.

32:02.31
Sean
Definitely. Yeah.

32:08.40
Sean
Yeah.

32:08.72
Andrew
And even then, I feel like forcing that stuff usually doesn't work that well in the long run. it like You're juicing growth, but if there's not like real concrete reasons for growth behind it, for viral loops like behind it, then it's not super sustainable.

32:20.66
Sean
sure. That's cool.

32:23.21
Sean
but sure

32:23.99
Andrew
i I have seen people do it. i think, oh yeah, one of my clients right now is doing it. If you complete their onboarding, they give you like 50 credits for free.

32:34.70
Andrew
So you can, if you like, you know, they're trying to encourage you to give you a little incentive to activate, which is cool.

32:36.99
Sean
it's cool

32:40.78
Sean
yeah

32:42.33
Andrew
So who knows? Maybe we'll try it at some point.

32:45.53
Andrew
I should ask them if if it moved the needle for them.

32:45.58
Sean
cool

32:47.93
Andrew
Like if it, like what the results were of implementing that.

32:49.44
Sean
Yeah.

32:52.12
Sean
I mean, Bolt did it with like, one thing that Bolt did that was interesting was like, tell us who you are and we'll give you a million of credits in Bolt, which is like enough for me to, I mean, it was a so that's a substantial amount that gets you pretty far into an app.

32:58.97
Andrew
Yeah.

33:07.39
Sean
Even, yeah.

33:07.45
Andrew
That's cool.

33:11.63
Sean
Actually, I don't know. I don't know how many. I haven't used Bold in so long compared to like Lovable. It was nice. It was like an it was like an easy, like like, tell us about yourself, who you are, what company you work for, stuff like that.

33:16.91
Andrew
i

33:23.98
Sean
I feel like that's like and nice, easy hoop to jump through, but also gives you intel into, you know.

33:28.89
Andrew
That is pretty nice. I do i do like that, actually. That would actually be even interesting if if all you did was that as your micro sass.

33:31.98
Sean
Yeah.

33:36.59
Sean
Yeah.

33:37.87
Andrew
It's like just a, you know, answer a question and get credits or something.

33:38.80
Sean
Mmm. Yeah.

33:44.18
Sean
Wait, wait, wait. Quick, quick. Pick an animal.

33:47.27
Andrew
Jaguar.

33:47.97
Sean
Okay, credit jaguar.

33:50.54
Andrew
We can probably do better than that.

33:51.83
Sean
Okay.

33:52.77
Andrew
Credit. Credit.

33:53.82
Sean
Chicken.

33:54.92
Andrew
Credit chicken. Credit. Crow. yeah

34:00.07
Sean
Credit crow is kind of... Credit crow kind of sounds like a like ah Equifax kind of thing.

34:06.60
Andrew
Credit panda. it's There's too many pandas already.

34:09.98
Sean
There are a lot of pandas.

34:10.98
Andrew
Feedback panda.

34:11.23
Sean
What pandas are they? Oh.

34:12.29
Andrew
There's like feedback panda. Like there's like some like teacher panda thing. Like isn't Arvid Kahl's original thing? ah Wasn't it a panda something?

34:21.06
Sean
I don't know. Podcast panda.

34:23.84
Andrew
podcast panda is pretty good pretty good still i still own link ferret dot com i don't want to give it up but it does definitely does not fit metamonster anymore

34:26.78
Sean
Podcast panda is pretty good.

34:30.06
Sean
Mm-hmm.

34:37.12
Sean
It's okay. Launch another one. Do another thing. You got cloud code now. Yeah. Apparently it's fast as fuck.

34:44.97
Andrew
that's funny that can be fun

34:45.09
Sean
Yeah. You should, you should, you should, you should add the internal linking template because like all of our clients are like, Oh, I wish we had internal links more on our blogs.

34:53.21
Andrew
So that's like one of the only... I'm going to test some prompts to like try to to do it. But internal linking is actually like one of the hardest things to figure out the UI for.

34:59.57
Sean
Hmm.

35:04.18
Andrew
it's like It's one of the only things that doesn't fit well.

35:07.28
Sean
Yeah.

35:07.56
Andrew
course, the two things that people want the most, image alt text and internal linking, like don't fit super well into this new UI. like There's like kind of hacky workarounds that I can do for right now.

35:18.50
Sean
yeah

35:19.69
Andrew
But like one of the big UX problems we haven't solved yet is like what do you do when issues are kind of like one page to many issues, issues like or one page to many elements? like H2s are another one.

35:35.94
Andrew
There can be yeah as many H2s on a page as you want.

35:36.25
Sean
Yeah. Yeah.

35:40.36
Sean
yeah

35:40.56
Andrew
H1s are a little bit easier because in theory, there should only be one. If there's not if there's more than one, it's a problem. And so you extract the first one. tell the user if there's more than one, and otherwise you just go off of that one.

35:55.10
Andrew
But then like H2s, how do you display that in a grid?

35:55.23
Sean
yeah

35:59.90
Andrew
And so I think we're going to have to have like kind of a page UI where you're looking at an individual page. We may also solve it by doing like having like just an images grid or an images view that they extracts all of the images from your entire site and solve image alt text that way and like have a link back to a page there.

36:12.15
Sean
Mm-hmm. Mm-hmm.

36:23.46
Andrew
The internal linking one is the hardest, though, to do well. we can We can do do it now. We can do it. um And I will probably i'll do some testing of that.

36:37.95
Andrew
But it's just, like it's go to like, it's basically going to be, a block of text in a, you know like, cell.

36:46.81
Sean
Yeah. Yeah.

36:48.76
Andrew
And you've got to like, kind of read that text and work with it versus...

36:52.45
Sean
Yeah. yeah It does kind of, it does kind of start to break the, the whole like table concept apart.

36:59.49
Andrew
Yeah. And we don't see the table as the end-all be-all UI here, right?

37:03.46
Sean
Yeah.

37:03.84
Andrew
Like it's the starting point.

37:04.27
Sean
Yeah. yeah

37:06.61
Andrew
we We want to give you multiple views of your data and like ways to structure them and work with them.

37:13.31
Sean
Yeah.

37:14.98
Andrew
Like we probably will have, you know, click into a page and then look at like the elements on that page and...

37:23.30
Sean
I mean, I think that makes sense. I think like with internal linking, i think the thing with internal linking is that there probably is still a level of like human in the loop there.

37:33.81
Sean
So clicking it into each one to review still probably makes, i don't know.

37:38.16
Andrew
Yeah.

37:38.30
Sean
It's hard to say.

37:39.22
Andrew
You really, with internal linking, you really kind of want like almost a graph view too, or like a flow chart view, because you want to be able to see like how the links, like see how the links move visually.

37:43.73
Sean
Yeah. Yeah.

37:46.23
Sean
Yeah.

37:52.99
Sean
Yeah, you care you care which one is your pillar page or your hub page for sure.

37:55.77
Andrew
Yeah.

37:56.90
Sean
so Okay, good luck with that.

38:00.52
Andrew
Yeah, it's a thorny, it's a really thorny UX problem that we don't have a solution for yet.

38:00.71
Sean
Sounds like a pain to me. Yeah.

38:06.87
Sean
Yeah.

38:07.81
Andrew
So I will have like a way to generate internal links, ah ah like a crude way to do it pretty soon. And then, but we are actively thinking about how do we do it better long-term.

38:19.92
Andrew
The other thing we have to do is we have to give the the agent that runs at the page level, the ability to access the full page, like the full grid context. now it only has row by row context and we need to give it the ability to search the entire grid to do internal links.

38:38.74
Andrew
Well, if you want to see links from other pages to your page, like we can include links, the internal links that are in your existing content, but yeah.

38:42.80
Sean
Yeah, yeah.

38:49.97
Sean
Yeah, out to other pages. Yeah. Hmm.

38:54.92
Andrew
This is the other thing. Like, you care about links in your existing pages out to other pages. You care about links coming into your existing page. And, like, and then ideally, like, you want to have, like, sort of more links going.

39:08.27
Sean
Yeah, i um I almost want like a sitemap.

39:10.64
Andrew
Yeah, you want a sitemap. You want, like, a flowchart kind of.

39:11.85
Sean
Yeah. Yeah. Yeah. Yeah, like a blog-specific sitemap. That just sounds painful, you know?

39:24.60
Sean
Like, when we import something into, like, our sitemap tool, like, an existing site, and it's a site that's been around for a while, like, it just generates giant amount of, like, one single row of a bunch of these, page like, boxes of pages, and it's like, oh, this is this is untenable. I just, like, I don't need this, first of all, but, like,

39:45.12
Sean
yeah because you can't, like, how how would you figure out the the, like, levels of it?

39:51.50
Andrew
I went to look back at LinkScout to see how they do it.

39:52.40
Sean
Mm-hmm.

39:54.33
Andrew
And it's pretty crude. Like each page, you can click into the page and the page has like a keyword associated with it.

39:57.75
Sean
Yeah.

40:02.77
Andrew
And then it has links coming to that page and links going out from that page. but it just automatically generates them all.

40:07.77
Sean
Mm-hmm.

40:09.46
Andrew
And so there's no way to see like what's on your site now versus what's not on your site. And like to edit individual links and stuff, it's pretty clunky actually.

40:21.65
Andrew
So I think we can do it better.

40:22.98
Sean
Yeah.

40:26.17
Sean
Cool. Cool. Sounds fun.

40:28.58
Andrew
the

40:29.48
Sean
Sounds exciting.

40:30.44
Andrew
Yeah.

40:30.54
Sean
wish I was working on a star with my my friend.

40:33.81
Andrew
Aren't you working on a startup with,

40:36.67
Sean
Yeah, i guess.

40:36.88
Andrew
your coworkers, I guess.

40:39.27
Sean
No, they're my friends. They're my friends. ah there I meant, you know, like...

40:48.86
Sean
indie hacking it and not I wish I didn't have client fires all the time but yeah yeah you're welcome how's your uh I gotta go in a second how does how's your consulting thing going

40:52.64
Andrew
Yeah. Well, thanks to you, I also have client buyers now sometimes. So

41:02.39
Andrew
yeah.

41:07.78
Andrew
Good. i'm Luckily, I have a lot of leeway, and so I'm taking my time on discovery. and just like So i'm I'm spending this week basically just talking to members of their team to ask them to get their perspective.

41:15.57
Sean
okay

41:22.63
Sean
Mm-hmm. Mm-hmm. Mm-hmm.

41:23.43
Andrew
The downside is like what I'm supposed to do isn't crystal clear. The founder kind of just hired me to be like,

41:34.34
Andrew
hey, help us find product market fit. Like, help us six fix things. Like, connect the dots between product and growth and stuff. And I had a bunch of, like, concrete project ideas.

41:46.32
Andrew
And then I started talking to their head of growth. And he's super fucking smart and is already, like...

41:52.09
Sean
Nice.

41:52.36
Andrew
actively working on a lot of those. And I'm like, okay, well, i you seem smart. And I will like you see if we can collaborate on some of this stuff for sure.

42:04.76
Andrew
But don't know that you need me to just run with these things.

42:04.84
Sean
Right.

42:08.40
Sean
Right.

42:08.90
Andrew
Maybe at some point I can like take a project off his hands or something. But yeah and so i was like okay um so i think my job is gonna shift a little bit i think it's gonna be a little less like concrete projects and more like helping them build process and build a product culture that is focused on like outcomes over outputs which is hard

42:34.68
Andrew
so that'll be interesting Like, I think I'm basically going I think there's a chance I end up becoming, like, kind of a fractional chief product officer or head of product or whatever.

42:49.19
Andrew
And not just, like, do some do some positioning sprints for them.

42:51.76
Sean
growth

42:55.50
Sean
makes sense yeah makes sense is that

42:59.74
Andrew
So...

43:02.02
Sean
Is that what you want to do?

43:04.34
Andrew
Oh, it's definitely, it's it's fun it's fun and interesting work. It's just like, I'm still trying to figure out where I can deliver the most impact to the team.

43:08.22
Sean
Okay.

43:13.02
Andrew
And it's like weird to have a consulting project where they're not like, do this one thing.

43:13.40
Sean
Okay.

43:17.36
Andrew
They're like, help us figure out what we need.

43:20.23
Sean
It is, for what it's worth, that is more true to like the consulting sense of your consulting stuff.

43:24.79
Andrew
Yeah. I'm like an actual consultant now and not just, yeah.

43:28.56
Sean
Yeah, yeah, or like an IC sort of person.

43:31.00
Andrew
Yeah.

43:31.61
Sean
So cool. Very cool. Yeah, that'd be good.

43:35.77
Sean
I wonder i wonder what what ah this this will look like like a month from now. So cool.

43:39.25
Andrew
Me too. But the good news is their head of product or head of growth, super fucking sharp.

43:45.35
Sean
Yeah.

43:46.19
Andrew
He's worked at like, he's also worked at like segment Twilio stat sig.

43:49.42
Sean
Oh, shit.

43:51.05
Andrew
He was a solutions engineer or like head of solutions engineering for a lot of those folks.

43:54.62
Sean
Got it.

43:55.49
Andrew
And so has been on the like go to market side of things and has also had a growth role in the past.

43:55.62
Sean
Got it.

44:01.62
Andrew
And he seems really thoughtful and like he's approaching things in a good way. So that's cool.

44:09.56
Sean
Cool.

44:09.93
Andrew
I'm excited to get to know him more and learn from him and work with him.

44:13.70
Sean
Cool. Sick.

44:14.54
Andrew
Yeah. Cool. All right.

44:15.86
Sean
Okay.

44:16.08
Andrew
You got to run. All right.

44:17.12
Sean
Yeah, going I'm to call. Bye. I'll see you later.

44:19.25
Andrew
Peace.

44:19.97
Sean
Bye.