Latest / 1 to 100 / E53: GaaS for SaaS with Ronak Shah and Ankit Patel

Transcript

E53: GaaS for SaaS with Ronak Shah and Ankit Patel · 1 to 100

0:00
Are there any like stand out best performing SAS brands that
0:04
you're seeing right now, I guess for SAS?
0:06
Yeah. I mean I think I think there is
0:08
probably a few. I think for us, like, I think
0:13
rocked is just like, I mean, has been our top client has been
0:18
like our best customer, but like everything they do with just
0:22
high level of intention, they understand like the pay it
0:25
forward model. They understand that, you know,
0:28
how to activate partnerships. They understand agency
0:31
partnerships, they understand, you know, how to go after the
0:34
brands you need. You know, we've been working on
0:36
this cool thing tool called D2 CMVP as well.
0:39
They've been building with Sean and you know, they even with
0:41
that, they're even like really, really good with just
0:45
understanding that feedback doesn't mean conversion.
0:48
So I really like them, but also just a very mature team.
0:52
And then I think like on the bigger size, we've been working
0:54
with Apple up and now for two years and we're, you know, doing
0:58
a lot of production with them and they've been really
1:02
interesting because it's like you have like a public company
1:07
that like actually works like a start up.
1:11
So it's like really, really fun working with them.
1:13
And I think they also really care, but they're definitely
1:17
like learning as they go. Like they don't understand the
1:20
models. They don't know you have to do
1:21
like these different, you know, events and partnerships use like
1:25
they're learning as they go and they're figuring it out.
1:28
I mean, the beauty is they do have an unlimited bankroll.
1:30
So that helps. I think just on the on the flip
1:34
side, a few others that I think we've enjoyed working with, but
1:40
more so even like just that I think I've I've done well is
1:43
Proxima. You know, Proxima is in a
1:45
really, really good job kind of taking the e-commerce space and
1:49
then like levelling up to now like partnering with Microsoft
1:53
and going enterprise. And I think it's been really
1:55
cool to see them like almost graduate beyond e-commerce be
1:59
beyond like SMBs and kind of their initial ICP.
2:03
So that that was really fun. But I think overall, I would
2:05
just say there is there's a lot of this noise around AI and what
2:10
it's doing and everyone has all the tools and capabilities.
2:15
And although I thought it would make certain SAS companies feel
2:18
like we got it, we got to figure it out, we got to pivot, we got
2:22
to create something new. It's at least from my angle, I
2:26
see it, people are very like calm.
2:28
And it's like maybe because like most of these companies are
2:30
built on true engineers and like, they probably know like,
2:33
all right, some of this 90% of what you're seeing out there
2:36
isn't even going to be effective or won't actually make it
2:39
through or there's going to be too many issues.
2:41
So there's this like sense of calm in how certain people are
2:44
building, at least from our peripherals.
2:47
And it's kind of cool because like on the other side, you have
2:49
brand owners are like, I got to learn AI.
2:51
I got to build landing pages. I got to spit out 50 creatives a
2:54
week. Now I got to do this, I had to
2:56
do that. And I feel like we're on a
2:57
rampage, like utilize everything.
3:00
But yeah, those are some of our, our lease, our our viewpoint
3:03
right now. Do you think that's like good
3:09
that you're seeing SAS companies become but brand owners not be?
3:13
I don't know. I, I, I don't really understand
3:18
what, what we should all be feeling.
3:21
Also like I, I just because we have access doesn't mean it
3:25
does, it doesn't mean it's going to work, right?
3:27
Like I, I, I talked about a lot about these people talking about
3:30
these static ads, like, oh, you can pump out now 100 ads in 5
3:34
minutes. But like it, what if it doesn't
3:37
work? Like just because you can do it
3:40
doesn't mean it's going to work. And I think like if you take
3:43
that theme and you apply that in multiple ways, it's like, wait,
3:48
like that's true. Like just because we all can
3:50
build landing pages, what if it doesn't convert as high?
3:53
Now what, you know, I just, so I, I don't know, I, I feel like
3:57
maybe the SAS companies have a little bit more just baseline of
4:01
understanding a lot more data that other people don't see.
4:04
And they can probably, you know, look at the back end and say,
4:07
all right, well, things are moving this way versus that.
4:09
So it was just a better bird's eye view.
4:12
I one thing that's interesting, all the companies you just
4:14
mentioned, none of those are widgets as I would look at like
4:19
a, there's like different types of SAS, right?
4:21
But like you just described an ad platform to ad platforms
4:25
basically and a a company that's a what ad audience platform?
4:31
Basically, yeah, it's a data product.
4:33
Well, what about? What about let's say you take
4:35
another tool like Video wise? Would you?
4:38
Yes, that's a. That's a good example.
4:39
So like for, do you feel like they fall into the same
4:42
category? Because I think I actually
4:44
believe. Growing, they're growing like
4:47
crazy this year and it's like in my mind, I've I've even asked
4:51
Harry who moved from chew on this to there.
4:53
I'm like, can somebody not code the ability to show video on
4:58
their website? Like is it, are we that far away
5:01
from it? And, you know, I think the, the
5:03
his piece was he's like the amount of data we have to tell
5:07
you what is actually going to work out of the gate is on, is
5:10
on like, like, and, and how we're going to, you know, set up
5:15
things and make sure it doesn't bug your site, make sure it
5:17
doesn't slow down your site. All these things that I'm like,
5:21
OK, that's fair. Like we can't just cause again,
5:23
just 'cause you can build it doesn't mean it's going to work
5:25
the way you intended it to. So I don't know if that's, I
5:29
don't know if that's going to change soon enough.
5:32
Video wise is an interesting one.
5:34
Yeah, also Jeremiah, like hosting and serving videos, as
5:37
you know, is like a giant pain. I mean, like, like, look, maybe
5:43
we're just going to find ourselves doing like, oh, but
5:45
this one is lucky for this reason.
5:47
And this one is lucky. Like maybe maybe that's what
5:49
we're going to find ourselves doing.
5:51
But dude, like nobody's going to yeah, no ones going to host and
5:56
serve like sorry vibe code host and serving videos because they
6:00
would just get a, they would get a giant AWS bill in like 3
6:04
months. Well, that's what you can do
6:06
that then to your point, you get the giant AWS bill.
6:08
I, I, I think what I'm curious about it was something like that
6:11
is, are we going to see a well, OK, I might be wrong about this.
6:16
And Ron, I actually would love your, your thought on this
6:17
because I mean, you've got a lot of interesting perspective too.
6:20
I think what AI is going to do is kind of put a ceiling on how
6:23
much people are willing to pay. So there's always going to be a
6:25
price that the, the out-of-the-box software is worth
6:28
because of the fact that you don't have to deal with it.
6:31
It's just done on its own. Like this example, right?
6:35
Like is there you could go vibe code video, a similar product to
6:41
put on your website, but there's a cost of doing that, that
6:44
you've got the maintenance, you got the actual like cost of
6:46
credits to do it, you've got the hosting cost, all those kinds of
6:49
things. And that whatever that cost is,
6:52
you're ultimately going to compare that against the
6:54
software subscription fee. And so if that software
6:56
subscription went to $100,000 a month, you're probably going to
6:59
buy code it because it's cheaper slash you get exactly what you
7:02
want out of this thing that you built yourself.
7:07
But if it's $100 a month, there's no way you're going to
7:09
buy Coda. It's not worth it, right?
7:10
So there's, and the, the in between is the unknown right
7:14
now, But I kind of believe there's like there's some number
7:17
there where AI is almost putting like a ceiling on what people
7:19
are going to be willing to pay for some of these things.
7:22
But I I don't know if that's actually the case or if that's
7:25
just like a point how things look early on.
7:29
Video Wise is a really interesting one because I'm a
7:31
customer of Video Wise and Claudia as a friend and he
7:34
actually made Video Wise free for me which I was really happy
7:37
with because spending like 1000 bucks a month on video Wise.
7:39
And I only came over to Video Wise because I rage Chen from
7:43
Tolstoy because I was paying $1000 a month in usage fees.
7:46
And I didn't realize I was like, fuck, this is expensive for like
7:49
showing video on my website. But I was just telling you guys
7:51
about how I've just vibe coded. Vibe coded my first Shopify
7:55
landing page that I'm actually happy with.
7:58
And look at one of the things it did using a video wise landing
8:01
page that I had as a reference. It went and built and this is
8:05
all in Shopify sections. It's not just like raw code.
8:08
It's actually used computer users and like the agent to use
8:12
a browser and built it into Shopify and then built its own
8:17
like custom widgets in the app. So these are like shoppable
8:20
widgets for different videos using like a Video Wise widget.
8:25
But Video Wise is so much more than this.
8:27
This is just one thing that I was using Video Wise for.
8:30
And yeah, it did that. I did another one, which I
8:32
thought was pretty cool. Part of my prompt was I want a
8:35
coupon code. Look at this coupon code and
8:39
then I like redirects to check out.
8:40
I think how cool is that with the discount?
8:44
And Jack, just to clarify, the video that you have there is not
8:48
video wise. It's completely, yeah.
8:51
Yeah, it's not video wise. Yeah.
8:52
So it took like a reference landing page that I had built.
8:55
That took me some time, right? Like I'm clicking around, adding
8:57
in the widgets, uploading the videos, right.
9:00
And it then it just scraped all the videos and built a new one
9:03
as like a custom section in the Shopify theme.
9:05
And I didn't ask it to, it just did it dude.
9:08
That is sick. And Shopify is covering the
9:11
hosting on that I assume. I'm into Shopify now.
9:15
You're not actually paying the hosting cost on those videos
9:17
either. Yeah, one of the things I'd like
9:19
to do, I'd like if it could embed the videos from YouTube,
9:23
like do an extra interim step so they're not actually hosted on
9:26
Shopify, they're on YouTube and they're like racking up plays on
9:29
YouTube. But but yeah, this is a nice
9:32
little midway I. Love this rod.
9:35
I don't know man, maybe you got to tell your SAS friends to
9:38
sweat some bullets. Yeah, I got, I get, I get Harry
9:44
on the call. You just got to be like, yo,
9:48
watch the pod, look at what Jack's doing.
9:50
You guys are too chill. What are you doing?
9:52
Yeah, this is a that was very nonchalant by the way.
9:56
Like just showed us a whole like a SAS built out already.
10:01
That's excellent. Well done.
10:02
No, it's, I mean, it's, it's sick, but it's not 10% of what
10:06
video wise does. Video wise does so, so much
10:09
more. And they're doing this like new
10:10
clipping thing where they like make ads from you based on your
10:13
content, which is sick. So yeah, I feel like people are
10:16
going to lose that detail and that attention to detail from
10:20
like engineers who are building solutions if they just go and
10:24
vibe code some replacement that's buggy and needs
10:26
maintenance, right? Or whatever happens to be right.
10:30
As somebody with A7 figure AWS bill, like I can say that the
10:35
margins matter on this kind of stuff.
10:37
Like it actually does add up over time.
10:39
And you know, if you've got 100 visitors, not a big deal.
10:43
When you've got a million visitors, it's a different
10:44
story. So I do think there's like
10:46
there's a lot of opportunity in those margins on these kinds of
10:49
things. Yeah dude, our Cloudflare bill
10:51
alone last month was like 70 K it's.
10:57
The same. So I just think people don't
11:00
understand what it means to serve video, that's all.
11:04
That's what I'll say about. Yeah, dude, I'm good.
11:08
I got to know, man. Like, you come across like, at
11:11
least my knowledge of you is like, you're like a you're like
11:15
a designer at heart. Yeah.
11:17
Do you feel like you've been unleashed with all of these
11:19
tools? Like most of the other designers
11:22
who I know are like, Oh my God, they're like kids in candy
11:26
stores and that. No, it's, I think there was
11:30
always a, there's a point where, you know, you're either a
11:32
packaging designer or you're a UIUX designer or you created
11:37
graphics for ads and there was all these like little niches
11:40
within the design world. And I was always somebody who
11:42
did a little bit of everything. Like I like to get my hands
11:45
dirty and everything, but it was hard to find people who did
11:49
everything. And it was, it was always a
11:51
software skill issue, you know, because you had, you know, Adobe
11:55
Photoshop did this one thing, you had Illustrator that did
11:58
this other thing, you had Indesign that is so in the
12:00
Epigma and whatever else. But it was always tough to find
12:04
somebody who just went and learned every single thing.
12:07
And what's recently happening, especially with the the design,
12:11
the code aspect now in the handoff being so seamless, it's
12:15
changed the whole game where you can, you can be a packaging
12:19
designer, you can create the ads, you can create the the
12:22
branding systems, you can create every single thing.
12:26
And you don't need 5 different designers anymore.
12:29
You just need that one guy who knows how to use every single AI
12:32
platform. And honestly, most of these
12:34
things are all within a suite and I can just kind of one shot
12:38
everything. So for me, it's been amazing.
12:41
But then there's also the other end of it where a lot of these
12:44
people who are who have specialized in this one, or I'm
12:48
just a packaging designer and I've never done anything else.
12:51
OK, well that can't fly anymore. Those people are getting scared.
12:55
So I think there's 2 ends of the spectrum where there's there's
12:58
half the designers are like, OK, I'm unleashed.
13:00
I can do every everything that I wasn't able to do before.
13:03
And then there's the other half that's like, shit, I don't even
13:06
want to go and learn how to go do everything else.
13:11
But yeah, that's that's kind of where I'm at.
13:14
So if I connect that back to like this, like, yeah, how
13:19
should like SAS people feel? And how's all of this shifting?
13:24
Like what you're describing is a world in which let's just take
13:28
some like larger company has like 20 designers on staff using
13:32
10 different tools. And now you're saying actually
13:35
it can be two designers using one tool.
13:39
I'm, I'm making it up, right. But just to, just to be
13:43
provocative, the, I think the average brand or retailer is
13:48
going to like choose that direction, which like, it makes
13:55
it feel like every software company must serve multiple
13:59
purposes. Because like what you just said
14:01
as a user of software is like, I kind of have this experience now
14:05
where this piece of software can do a lot.
14:08
And So what it does is it imposes the pressure on other
14:10
pieces of software to also be kind of full suite, right.
14:14
And if you're not full suite, then what are you doing?
14:17
You know, we're in like the greatest era of full suite
14:20
platforms. And, and the second thing it
14:22
does is it puts pressure on this idea that you're not going to be
14:28
able to sell it to 20 people. You have to sell it to two
14:30
people and make them like have Iron Man suits, for lack of a
14:33
better phrase. Do you, do you feel like the,
14:37
the sort of total change that is going to happen as a consequence
14:41
of what you just described as your own experience?
14:44
Where, where in that cycle are we from your perspective?
14:48
And like, who's like, you know, Ron was jokingly saying like,
14:52
you know, all the brand owners are like everywhere.
14:55
But I actually feel like the reason they're everywhere is
14:57
like, because there's the, they're at the tip of the spear
15:01
of like how this change is going to happen is, is that, is that
15:04
how you would feel too? Or how would you describe like
15:06
the, the, the, the rate of change that's going to happen
15:09
across the industry and like who's leading first?
15:12
Yeah, I think it's already happening.
15:15
If you were to talk to me about this about six months ago, I'd
15:18
say, oh, it's AI is great, but it can't do this, this, this,
15:22
this and this. Well, six months later, we've
15:26
gotten to a point where you have no, you have no excuse but to
15:31
adapt to it. And if you don't adapt to it
15:33
right now, like we are at that, that pinnacle where if you don't
15:38
adapt right now, you are actually going to get left
15:41
behind. You have to change all of your
15:44
systems, how you are working. We all, I'm sure most designers
15:50
went to school and they were taught a certain way of doing
15:53
things. Hey, you, you know, you, you
15:57
create this thing in Figma and then you hand it off to code.
15:59
And, and that's, that was kind of the, there's, you know,
16:02
certain protocols that you had to follow.
16:05
Well, now people are just looking for designers who can
16:07
ship. So these and it's a popular
16:11
Twitter term. Where?
16:14
You, that designer, has to be able to go and 1st create the
16:17
product. Like let's say we're just
16:18
talking about a tangible product, then go and create the
16:22
LP's for it. Go and create the the visual
16:24
identity. Go and create.
16:28
Honestly, these days designers are more looked upon to create
16:32
ad copy and things like that as well.
16:34
Like you don't need something like they should know the
16:36
product in and out. That person needs to be the, you
16:40
know, people. People always used to say too
16:42
many chefs in the kitchen. That person needs to be the chef
16:44
and really own everything. And that's why you see a lot of
16:49
CMOS now who weren't traditionally trained in design
16:53
now becoming those people because they already had the
16:56
mindset. They already knew the product,
16:58
they already knew how to sell it.
16:59
The only thing they couldn't do was design it.
17:03
Now they can design it and they can build a landing page and,
17:06
and Claude or Manus or whatever. And you know, they can, they can
17:11
pretty much design and ship and they can build 100 static ads
17:14
like Ron said. Now the only problem I see with
17:17
this is we don't know the incremental value of of that.
17:23
You can go and create a hundred static ads and we've been
17:26
testing this on Avi right now and you can go and create a
17:28
bunch of landing pages. But honestly, Ron, we haven't
17:33
seen much of A lift versus when I was doing it by hand and I was
17:40
creating 15. So it's it's one of those things
17:45
where, yeah, you can get a lot of output, but now the quality
17:51
comes into into discussion because something is missing.
17:57
And even though it looks all that all the same, something's
17:59
missing. So code is obviously the domain
18:03
in which AI has moved the fastest, right?
18:07
And I think it's, it is because it has the best feedback loop.
18:12
And so like these coding agents can like write the test
18:17
themselves. They can understand what to ask
18:19
the user for. There's, there's like all of
18:22
these sorts of reasons why coding has the best feedback
18:25
loops. Don't you feel like we're like 6
18:29
months away from what you just described, having the right
18:33
feedback loops like this landing page that you build?
18:37
Like, man, there's just no way. You're like, I can't imagine
18:42
that in Christmas if you and I were chatting that that feedback
18:45
loop problem has not been solved for, for like landing pages or
18:52
ads or whatever it may be. It feels like we're at the
18:56
precipice of solving the feedback loop problem in tons of
19:01
domains over the next 6 to 12 months.
19:03
And maybe some come earlier than others.
19:05
Like clearly coding happened first.
19:08
Clearly legal is happening really fast, but man I got to
19:13
believe that the that it'll get solved fast.
19:17
I agree with you just because of how fast we've come here in the
19:20
last six months, coding happened first, but then you look at
19:25
something like, you know, people are looking for feedback on Clod
19:28
for their their meta ad accounts.
19:31
And if you, if you look at the feedback you get on Clod, it's
19:34
terribly outdated. It's like it doesn't work.
19:37
It's literally just, it's doing what ChatGPT used to do, and
19:41
we're just kind of given you positive answers and things like
19:44
that. But like you said, I think that
19:46
gets solved relatively quickly. When does that happen?
19:50
Yeah, within the next six months.
19:52
I could. I could see that getting solved.
19:54
But then when everybody has the answer, then what happens?
20:03
Dude, boom time. It's boom time.
20:07
The greatest boom since, I don't even know.
20:12
Yeah, I think like, you know, like, yeah, we will be, we will
20:19
be restrained only by the rate at which the labor market can
20:23
keep up in terms of GDP growth rate.
20:26
Is is is is like what is possible right?
20:31
Now I'm, I hope so, I mean. The reason why we're seeing like
20:38
founders get so stuck in where maybe we're not seeing other
20:42
people, I don't know, maybe they're more comfortable.
20:46
They're just not founders is because, you know, founders, you
20:49
know, they have their wide range of different motivations.
20:52
Maybe they're scared, right, or they're excited about the new
20:55
tech or they just want to kill their competitors, right?
20:58
They just want to, you know, not have those competitors, those
21:01
copycats, whatever happens to be.
21:04
But the high agency people, by like definition, they have like
21:07
a bias to action. They have a high pain tolerance
21:10
or what they're doing doesn't feel painful.
21:14
They just get things done. I mean, I used to work in agency
21:18
consulting agencies. I remember talking to e-mail
21:21
strategists. These people didn't even design
21:23
emails. They had designers.
21:26
They were Clavio people, right? And or, you know, the thing that
21:30
I would talk about is, hey, we have to do more, right?
21:32
We have to, you know, achieve the outcome.
21:34
We have to test new pop ups. We have to write and send SMS.
21:36
We have to, you know, send push notifications, right?
21:39
And we have to stop defining ourselves as I do e-mail and
21:43
start defining ourselves as we, we do retention, we maximize the
21:47
value of the contact information to the brand that we're working
21:50
with. But they didn't want to, it was
21:52
like, I do e-mail. That's my, that's my job title.
21:54
I'm not going to do more than this, right?
21:56
So I wonder if that's one of the reasons why we're seeing
21:59
founders get stuck in and do stuff is because they actually
22:02
do stuff that's in their blood. All right.
22:05
I agree. The guy was saying before some
22:08
people are they want to be specialists.
22:11
They almost take pride in being specialists.
22:14
And if you're founder of a company, you know, you can't be
22:16
a specialist. You kind of have to be able to
22:19
wear many hats and then do a lot of different things.
22:23
So I, I think it's, I think we're in a great time where
22:28
people are becoming or people are learning faster than they
22:31
ever have before. It's it's crazy to to know that
22:36
just just a year ago or two years ago, how less educated
22:40
everybody was on everything. And it's almost like everybody's
22:44
on the same playing field right now.
22:46
You could tell two years ago who really knew their shit and who
22:50
didn't. Now it's like, man, everybody,
22:53
everybody seems smart. That's I don't know if that's
22:57
just because they're getting fed all the answers.
23:00
I don't know if it's a good thing in the long run, but I
23:04
know output wise that's it's a great thing.
23:09
Yeah, You know what? I'm really grateful I haven't
23:12
jumped into a call with someone who's just, like, firing out
23:14
contrast negations. It's not this, it's that, you
23:17
know, they got ChatGPT up. Just feeding them the next thing
23:20
to say, it would be a nightmare. That's funny, you guys.
23:25
See this? I imagine you probably saw
23:27
Rashad, but I was excited to talk to you about this.
23:32
But the thesis is that the next $1 trillion company will sell
23:35
work and not software. The idea being if you sell a
23:38
copilot you're competing with every new model release.
23:41
If you sell an outcome, close books, contracts reviewed,
23:44
claims handled. Every AI improving makes your
23:46
margins better. The entire SAS playbook was
23:50
about capturing the software dollar.
23:51
The AI playbook is about capturing the services dollar at
23:54
software margins. I think this example puts it is
23:58
contrast indications, but hey I think it puts it nicely.
24:01
Not AI for accountants, but the AI accounting firm.
24:04
Not AI for lawyers, but the AI law firm, right?
24:07
They weren't like SAS companies that look like service firms
24:10
rebuilt on software infrastructure.
24:11
What do you guys think about this?
24:14
I mean, I think VCs are going to talk like VCs.
24:17
That's that's for sure. So that's like you, you have to
24:22
like, remember, it's their job to like be pithy and provocative
24:28
to drive engagement and talk their buck.
24:32
Because this is just like not, I mean, I hate to say it, but this
24:35
is just like, not a new idea. Well, that's yeah.
24:38
This is the idea of software, right?
24:40
This is what software? Yeah, exactly.
24:41
For a long time is it. It would replace the things that
24:44
humans had to do by hand. That I agree with the like basic
24:49
premise that the the dollars are in human labor.
24:52
If you replace human labor, people are willing to pay a lot
24:56
for that. But ultimately, like that was
24:58
always software, was always a version of trying to do that as
25:01
well. It's just like a different
25:04
version of it. Like the core problem, said
25:06
differently, the core problem is the art.
25:10
Or there's a temporary argument to be made that if you can dress
25:13
up a software company as a services company, you can ask
25:17
for a higher price point in the short term.
25:21
But do the problem is software companies compete with software
25:23
companies and like we live thankfully in a highly
25:26
capitalistic competitive society.
25:29
And so ultimately these things will just get repriced, right?
25:32
Because at some point someone's going to compete on price.
25:35
And so that's just the beauty of capitalism.
25:39
Now, I, I, I think actually the bigger insight is what it does
25:44
to, not the software industry, like to people who run brands is
25:50
they used to have to pay whatever amount to an agency or
25:54
a service provider. And what they can look forward
25:57
to is that that amount is going to become a quarter of what it
26:01
is today, 1/10 of what it is. And it changes the economics of
26:05
all of those businesses, which is why I made my like boom time
26:08
comment. I, I think that's actually the
26:12
insight. The insight is like, you're
26:14
going to see a bunch of services start to get priced like
26:17
software over some period of time.
26:20
They're not going to price like services.
26:22
There's just no way. And all those dollars that are
26:25
saved are just going to go to the ad platforms.
26:27
That's where all that money is going to go.
26:30
No, You think so? I think so like think about so
26:35
well, let's just use Avi as an example.
26:37
If your cost of running your business got cut in half, where
26:42
do those dollars go? I mean, probably some of that's
26:47
profit, right. But like if you're going to
26:49
reallocate those dollars because your, your OpEx gets cut in
26:55
half, where, where, where do you put those extra OpEx dollars?
27:00
Media, you know. I think it's all, it's all going
27:03
to get spent on on advertising, at least in this ecosystem.
27:08
Maybe that's not true everywhere, but.
27:09
Yeah, I, I just think that if it all gets spent on advertising,
27:12
what it means is that there's enough money going somehow to
27:18
the somehow the consumer has to buy stuff, right, Right.
27:23
Because you're not going to sit here and like, advertise that
27:26
like cocks that don't make any economic sense.
27:29
Like that's not going to happen. That's there's like a macro
27:32
issue there potentially that has to be dealt with first.
27:35
Yeah, you'll see like a temporary hockey stick and then
27:37
it's like, well, where does it all get eaten up?
27:39
Like it can't. And there still has to be the
27:41
the demand and consumption model.
27:44
So, but no, like I, I, I think on the same note of like how I
27:51
don't, I don't know about like this whole like AI accounting
27:54
firm and whatnot. Like, but what I do do think was
27:58
that was pretty relevant was I think it was Khosla, Khosla
28:05
Ventures he had put out. He had done some podcasts or
28:08
some clips for TikTok and one thing he said, which is like
28:11
really interesting, he said, I think we're less than two years
28:13
away from websites being completely irrelevant.
28:18
And I thought that was really powerful where he was basically
28:21
talking about the fact that there is why would if if you're
28:25
a woman, why are you going to in a year from now, even go to aloe
28:30
yoga, Lululemon and a third website to pick leggings?
28:34
Why wouldn't you just go to your agent who knows you probably
28:37
better than you know yourself at this point, right?
28:39
Whatever you're using, whatever tool you're using and say, Hey,
28:42
I need leggings. And it's going to have your
28:44
information and you're going to trust the number one source it
28:48
gives you, whether that's aloe, whether that's lemon or it's
28:50
Walmart, you're going to buy it 'cause that's what you need.
28:53
And then there's of course, there's a layer of like, well, I
28:56
want a certain look and feel. And I think we're going to get
28:59
to a point where like if everyone stops shopping that way
29:04
or stop thinking that way and you start to get into a little
29:07
bit of this, like, I don't know, utilitarian, like this world of
29:11
where like some of these things don't matter.
29:13
And this is how everyone else is creating efficiency and how
29:16
they're doing it. It's like, why would you have a
29:18
website? You would just go and basically
29:21
make sure that like if I need a collagen, why am I going to go
29:25
to vital proteins or gain of the etcetera when there is some
29:30
known AI that is probably smarter than any research I can
29:34
do. So questionnaire, even AI agents
29:38
doing all your shopping, who's the brand actually marketing to?
29:42
Is it the human or is it the AI agent?
29:45
I think it's going to become 5050.
29:46
I, I think it's like exactly, I mean the S, the SEO layer
29:50
turning into Geo and all of that stuff is just going to have to
29:53
accelerate. But the problem is, is, yeah, I
29:56
still need to pick one thing, whereas Google Shopping and all
29:59
these other places can show 10 things.
30:01
And I think that's where the real estate piece is going to be
30:04
very expensive, where I think a placement on a ChatGPT or one of
30:09
these things are going to be like a Super Bowl commercial.
30:12
It's like, hey, you want to be the most recommended leggings?
30:15
Well, it's going to cost $2,000,000 for this month and
30:19
the placement and need to have all these characteristics and
30:22
criteria to make sure you can even be the most recommended
30:25
one. And then out of that, we'll
30:27
create a field of opportunity, and that's what it is.
30:31
That's just like selling to Costco, man.
30:36
Yeah, Ronna, I think that's, I think you're right that a lot of
30:39
shopping will go that way. But I don't think websites die
30:42
because I think people like to browse like this is actually
30:46
like a like, maybe maybe it goes towards marketplaces as opposed
30:51
to individual brand websites. But like you go to the farmers
30:53
market because you want to browse, not because you're
30:55
actually like, that's the most efficient way to buy your your
30:59
fruit and vegetables, right? It's like that that desired
31:02
shopping malls at the same way, right?
31:03
Like most retail still happens in a an environment where
31:08
somebody can browse and shop as opposed to like online is more
31:12
efficient shopping than buying in retail.
31:15
But yet most shopping still happens in retail.
31:17
So I don't think it goes completely to that.
31:20
How do you think that I like I think you're right about like
31:22
that Super Bowl idea, like there's going to be like a
31:24
massive premium to be the number one recommended thing in the AI.
31:28
But I still don't think that the idea of like online browsing
31:31
goes away. When I watch the way my wife
31:32
shops, it's not the way I shop and it is very browsing heavy.
31:37
Like there's a lot of browsing there.
31:39
And that is, that is actually the desired behavior for her,
31:42
not the utilitarian get the exact thing that I need.
31:46
Yeah. Can I just one provocation,
31:48
Jeremiah, because I first of all, I agree with you, but I
31:51
have this dissonance because in B to B world, actually it's
31:56
really easy to see. I don't buy tools that don't
32:00
have an MCP server or CLI or like a plus API.
32:04
And the reason is I don't want to look at the website like I
32:08
have the inverse. So like there's a tool that we
32:11
use that I have never seen the interface of because all I've
32:15
ever done is connect through the MCPII literally could not tell
32:18
you what the interface looks like and I interact with it
32:20
entirely through the MCP. Salesforce just announced that
32:23
is going, they're marketing it as headless, but the reason
32:27
they're marketing it that way is they've invested a lot of time
32:30
over 20 years being very strong at the API layer, which is what
32:35
allows for that whole like industry of Salesforce
32:37
implementers. But anyway, the, the, the idea
32:41
that you'd never have to visit the website in B2B is really
32:45
easy to understand, like ridiculously easy.
32:49
And so if it's so easy to understand in B2B, yes, I
32:53
understand what you're saying for shopping, but at some point
32:56
you got to ask yourself, wait a second, there are categories of
33:00
things where I don't want to see the website and like how far
33:05
does that extend? Totally what Ron just described
33:08
collagen and is a good example, right?
33:10
Like why? Why would you need to see the
33:12
website? You're just looking for the
33:14
thing that's the best fit for your specific need.
33:16
So, but if you're looking for, I'm browsing dresses and I'm
33:21
trying to figure out which dress I want to wear to this, you
33:24
know, this ball that I'm going to in two weeks, you might want
33:27
to look at 500 dresses. Right, right.
33:30
Experience. There's an experience component,
33:32
right? We're talking because we're all
33:33
male. We're all male.
33:35
So we're thinking problem solution, right?
33:37
But females aren't, well, sometimes they are.
33:40
My girlfriend right now, she's a Gemini maxi.
33:42
Everything is organic, you know, natural rubber, whatever it
33:45
happens to be. So she has problem that she's
33:47
meeting with solution. But like, you know, at some part
33:50
of it's the experience. I like going to the farmers
33:52
market as well. That's the only time I'm ever
33:53
going to do my groceries is if I'm going to a place where it's,
33:56
it's nice to do it, right. It's an experience.
33:58
So yeah. But Jeremiah, can I provoke you
34:00
a little? Yeah, totally.
34:02
I'm I'm not saying I'm right about this and I I think I'm
34:05
just, I just don't think they go all the way lol as my my belief.
34:12
Yeah, but here's here's the problem with the your dress
34:14
example. What's better than browsing like
34:17
for brands websites, It's I have an MCP connected to every one of
34:21
those brands websites and I pull in all of the images.
34:25
Then your wife takes 20 photos of herself and she says impose
34:29
the dress onto this image of myself.
34:31
I want you to render it as a 3D render.
34:34
And then now I want you to build an application for me that takes
34:37
the dresses from every brand that I love, renders it onto my
34:41
body. And I just slipped through it.
34:43
And she swivels and she looks at herself and she says, now make a
34:45
little bit of a cart and do a little bit of a modeling show.
34:49
And now I'm going to text it to Jeremiah.
34:50
So that way he's in the middle of recording a podcast and he
34:53
can watch We model for dresses and give me his input on which
34:56
one he thinks look best. And I've never gone to the
34:58
store. I've never interacted with the
35:00
website. And she doesn't even need to
35:02
take the photos because it's just taking them from her
35:04
Instagram and her TikTok just put them on videos of her as
35:06
well. And this is easy.
35:07
I could do this today. I could build this out for you
35:09
today. Wish you've already built this
35:11
guys, It's live. And for the low, low price?
35:16
I yes, I I agree. I'm yeah, it is interesting.
35:23
Like where does the website live in that context?
35:25
But now she gets to do everything she loves.
35:29
She gets to browse, she gets to look at herself, she gets to ask
35:31
for your input. She like what she wants to do is
35:34
satisfy those desires. She doesn't actually care about
35:37
the Outlaw website. Yeah, totally.
35:39
I mean and websites are new invention anyways, right.
35:41
If you look at like shopping, it's just a, it's the way that
35:45
we distributed products online because that was the format that
35:49
somebody decided was the right format 30 years ago.
35:51
And that's why we're doing it. It's not that it's the best
35:55
format. I think it's Mr. Bezos who
35:57
decided it, but yeah. That's.
36:01
True, guys, you know what this is reminding me of, I'm not sure
36:04
if any of you've seen like Bellagi's Earth versus Cloud.
36:06
I'm not sure if I'm getting that right, but it's like this law
36:09
index thinking that like, you know, it's not, it's not going
36:11
to be blue versus red in the future.
36:12
It's going to be, you know, people who want to get outside
36:15
and people who want to get things done online.
36:19
I can't find it. Maybe it's earth versus cloud.
36:22
Maybe it's orange versus green. I think it's orange versus
36:25
green. It's a cool idea.
36:26
What do you guys think about like, because I'm sure, you
36:29
know, as we're talking about this, you, you're all thinking
36:31
there are people in my life who have no idea about anything AI,
36:34
right? They're not thinking about any
36:36
of these things. And they're they're refusing
36:39
outright to use AI. They're concerned when the New
36:40
York Times writes an article about how some soft person made
36:44
a poor decision because of AI, right?
36:47
And yeah, they're becoming increasingly, how should I say
36:52
it, hostile. It's not even just like
36:55
adversarial. They're becoming hostile towards
36:56
AI, right? And I, you know, I hate to say
36:59
it, but I, I wouldn't be surprised if the vast majority
37:03
of people end up here, right? If they end up like actively
37:06
hostile towards technology, you know, they've done it before.
37:11
So, yeah, what do you guys think about that?
37:14
Is there, is there a world where some of this stuff actually
37:18
fails on the demand front? For sure.
37:22
Yeah, dude, I'm, I can't share the context in which I am
37:26
observing it effectively first hand, but yeah, it's all, I
37:30
mean, it's already, it's already happening.
37:35
So, So yeah, there are, there are many places where people are
37:37
very, very adverse to it. Dude, Uncle, I've got to ask you
37:40
as a brand builder, one idea that I've been interacting with
37:43
is or sorry that everybody's been around dragging with like,
37:47
do you know Brian kennels new like one person brand like
37:50
thing, whatever, like running a bunch of agents to build brands.
37:54
There's another one. This guy started this thing
37:56
called Anwell just got funded by a general Catalyst.
37:59
And then of course, there's that GLP one story of it's like 1.
38:03
If it's like two brothers, do you think we're about to see a
38:06
wave of like 100 brand platforms being built with like one or two
38:13
people and a bunch of agents is like, is that what we're about
38:16
to see happen in the next three years?
38:18
Yeah, dude, if I wasn't doing this, I would be doing that
38:23
because it's the easiest thing you could do right now.
38:25
And, and I think every kid, there's kids growing up with
38:30
this now. So it's only a matter of a
38:32
couple years where they've had enough time with these tools to
38:37
go and do this themselves. And they don't need anybody else
38:40
because hell, there's a ton of kids with no friends and they're
38:45
going to, they're going to start building alone.
38:48
They're not going to need anybody else.
38:49
It's literally going to be them and their AI agents and that's
38:52
it. And man, I wish, I wish coming
38:56
out of high school we had the tools we had now, because I was
39:00
probably one of those kids. And I can only imagine what
39:05
they're going to, what they're going to be able to do with all
39:07
the free time that they have that we don't have now.
39:12
It's it's, it's going to be a sight to see, but that's going
39:14
to happen relatively quickly. I'm thinking, you know, within
39:17
the next year, you're going to just start seeing a lot of
39:20
saturation. And then in two or three years
39:24
it's I get crazy. We work with a bunch of
39:27
enterprises right now. And my biggest worry is like all
39:32
of those companies are asking themselves, how do we like end
39:34
up on the right side of AI? I think the biggest issue is
39:37
that actually who they're going to end up competing with this
39:41
people like this who have fundamentally different OpEx
39:44
structures. And when your OpEx structure is
39:47
that different, how do you transform, dude?
39:51
That's like an impossible challenge in my mind.
39:53
It's, it's one of those things where they haven't built habits,
39:59
they haven't been taught, so they're building in a completely
40:02
different way. And how do you combat an enemy
40:06
that you don't know what they're doing?
40:09
It's, it's like playing poker with somebody who doesn't know
40:12
how to play. You hate playing with that guy
40:14
because he, you have no idea how he's going to play it.
40:18
It's literally that. I I imagine some of the
40:21
enterprises that you work with Rashaab started as, you know,
40:25
low OpEx compared to the incumbents, right?
40:28
That, you know, Shopify in all of its technology pre AI kind of
40:33
eliminated the need for a web developer.
40:36
All right, in a similar way that AI eliminates the need for, you
40:40
know, an entire team of web marketers that are going to like
40:43
spin up landing pages and take weeks to do it.
40:45
It will do it in, you know, one prompt.
40:48
You know, I'm not sure if you guys have seen baseball
40:49
lifestyle one-on-one. I think that they're, I think
40:52
that's what they're called. I think they're called baseball
40:54
lifestyle. I think they're on the dorm
40:55
theme. This is a massive brand that
40:58
does I think 10s of millions a day at peak and.
41:02
They were going to, I think they were planning on hitting nine
41:06
figures last year, if I remember seeing those numbers correctly.
41:10
And yeah, just massive growth, like 500% year over year growth.
41:14
Insane, insane, incredible. They're eating someone's lunch
41:17
at those numbers rather than incumbent.
41:19
That's not selling backpacks and shorts to the kids.
41:22
Yeah, I think the business to be in is manufacturing.
41:24
If you can figure out how to manufacture for these, that's
41:27
where the money is because that's not that demand is just
41:29
going to increase. And if I was an incumbent who
41:33
had my own manufacturing, that's how I would be thinking about
41:35
it. I'd be thinking about how do I
41:36
manufacture for all of these people that are going to be
41:39
starting up. You make your money regardless
41:42
of how successful they are in that context.
41:44
They pay you to do a run a product and you're good to go
41:48
and move on. But I don't know anything about
41:49
manufacturing, so I'm not going to get into that business.
41:53
A lot of like just. Yeah, I don't know, man.
41:56
I think in the if you're anywhere are in and around ECOM.
42:01
I think the thing that Ankit was saying like Ankit sing the
42:06
words, if I wasn't doing this, that's what I'd be doing.
42:10
I think it's the thing that the, that people in and around ECOM
42:12
are spending the least amount of time talking about and should be
42:15
spending the most amount of time talking about because it's, it's
42:19
literally exactly what the next like Shopify and Shopify wave
42:24
is, is like, not like, hey, how do we fit to generate 100 new
42:29
ads? Like you guys literally said it,
42:31
it had no business impact. Like who gives a shit?
42:35
The thing that's going to have massive business impact is you
42:37
build a business that has an OpEx structure that's
42:39
fundamentally different than the old business.
42:42
And I just don't know why we're talking about other stuff.
42:47
You know what I mean? Like the other stuff doesn't
42:49
matter. Yeah.
42:51
I've been thinking about this a lot in my business, like how if
42:53
I were doing it over today, what would I do and why am I not
42:57
doing that today? And then how does that look?
43:01
Because that's the, I think that's the real question.
43:04
It's almost like you're, you're almost constrained.
43:05
You have to do it a certain way because of the thing like there,
43:10
there are obviously guardrails that have either been put up by
43:15
you or put up by somebody else. But if there were no guardrails
43:19
and you were doing this in your closet, how much more efficient
43:24
could you do this if you just started with a blank slate, like
43:28
you said? It's like, it's almost like no
43:31
one's talking about it because nobody wants it to happen, but
43:34
it is going to happen. But isn't that, I mean, anyway
43:38
that I feel like this is like a great place to end, but it's
43:41
like a dude we got we've got to talk more about that that actual
43:48
thing, because I think otherwise we're all just sort of like
43:51
pretending and the tornado is coming.
43:53
You know, I'm still I'm still bullish on people, but I'm
43:56
bullish on Co founders and founders, people who are like
43:59
hungry, waking up every day doing something interesting or
44:02
like at least setting up the leverage, right, Whether they're
44:04
finding someone really hungry who who can do it.
44:07
I'm still bullish on people, but those people, right, The people
44:10
who are like actually want to make things better.
44:13
Yeah, I'm a little concerned for everyone who's, you know,
44:17
looking at the tornado and going on would be fine, all right.
44:20
But the other, the other people who are out there, the building,
44:23
solar building or the building with their friends from like
44:26
everybody's high school or the building with, you know, other
44:29
very industrious, motivated people, hungry people.
44:34
Yeah, I'm bullish on. I'm bullish on most people.
44:36
Amazing. All right, dude.
44:38
Uncle, this was really fun, man. Yeah.
44:40
Thank you for coming. And thanks Ron, too.
44:43
Great. Catch them with you guys.
44:45
Bye.