Latest / 1 to 100 / E20: In Stealth Mode with Jacob Posel

Transcript

E20: In Stealth Mode with Jacob Posel · 1 to 100

0:00
Sidney Sweeney has good genes. I haven't seen that.
0:03
I saw just a buzz of some people complaining.
0:06
I'm that would be an interesting one to talk about because I have
0:08
no contact. So I'm curious what you guys
0:10
think about that, but. Guys, I thought one of them was
0:13
fake. There's there's so many
0:14
different clips that they came out with and one of them is just
0:18
balls to the wall crazy. Let me see if I can show you
0:21
guys. I assume no Zach today.
0:23
He's in Italy right now. He might join but I'm not sure.
0:27
Listen to this. It's incredible.
0:31
My body's composition is determined by my genes.
0:36
Hey, eyes up here. Sydney Sweeney has very genes.
0:45
I think that's real. I think that's real.
0:47
No, no, This one's real. This one's.
0:49
Real. That's real.
0:50
Dude, first of all, all right, we want to talk about SAS.
0:57
OK, so we're going to talk about SAS.
0:59
But I do think that ads of this variety are a plus, OK?
1:06
Just like the astronomer, like Gwyneth Paltrow Reed is A and
1:11
the reason they're A+ is they like, they're tapping into the
1:15
zeitgeist in a way in which like most other ads just don't.
1:20
Like most ads are like so worried about like a problem
1:24
statement or a solution or whatever, like all the like
1:27
stupid normal things that marketers do, but so few like,
1:32
tap into the zeitgeist. And like Sidney Sweeney is just
1:37
like leaning into the fact that she did the SNL skit, which
1:43
became like a meme, then Dr. Squatch hired her for ads and
1:51
she like is in the zeitgeist for a particular reason.
1:55
An American Eagle tapped her to play into that same like element
2:00
of the zeitgeist. And I think that that is
2:03
excellent. Like I think like to not do that
2:09
is like missing a marketing moment basically.
2:12
And I think you should, I think it's true for consumer like what
2:15
they did with the American Eagle ad.
2:17
And I think it's true for SAS like the astronomer ad or
2:20
whatever. The read by Gwyneth Paltrow is
2:22
so good because it like takes elements of like how like how
2:27
does the consumer or how do people, just normal people think
2:30
about things? But everybody now knows the
2:33
words Apache Airflow, even though they have no effing idea
2:36
what Apache Airflow like. You know what I mean?
2:39
Like nobody knows what Apache Airflow is, but everybody knows
2:42
what Apache. Airflow is and you only need 100
2:45
CTOS who care about that who reach out for a demo and it's
2:49
like the whole thing isn't worth it for them I'm sure.
2:52
No, dude, there's not even, you know what's going to happen in
2:55
practice. By the way, like Jacob sees me
2:58
do this, so he can attest to like how do SAS like, like how
3:04
does a sassy yo who doesn't know anything about technology
3:06
actually operate? It's like I hear a word and I'm
3:10
like, guys, and then I say the word, OK, so it's like I see
3:15
this kind of AD and I'm like Apache.
3:17
I'm like, guys, should we be thinking about Apache Airflow?
3:20
And I don't even know what the hell I'm saying.
3:23
But like the number of CE OS who are going to do that is going to
3:26
be massive because I'm more the average SAS CEO than like some
3:32
like deeply technical person. That's true.
3:37
Speaking of Jacob, I love Jacob, but why is Jacob on here today?
3:42
We should we should dive in on that real quick.
3:45
What's the? I was Jacob.
3:47
Jacob has been teasing the Internet with building a SAS
3:51
company and I wanted to know what kind of SAS company.
3:56
Yes, it's true like. How Jacob is thinking about what
3:59
kind of SAS company to build. And I can't get him to tell me
4:02
in private. So I figured, you know what I'm
4:04
going to do? I'm going to put him onto a
4:05
podcast and ask him. And we're going to and we're
4:08
going to figure out together what kind of like how does Jacob
4:12
think about how to build a SAS company?
4:14
Because you, you just hosted something.
4:17
Was it you said you signed the largest contract in your life or
4:20
something like that recently? Yeah, I I can't talk about that
4:24
publicly. That's fine.
4:26
But yeah, I mean, I, I, I don't, I don't even have an answer for
4:30
private, but I can tell you, I can tell you the the ways that
4:35
I've been navigating the journey and what, the lessons that I've
4:38
had and what I've been converging on.
4:41
Yeah, it would be cool too, because I mean you, I feel like
4:44
like even learning a little bit more your background, I know a
4:46
little bit. I've seen some of the stuff like
4:49
you are a common thread and I, you know, it'd be interesting
4:52
that to learn more about like what problems you saw, like what
4:55
were the things that you've been looking at and evaluating,
4:59
especially around the world of AI and all that over the last
5:01
couple of years? And how that makes you think
5:02
about like what SAS can look like in this context too,
5:05
because I think you've got a pretty interesting perspective
5:08
on that. Yeah, yeah, I have strong
5:10
opinions on SAS and AI for sure. For sure.
5:15
Yeah, just like strong opinions in general.
5:17
Just strong opinions in general. Loosely held, though.
5:19
I've heard about this though, like this whole stealth mode
5:22
thing. What is that?
5:24
Why do tech founders keep the the businesses they're building
5:27
a secret? Like what's their competitive
5:29
advantage? Relevant to me, the reason why I
5:31
don't want to say yet. So the like big contract thing
5:35
that is completely different. It's a completely different
5:37
thing. But to me, I think because I
5:42
actually have people who care about what I build, because I
5:46
have like some kind of inbuilt distribution and I've spoken to
5:49
Rishabh about this. It's like I think there's only
5:52
so many times I can I can use that.
5:56
There's only so many times I can leverage that distribution.
5:59
And so I want to be careful about doing that.
6:02
I want to be careful about kind of saturating that and launching
6:05
and relaunching over and over with different things.
6:07
I also think people aren't taking me seriously if I sell
6:10
them two different things, especially as I'm not totally
6:13
sure. But now that now that I'm
6:16
narrowing down into something that feels closer, I'm like
6:19
being like I've like kind of updated.
6:22
It doesn't say that I'm just like building something new.
6:24
It says building AI and advertising on my like socials.
6:28
And when I talk to people, it's I'm like very open with what I'm
6:32
thinking about building and what I want to build and stuff like
6:35
that. So that's why that, but that's
6:37
why I was initially pretty also because like it's fun to build
6:41
up demand and to get people asking you, you know, like build
6:44
up latent attention and then capitalize on it as the game.
6:47
But if you can't capitalize on it, then it kind of sucks.
6:50
So how and how have you been thinking about?
6:52
Like what is the way to Yeah, like what is the way to start?
6:59
So I'm sure a lot of people who listen to this that like we talk
7:03
about a lot of random things we talked about like the structure
7:05
of the company, like how do you fund it, all that sort of stuff.
7:09
But the very, very beginning, like the stage that you're in is
7:12
super interesting, right? And I mean, of course, like
7:17
Jeremiah, Zach and I all all did that a few years ago, starting a
7:21
SAS company right now. When there's like this massive
7:25
AI like undercurrent, like what does it make you think in terms
7:32
of either what like, and you pick where to start, like what
7:37
customer base to to address what, what is the structure of
7:41
the company? Like who are the first people
7:44
who you would hire? Like, is it a, is it a business
7:48
person? Is it another engineer?
7:50
Yeah, just like start, start with anyone dimension of company
7:53
building and and and share how you're thinking about it.
7:58
Yeah, So, OK, biggest lesson initially for me was I thought
8:02
that I could just build anything that I felt was the greatest
8:06
opportunity, which I convinced myself was true, but then kind
8:11
of proved myself to be wrong. I, I think that I need to be
8:15
somewhat inherently excited about what I'm building.
8:18
And so like going through the idea maze of this is a cool
8:22
thing to build. Should I build for consumer?
8:24
Should I build B to B? Should it be in a certain
8:28
vertical, a niche or should it be SAS or me?
8:30
Should it be e-commerce or marketing advertisers?
8:33
I, I started building a product which you, you have seen and
8:38
like, probably there was something there, but I just
8:42
couldn't get myself like inherently excited about it
8:45
enough. Where the problem was, I think
8:48
that a unique edge of a founder is you always have to be as
8:51
close to the frontier or like as close to improvement as
8:56
possible. And if I wasn't obsessively
8:59
thinking about it all the time, then I always felt a step behind
9:03
and there wasn't a clear vision to me of what I'm working
9:05
towards, what I'm building towards.
9:06
So with that, it was like selling to B to B SAS.
9:10
I didn't fully understand the problem space.
9:13
And so I felt, I always felt a step behind, but there's certain
9:18
niches or verticals where I'm like literally obsessed with it.
9:21
And I'm always up to date with it and I'm learning new things
9:25
all the time and I'm thinking about it and you all the time.
9:27
And that to me feels like I have a true competitive advantage in
9:30
building in that space. And so that's how and I just
9:34
kept getting pulled back to like marketing, advertising,
9:37
marketing, advertising growth, like I just getting kept getting
9:40
pulled back to it. Like the podcast I was listening
9:42
to the the articles I was reading, the part of Twitter
9:46
that I was hanging out and you know, so that's why I was like
9:49
wandering in the desert and thought thought this castle was
9:53
my castle, but then I came back to came back to the space.
9:57
So that's like the idea side, I think.
10:00
And to me related, like I have pre built distribution in that
10:07
area that I can capitalize on and that's exciting to me.
10:10
Like, so I've started building and I have, I think the best
10:16
market is that I know in my phone that I'm texting and
10:20
getting ideas and getting advice and like want to work with me.
10:23
And I've CE OS of big businesses who are interested in testing it
10:26
and trying it, you know, like that that sets me apart so
10:30
significantly if I was to build in different areas.
10:32
So I just can't discount that. And specifically related to AII,
10:37
think if you build it to me, I want to build something which
10:39
delivers valuable outcomes to people and the people who know
10:42
how to design products that deliver valuable outcomes, all
10:45
the people who are doing it. And so having that connection
10:49
feels invaluable too. Like if I can call on that
10:51
person and say, how, like, how is this maximally valuable to
10:55
you? How, how, how would you use
10:56
this? That feels, that feels
10:59
incredibly invaluable. And then I guess in terms of
11:03
hiring, this is the biggest. I'm sure like every CEO feels it
11:06
right now, but this is the biggest challenge.
11:11
Everyone good has a million opportunities.
11:14
And so I think it's like, how do you convince people to work with
11:18
you? And I have like a couple when
11:20
I'm, I think when I'm ready for it, I have people who who will
11:23
jump on board and be excited about it.
11:25
And to me, those have to be technical people, like super
11:29
technical people to build an amazing product.
11:31
Because I've also realized that the best businesses also just
11:35
have great tech, amazing tech. And I don't want to spend my
11:38
time in the weeds trying to figure out how this is the best
11:42
possible technical, how this is the best possible product.
11:45
Because there's a million different ways I think I can be
11:47
more valuable and provide more leverage to the business.
11:50
That's interesting. What, what when you so like one
11:55
of the the debates like we have had on this pod and I know
12:00
Jacob, we've talked about privately too, is like your ACV
12:06
basically determines your Tam basically determines what type
12:10
of business you can build. And like, I hate that the world
12:15
is so reductive, but you know, that kind of is helpful.
12:20
Like it, it is helpful to for things to be simple and when,
12:28
when you're building a product in the sort of like AI as like
12:34
world, how do you think about that question, right?
12:40
I think about it a lot because of you.
12:42
So like being like I've spent a lot of time in the format
12:45
offices and that that's made me feel like I've gotten my SAS
12:49
MBA. So like I put on now when I
12:50
think about I think like in two different ways.
12:53
I think in my like product minded way and my like SAS MBA
12:56
way. And I'm very cautious of that
12:58
now. Like how do I how do I grow a
13:00
business that can get me to 10 KCB kind of and easily
13:04
justifiable, easily justified that future.
13:08
And I think it's feasible when you're eating into when you're
13:12
eating into different, different parts of the budget of these
13:16
businesses. So like I compare myself to a
13:21
service business, which is like, I think where you have to go if
13:24
you, if you are ambitious in this space, like you have to
13:27
think, how am I no longer a tool or a platform, but I'm
13:30
delivering a service that people are willing to pay for.
13:33
And my belief is that the services that I'm going to be in
13:38
the short term augmenting or in the long term replacing are able
13:44
to charge significant contract, are able to like demand
13:48
significant contract sizes. And I'm not doing work, which is
13:53
exceptional, like to say the least.
13:56
And so I understand there is a difference there.
13:59
There is like an inherent difference in a human delivering
14:02
a service versus technology. And there are in there, there's
14:05
just kind of intrinsic differences and be able to pick
14:08
up the phone and calling someone.
14:10
But I'm trying to think how can I sit in that spectrum to where
14:13
I'm delivering enough value where the size of the contracts
14:17
that people are willing to pay what I need.
14:20
And there may be a user or someone sitting on top of that,
14:25
that at a cost to me makes sense or at a cost to an agency, for
14:28
example, makes sense and they're delivering that little piece
14:31
that is required. And I'm validating that by
14:34
asking people how much they're willing to pay or like having
14:38
having at least honest discussions with people who I
14:40
trust and how much they're willing to pay.
14:42
And like everything is leaning in the right direction.
14:45
Like it's heading in the right direction for sure.
14:48
So I think I do think about that a lot.
14:51
So let's talk about Icon, which has had this exact question this
14:56
past week and has spiced up the Twitterverse and Jack and
15:00
Jeremiah. I want, I'm like, very curious
15:03
how you interacted with that. Jeremiah, have you seen this
15:07
thing? Yeah, yeah.
15:08
OK. OK, OK.
15:10
So just to make sure everybody's on the same page, Icon released
15:14
a managed service offering. And for whatever reason, that
15:19
managed service offering drew a lot of criticism.
15:23
Well, at Taylor's post was perfectly reasonable.
15:27
There are other parts of the Internet where it yeah, where it
15:30
drew, it drew some more criticism.
15:32
I would, I would say. And I, I, I think the reason
15:39
people were criticizing it was they felt that it was advertised
15:46
as a tool and now it is like 180 ING into it's a tool and a
15:52
service, but basically. And then I, I think it begs this
16:00
much harder question of how should like, yeah, first of all,
16:06
I guess I'm curious how you what you guys thought of this?
16:08
What is your own opinions? And and then I'm curious
16:13
secondarily, where do you think the future of software is going?
16:19
Like does all software end up having a managed service
16:23
offering basically? Yeah.
16:25
Jeremiah, I'm Why don't you start?
16:28
I mean, it's interesting because enterprise software has always
16:31
been managed services essentially.
16:34
Like there's like a anybody who I know who works in like true
16:38
enterprise software, at the end of the day, they are these
16:42
massive contracts have humans who are doing the work of
16:46
implementing the software, managing the software,
16:48
customizing the software, doing all that kind of stuff for these
16:51
brands and I think are brands for these companies.
16:55
But I think AI just enhances that, right?
16:57
So like, I'm not surprised by this.
16:58
I think Jacob, you're, you know, you're talking about like the
17:01
line item that you replace ultimately, like the if you want
17:04
to make money, replacing humans as a line item in a business is
17:10
how you can make the most money. There's other things, but like
17:13
that is a an Ave. to make a lot of money because humans are
17:16
expensive. And so, you know, it sounds like
17:20
you're implying a little bit at least that you're kind of going
17:22
up against service providers for some of that stuff with what
17:24
you're building. I think I do really think like
17:27
what the world is moving towards a combination of two.
17:29
And so you'll have like, there are probably cheap software
17:34
solutions out there where you're willing to pay a small amount of
17:38
money to manage this thing yourself.
17:40
And then there's other more expensive services that are
17:42
going to require basically like offloading the, the thought
17:48
required to manage that system. And I, I actually think that's a
17:51
really underrated part of software.
17:54
Like something that that I don't believe people are just going to
17:58
vibe code everything in the future because at the end of the
18:01
day, you're not an expert in those things.
18:03
So like, when it comes to surveys, I have an entire team
18:06
of people who spend 40 plus hours a week just thinking about
18:10
how to use surveys the best way they possibly can for a
18:13
business. And can AI fill some of those
18:16
gaps? Yes, maybe.
18:18
I mean, maybe it gets really good in the future.
18:19
I have no idea. But I, I think at the end of the
18:21
day, I, I think most people are still going to trust a human to
18:25
support a lot of those capacities in a way that you're
18:28
not going to just trust a machine.
18:29
And yeah, that may not be universally true, but I think
18:32
that there's a piece of that that will always the the people
18:36
who are building the thing are going to be experts that are in
18:39
a way that like your internal team might not be.
18:42
So that's that's my time. But yeah, I'm curious what you
18:45
all think about that. I think of first of all, I'm
18:49
going to disagree with what you just said, but I'm going to
18:51
answer the other thing first. I think that lays like a pyramid
18:54
of value that is provided to to a business to like deliver a
18:59
certain outcome that they need. And you have, and that pyramid
19:03
also sits unlike a dimension of the total value that could be
19:06
provided at the tip of that pyramid.
19:09
And you have tools and platforms at the bottom, which, which are
19:13
like the very basics you that that you need like, I don't
19:16
know, like an ERP or something like that, just like sits at the
19:18
bottom. And then you have people that
19:20
need to sit on top to actually deliver the outcome which which
19:24
which you need for your business to function like like closing
19:28
your books or like creating your next purchase order with your
19:31
supplier, for example. I think that the assumption has
19:35
to be that AI is going to squeeze enough of that, enough
19:40
of that value, like towards the top where the the work being
19:43
done by the human is now at a scalable margin to where it can
19:47
be considered SAS, not like tech enabled service or a service or
19:52
you have to believe that you're able to move the value upwards
19:55
so that the total value being provided from the human plus the
19:58
tech itself is now at a greater is now greater.
20:02
And I think those two things are happening.
20:04
So like in marketing, advertising, I don't think
20:06
marketers and advertisers are sophisticated enough to say that
20:09
that tip, that top value point is fixed or it's saturated or
20:14
it's the maximum pop or it's optimal.
20:17
I think there's many different levers that you can provide.
20:20
If you can abstract away all of the busy work and the work being
20:22
done right now to then provide an even greater service and
20:26
charge even more for that. So like just just like to like
20:30
the difference between a really bad marketing agency and like
20:37
Eric who Eric Seaford, like Rishab's friend is probably in
20:42
in if we're thinking like contract value sizes is probably
20:44
like 6 figures in different. And it's like, well, what if you
20:48
can do everything that a marketing agency would do right
20:51
now? And then if you need a person,
20:54
that person is now doing so much greater, more like more valuable
20:57
work that you actually are able to charge more and be more
21:01
scalable and do it in a margin that makes sense.
21:03
So like that's the way that I think about software and service
21:07
kind of coming together, or at least like how, how you use AI
21:11
and, and a human to fill in the gaps.
21:13
And then to Jeremiah, to your last point, I, I disagree with
21:15
you. I think that I think that
21:17
software providers are actually exceptionally bad at
21:20
understanding like what their users want and building products
21:23
around that. And I think that in most cases,
21:28
the businesses themselves are a lot better at knowing what they
21:32
want. They just didn't have a
21:34
mechanism for describing it. And now that they can vibe code,
21:38
they are, they have a mechanism to kind of display that without
21:42
having to get on a call and explain to their like AE why,
21:45
like why they do what they, what they want kind of thing which
21:47
they don't want to do. And I'm seeing like it's cutting
21:50
through a lot of the noise and generating a lot more signal.
21:53
And so like that's, I actually think that that is going to be a
21:57
problem when it comes to vibe coding.
22:00
Guys I have a question, does the service the icon are offering do
22:04
the thing that the software is supposed to do?
22:08
Well, I guess Jack, the reason I wanted your opinion on this
22:11
question is like you guys sell as a tool, don't you?
22:17
Yeah, yeah, yeah. Workflow tool for sure.
22:20
OK. You don't have a service
22:22
offering that you sell. No service.
22:24
Why not? That's a good question.
22:26
I think that the end user is the service provider, its agencies
22:31
and its marketers at organizations, right?
22:34
So they are providing the services using the tool.
22:39
I think that's actually part of the problem.
22:41
I think that's why there was such reaction to the icon
22:44
announcement of having services is that ultimately like the
22:47
consumer of that product is largely service providers.
22:51
Yeah, they're, they're pissed because they saw it as
22:53
competition. They're pissed that Icon is able
22:55
to charge less because they raised a trillion dollars for
22:58
the same service. Like that's the, that's the only
23:00
reason why people were upset about it.
23:02
They don't actually, like, I don't think they actually felt
23:04
morally or ethically like hurt by the fact that Icon now said
23:09
that there were a service business.
23:10
I think they were just pissed by the competition.
23:12
It was all creative agencies that were upset about it.
23:16
Yeah, wait. Hold on, what is the sorry?
23:19
What does the price have anything to do with raising
23:26
money? This part of the like debate
23:29
that happened I actually did not understand.
23:35
You can pull up one of the one of the responses to to my tweet.
23:38
It was someone who runs a creative like an ad creative.
23:42
I saw that, was upset that they are now delivering a service
23:46
which is competitive to their service and the one that I
23:48
shared with you that went viral is the same.
23:51
The guy advertised his own creative services in the first
23:53
comment and they are. They are pissed because they
23:58
think that because Icon raised so much money, they're able to
24:01
charge cheaper prices, like $1000 a month I think, or
24:05
whatever it is, is absolutely cheaper than a creative agency
24:09
and I think that's the real reason why they're upset.
24:12
But guys like this, I guess there's just two questions we
24:16
have to answer. Does icon the software do the
24:19
thing they said it was going to do?
24:21
And does the service that they're introducing do the thing
24:24
the software was supposed to do? I think that the, like, don't
24:26
get me wrong, I think they're, they're super ambitious.
24:29
It could be an awesome company even as a service business,
24:31
right? And not a software business.
24:33
But I think the answer to both those questions is yes, it it
24:36
doesn't do the thing that it's supposed to do, and the service
24:40
does the thing that the software was supposed to do.
24:43
It if I was in their shoes, the way I would be thinking about
24:46
this, so I don't know if this is how they're thinking about, but
24:48
here's here's how I would be thinking about it given what
24:51
I've seen. They know they want to charge
24:54
1000 or $3000 a month per customer.
24:57
They know that the technology, like for the technology, and
25:01
they're looking at it and saying the tech doesn't deliver what we
25:06
promise it would deliver today. If we insert humans into the
25:09
equation, we can get there and eventually we'll remove the
25:11
humans from the equation. I think that's how they're
25:13
thinking about it. That's how I would be thinking
25:14
about it in their shoes. And so at least it gives you the
25:17
ability to ramp up this business.
25:18
Essentially, you have a blend of service and software and
25:21
ultimately it becomes a software business.
25:23
That's how I think. That's how they're thinking it.
25:27
Yeah. So, so, so Jack, I think the
25:29
answer to both of your questions is yes.
25:31
I also think that for this creative, like the person from
25:36
the creative agency who was upset, I, I, I, I would think
25:43
about it the way that Taylor thinks about it, which is like,
25:48
you have to embrace the fact that technology is going to give
25:50
you cost leverage and then continuously drive down the cost
25:53
of providing any, any given service over time.
25:56
Because by definition, technology is going to compress
26:00
the cost or the price of any given service.
26:02
So like, like assuming the same output, the price will go down,
26:08
which is not the same thing as like over time, you cannot grow
26:11
the pie for everybody because you're just doing more stuff,
26:15
right? I think that at the bat and I
26:20
have no idea whether they're going to be right or wrong, but
26:22
I think that they're bad is that technology will get to the point
26:27
at which they can deliver this total thing at a price point at
26:32
which they can support the building of a large company.
26:36
Because like the biggest problem with Icon when it first
26:39
launched, which has nothing to do with what anybody else is
26:42
criticizing them for, but I have like has been my forever drum
26:47
beat is the price was too low. It is like no more or less
26:52
complicated than that. And like you can't build a
26:56
venture skilled business if you're selling a $500 a month
27:00
product. You just can't, you can't do it
27:03
like I don't know any, I don't know any examples of a venture
27:08
scale business on a $500 a month product.
27:10
And so like they have correctly. It's got to be.
27:15
Well, OK. Yeah, yeah, yeah, yeah, yeah,
27:19
yeah. Unless it's like a consumer
27:20
product, yes. Or like a Slack or something
27:22
where it's like you can get a 10 million seats sold.
27:28
Yeah, yeah, yeah, yeah. I think like as soon as you say
27:30
seats, like now you're like playing a trick.
27:33
I know, I know. I know all the tricks.
27:35
OK. All right.
27:39
I'm, I'm trying to say I agree with you, there's a few
27:41
exceptions, but like it's totally different Tam, right?
27:44
Like the Tam is not a typical B2B Tam in that context.
27:47
Yeah, exactly. Exactly.
27:49
Exactly, exactly, exactly. Yeah, I think the.
27:51
Big challenge is the icons created for itself as well as
27:54
and I think there's nothing wrong with swinging for the
27:56
fences and trying to build something that can't be built
27:58
right now. Like I don't think there's
28:00
nothing wrong with that. But this kind of reminds me of
28:03
what you were saying earlier, Jacob, about not telling people
28:05
about what it is that you're working on, what you're
28:07
building. When you're building in
28:08
different directions. You don't want to cash in on
28:11
that check too early and your reputation is going to change.
28:14
If it's a product that doesn't have product market fit, that
28:16
doesn't go as far as your ambitions for it, right?
28:20
If it doesn't, if it isn't the right product, let's say, I
28:23
think their reputation with their market is potentially if
28:29
irrecoverably damaged, right? I.
28:32
Disagree there. I disagree there for sure.
28:34
Like everyone knows Icon, everyone's talking about Icon.
28:36
At the end of the day, they're not going to let their thoughts
28:39
on Icon get in the way if Icon actually delivers them 100 ads a
28:44
month at a beatable price because they want to grow their
28:47
business. So like, I, I don't think that
28:50
it's, I don't think it's wrong that Icon launched and
28:52
relaunched a million times and gained like created this much
28:55
attention. Like then there is product
28:57
market fit. Like given a product that works,
29:00
there is absolute extreme, one-of-a-kind product market
29:03
fit. It's just the only risk now is
29:06
does the product get there? And then like if the product
29:09
doesn't get there because the tech doesn't get there because
29:11
that's what they rely on, I guess the big risk for them.
29:14
Then Canon just like kind of fails and his reputation is
29:19
damaged, but everyone will still use them.
29:22
Everyone will use them. I'm going to share something
29:25
that I think is deeply ironic is at the end of the day, like you
29:31
have to like see what people do, not what they say.
29:34
And I kind of agree with Jacob like the market just does stuff
29:37
even if they like say it's unsavory, right?
29:42
Like this is like a tales all this time.
29:44
Now, on the flip side, his execution risk has actually
29:48
gotten extraordinarily large because the people who actually
29:52
get the most turned off is potential employees.
29:57
And the potential employees are the people who are the ones who
29:59
are going to build the technologies that actually get
30:01
them to that point. And that's where like image risk
30:08
comes in, because it's like you have to actually affirmatively
30:13
decide whether or not you want to be known as someone who
30:16
worked at Icon. And that I think is hard.
30:21
The rest of it I think is actually a non issue.
30:26
And so I think he has added risk to himself in a massive way, but
30:29
I think he has done it entirely on execution, not on market.
30:36
One thing that's that I think is really interesting is the way
30:40
they talk about like being an employee.
30:42
They're like, if you want to, I don't know if you guys are
30:43
probably saying someone else to like we work seven days a week.
30:45
We have you have to you like work has to be your identity.
30:51
Expect calls at 3:00 in the morning that kind of like, I saw
30:54
a bunch of this stuff and I'm actually curious how you guys
30:57
think about that because I, I, I have very mixed feelings.
31:00
I I actually respect the idea of like, if that is the culture
31:05
you're trying to build, being very clear about that and going
31:08
all in on that. But it also does seem like your,
31:10
your work pool and your, your talent pool.
31:14
I don't. Oh God, yeah.
31:18
Sorry, I yeah, I cut you off. I was just going to say I think
31:21
all of that can be fine if you can show that you can deliver
31:26
what you say you're going to do. The problem that I have is that
31:29
on the one hand you're saying, hey, we have this extreme
31:31
culture. And then on the other hand, the
31:32
market is like there are haters essentially, who are saying like
31:36
your product can't do the thing. Now my problem is like, I mean,
31:40
dude, I want to work somewhere where you call me at 3:00 in the
31:42
morning, but because we are killing it, not because we're
31:46
flubbing. That's, that's like the
31:49
fundamental problem, right? Like people at, but like, you
31:54
know, people at Fermat, it's like, I like, I, there's a
31:59
certain expectation of like the culture here.
32:01
I'm sure you guys have a certain expectation of the culture at
32:04
No. And it's like, and that culture
32:06
delivers this product. You know what I mean?
32:08
It's like you got to know what you're signing up for from the
32:13
inputs all the way through to the outputs.
32:17
I think to understand their culture, you have to understand
32:20
Ken a little bit. And he's like, he is, he is that
32:23
like he's not playing it like he is that person.
32:26
Like he, he a while ago, he, he tried to hire me.
32:29
So he, he called, he messaged me on LinkedIn.
32:32
This is when I can kind of blew up.
32:33
And then he texted me and then he Facetimed me instantly.
32:37
And like, I believe that he hadn't slept in three days.
32:39
I believe that he had spent all the time in the office.
32:42
He asked me the most insane questions that Aceo would never
32:46
ask someone. He told me things about.
32:48
He was like, he was like, yeah, I'm like neurodivergent.
32:51
Like there's something, there's a massive chip on my shoulder.
32:53
Like there's something, there's something wrong with me.
32:55
I don't remember exactly what he said, but like, he is, he is
32:58
literally that guy. And so like, I think like
33:00
unapologetically, like absolutely.
33:02
He believes that should be the culture and he wants to surround
33:04
himself with those people. And so like that, I think that's
33:08
like super important. I think there's a way that you
33:10
can, people can play that culture like they can play the
33:12
like 7712 or whatever it is, like seven days a week kind of
33:18
thing. I forgot what it's called.
33:20
But anyway, yeah, Canon. Canon is like literally like
33:23
that. So I think it makes sense.
33:25
I think that that culture works again, as Rishab said, like if
33:29
the product works, but you will, if you fail over and over, then
33:32
you will fizzle out and die. But momentum will drive that
33:35
culture because like everyone actually really works like that.
33:38
Like my friends in Silicon Valley, like they work like that
33:40
anyway. Like, we're all working all the
33:43
time, grinding it out in like random offices anyway.
33:47
It's just not publicized. And their employees are too.
33:50
But momentum is like the most important drive.
33:52
And so I think if momentum fails, then they're in a lot of
33:55
trouble, right? I mean, I don't think anyone's
33:57
immune for this. Imagine you bought a product and
34:01
it let you down. It didn't do the thing it was
34:03
supposed to do, and then you left at some time and it came
34:06
back with a new promise. It can do the thing now and you
34:08
buy it again. You spend some number of
34:10
hundreds or thousands of dollars and then it doesn't do the thing
34:13
that it said it was going to do. There is, it isn't 0 cost to
34:16
evaluate whether this product is going to do the thing.
34:19
And if it's you've already been burnt once or twice and these
34:21
are important people. Maybe your Tam has shrunk
34:23
because your ACV is nice and high.
34:26
How many times can you play that game right?
34:28
And no one's immune from this. I'm trying to think of a few
34:31
products. Cyberpunk was a video game,
34:34
right? I'm not sure if you guys
34:34
remember. Cyberpunk came out.
34:36
It wasn't finished. People were upset.
34:39
I think sales were impacted by the fact that the game wasn't
34:41
finished. There was another one called No
34:42
Man's Sky. I'm not a gamer, but I just know
34:45
that these products kind of didn't hit well with the market.
34:48
And then there's Apple Maps. Has that ever recovered?
34:51
Is that being used or are people just using Google Maps now?
34:55
Like Apple Maps was terrible when it first came out and
34:57
people just don't use it I'm aware of.
35:01
Let me give you a hot take on why this is different.
35:03
Hot take for for the podcast. This is this is the equivalent
35:06
of selling ads to e-commerce brands.
35:09
Imagine you went to a heroin addict and you told them that
35:12
and they really wanted heroin and you told them that this hit
35:14
of heroin was valid heroin and it wasn't you.
35:18
You could still offer it to them like 50 to 100 times and they
35:21
would still try it every time because they're an addict.
35:24
Like ads are the lifeblood of these businesses.
35:27
So, and it's like, and the way that these algorithms work is
35:30
it's literally an addiction. Like it's literally a you, you
35:33
become addicted because that one ad can get, can just get you
35:35
there, You know, like just that, that one, that one killer ad
35:38
that can crush can like take you to the next level.
35:42
And so I think he can keep doing what he's doing over and over
35:44
again until the product works. And I don't I don't think he has
35:48
harm of saturating his market. I mean, I understand what you're
35:53
saying. I do, though, also agree with
35:55
Jack. Like there is like the
35:56
evaluation becomes, there's more scrutiny and future evaluation.
36:01
There's more like time that goes inside it.
36:03
Like if you just wanted to quantify it like your sales team
36:06
is going to have to, it's going to cost you more from a sales
36:08
standpoint to close those deals than it would if it just works
36:11
the first time, right? So like even thinking about ad
36:14
platforms, because that is maybe a good analogy there.
36:17
TikTok, for example, like they do a lot of really like they've
36:21
been very powerful for a lot of businesses, but you also see
36:23
businesses that will try it. They they do keep trying it.
36:26
They don't feel like they get enough traction.
36:27
There's other platforms that are like that as well, Snap and like
36:32
basically anything but Meta has this problem where even Meta
36:35
probably doesn't. It's just not the same scale
36:37
where people will keep coming back and trying for sure.
36:40
But like the cost of that evaluation, the time it takes,
36:43
the amount you're willing to commit in future things, it's
36:46
going to like there, there is like an erosion of like your
36:49
expectation changes of, of what it's going to do.
36:51
And it does have some sort of impact.
36:53
I mean, I think about this in the businesses that I run, we
36:55
have this problem too. Like it's not, it's not the same
36:59
in like, I don't know, We, we do a pretty good job of delivering,
37:02
but the reality is we don't always deliver the first time.
37:04
In fact, like a lot of times the first version of something we
37:07
launched doesn't deliver and it's a hard thing.
37:10
Like it is hard to get somebody to come back and try it again
37:13
after it hasn't delivered. Like I, you're right, Jacob,
37:16
feel like this is a little bit different because of like the
37:18
cost and the business and all that.
37:19
But I still think there's some cost of evaluation and like
37:22
building of trust that happens. So and and the reality is every
37:25
SAS struggles with this. It's yeah, yeah, yeah.
37:29
Going to ever nail. Like the odds of you actually
37:32
nailing at the first time when you launch something new are
37:34
extremely low. Right, but how many first
37:37
impressions do you get right? Well, actually I would say it
37:41
the other way around. I'm not sure it's like you first
37:43
impressions you get. I think for any piece of
37:47
software for any given need, there is a cycle time and the
37:52
more extreme your product market fit, the more that time
37:56
condenses. So like Meta ads, I would bet
37:58
it's like the three months like some ridiculously short amount
38:03
like that. It's like, man, I couldn't get
38:05
it to work and then the new quarter starts and you're like
38:07
this quarter got to get meta apps to work, but it's not 3
38:11
days. I want to be super, I think
38:13
clear about that like that doesn't exist anywhere.
38:15
OK ChatGPT, probably 3 months, right.
38:20
People who are just like this AI thing every quarter it's like
38:25
they try it, you know, like and these are like the most extreme
38:28
product market fit products on the planet.
38:31
It's interesting, at Fermat I actually measured this precisely
38:34
for this reason, and we're at like somewhere in the 12 to 18
38:39
month time horizon, which I'm effing thrilled about.
38:42
That's really short. You get like every year you get
38:46
a shot. If I get a shot every year, that
38:48
is insane because that's like roughly the same timeline as a
38:52
renewal. That's crazy.
38:54
OK for for like, you know, for most SAS.
39:00
And so the real question for Keenan is just like, what is
39:03
this cycle time? Right?
39:05
And my guess is it's like definitely not three months, but
39:10
it's probably less than 18 months.
39:13
And so like, he just has to like weigh that cycle time against
39:17
like how he chooses to operate his, his business and his runway
39:22
and that's it. And that's the game.
39:24
This is exactly right. Like this is a.
39:26
This is the problem and I think you're, you've outlined it
39:30
really well. My guess is his is probably like
39:31
6 to 12 months to be honest, like because of the space he's
39:35
operating in, the problem is solving the the potential impact
39:38
of that. It's actually probably fairly
39:40
short. And so I guess going back to
39:42
Jacob's point, like it's a it is lower risk of him screwing up
39:45
than it is for me to screw up because mine is.
39:47
I don't explicitly measure it, but anecdotally I know it's not
39:51
like 12 to 24 month range as well.
39:53
I'm definitely not the expert on the consumer journey.
39:57
Jeremiah would be are better than me to speak on this.
39:59
I guess as a consumer, I can speak about it of marketing
40:02
tech. I feel like there are two first
40:05
impressions you get. You get what is it and what does
40:07
it do and does it work right and decision makers are trying to
40:12
answer those questions as quickly as possible, right.
40:15
One of the nice things about no. Is, you know what it is on face
40:19
value, the branding is really good.
40:21
I want to know more about my customers.
40:23
And the joke about Fermat is I don't know what it is.
40:25
I actually have to explore a little bit to understand what
40:27
Fermat is and then whether it works, right?
40:30
So that's kind of the effect of branding, right?
40:32
The, the whole understanding what it is on at a glance.
40:37
So I, I, I think that matching the way you go to market with
40:43
the consumer journey is pretty important, right?
40:46
Knowing your customer and then making that easy for them right,
40:50
to buy. It's actually.
40:51
It's funny too, because it's like, you know, you or should be
40:56
like you, you need a sales conversation, right?
40:59
So it's actually like a it's a conversation starter potentially
41:01
in that context, right? I'm looking for self-serve as
41:05
much as possible in the context of my business.
41:08
And so like as on the nose as possible to get somebody to
41:11
understand. It's actually like take action
41:12
is beneficial for me. So yeah, it is.
41:14
It is an interesting way of thinking about.
41:16
By the way, Jeremiah, you should do this study.
41:18
You want to know why not that you have an outbound sales team
41:21
right now, but eventually at some point you will.
41:25
Once you hit that number, you just outbound all of your
41:29
previous customers and you have like a surprisingly high close
41:33
rate. OK, I, I actually, how, how do
41:36
you measure that? So like what I, what I know when
41:38
I say anecdotally, I know that people come back, right?
41:41
So it's really common for somebody to work with us for 6
41:45
to 12 months, stop for a year, come back.
41:48
That's a really common thing, right?
41:51
For a variety of reasons, right? So I know that I know like those
41:55
conversations. You have to measure.
41:56
The only thing that you can measure is the affirmative
41:59
boomerang time, because you can't measure the negative,
42:02
right? The negative is infinite.
42:03
So like you have to you just take that number.
42:07
So everybody who comes back, look at that, look at the range
42:11
of like 20th to 80th percentile. Take the lower bound on that
42:16
range because you have done nothing in there Boomerang gang.
42:20
And so now all you with a little bit of an intent push, more
42:25
likely than not, you're going to end up at the starting point of
42:28
that range. And that's when you start
42:30
hitting the phones and you hit those people.
42:32
We did that this week and yeah, like the data is insane
42:40
actually. This is actually just like I
42:42
love this because I have hundreds of customers that have
42:45
churned so could be could be good.
42:48
I want to talk about an exchange you and Rashab had a while back
42:52
now about Rashab building tools for himself.
42:56
Do you remember? Yeah, dude, you said that.
42:59
You said that I would be vibe coding things for myself and I
43:02
said there's no way. Like it is not worth my time.
43:06
I think I said that there's a lot of things that you could
43:08
theoretically create that would be worth your time.
43:12
Yeah, dude, give me one. It's totally fine to do it on
43:14
the pod. Yeah, OK.
43:16
What if you, we'll just take the one that I said right now.
43:19
What if you created, what if it took you half a day to create a
43:21
tool that went into Salesforce, looked at churn date, then
43:24
looked at resign date or like when the second opportunity came
43:27
in and automatically created an outbound flow for every single
43:31
one of those customers. Like if that took you half a
43:35
day, then that's that's worth your time.
43:37
That like it's 100% worth your time to do that.
43:41
Dude, but this is the problem, man.
43:42
The problem is that like I already like I pay Salesforce to
43:47
take care of that for me. Like I just create a report in
43:50
Salesforce. Let's just pretend, let's just
43:52
pretend the Salesforce didn't do it for the sake of the argument.
43:55
Like I would much rather pay software to run that report and
44:04
then like my SDR team can like hit them basically right?
44:09
Can I tell you my problem with it because I have very strong
44:11
opinions on this as well. I'm 100% with you, Risha.
44:15
Like the my problem is that it's never done.
44:18
So I will go vibe code something.
44:20
I do this all the time. I build something for myself, a
44:22
first version of it, but I can't maintain that.
44:25
I don't have time to go back and try to figure out how to like
44:27
actually get the things that I need.
44:29
And so it's just a waste of my time at the end of the day.
44:31
Like this is what I've stopped doing it because inevitably what
44:35
happens is I get something out there, it's not good enough and
44:37
it never ends up moving forward. I just wasted 4 hours.
44:40
So like, it's better for me to find a solution that can
44:44
actually go execute that for me as opposed to me spending my
44:46
time on it. That's my, my opinion, but I
44:49
understand like I, I get it, but it's also like, I think stage,
44:53
like the amount of things you have to manage and where you're
44:55
at has has a different impact. Like five years ago it would
44:58
have been different for me because I could have spent two
45:01
days worth of time in a week on this.
45:03
Now I'm lucky if I have two hours in a week to actually be
45:06
able to think about something. And by the time I come back to
45:08
it, I've lost all the context and and the ability to actually
45:11
make any real progress on it. OK, that that wasn't, that
45:14
wasn't one of the examples I was, I was trying to get anyway.
45:17
I like, I think there's elements of the Fermat product that
45:22
Richard could spin up in a certain, in a certain
45:27
environment that would be closer to what he likes and that what
45:31
he wants for, for their like AI features, for example, because I
45:36
think there's like room to grow there in terms of some of those
45:38
features. And then he could hand that off
45:41
to one of his engineers and they would scoff at him.
45:44
They would scoff at him and they would say, I don't want to look
45:47
at this like you're not an engineer, blah, blah, blah.
45:49
But like if if it were properly handled.
45:53
This use case I agree with completely.
45:55
Sorry. Yeah, yeah, yeah, sorry to cut
45:58
you off, but OK, now we're talking now we're now we're
46:02
fucking talking 100%. I agree with this.
46:06
OK, OK yeah like AI making the CE OS ability to meddle
46:11
significantly higher held. Yes, I'm all in because like I
46:16
can go in and say like guys like yeah, so so if that's what
46:23
you're suggesting, like, hey, if the CEO can do X, then like
46:27
everything is going to change. I am 100% in agreement.
46:32
Could you have like a chief meddler, right, that you just
46:34
Slack message and they go and meddle for you, right?
46:38
And they're just so much faster, right?
46:40
And then you can go and do something else.
46:42
I think there's information loss that happens in a company where
46:45
like there's a clear goal of what you want, or maybe it comes
46:48
from a cut like you, like Richard, you see the most, you
46:52
have the most information, the most context on value in your
46:54
business. And then you go and then you say
46:56
tell Shreyas and then Shreyas tell someone else.
46:58
And then it goes to an engineer and then the engineer like has
47:01
their own spin on what happens. And then that spin is always
47:05
going to be different to what you wanted.
47:07
If you you could cut through and you could provide them the most
47:10
information and the most signal by providing them code that
47:13
represents what you want, I think that would be worth, I
47:17
think that'll be worth your time.
47:18
And it's a sick product. I think some, I think someone
47:20
should build it. Yeah, it's a sick product too.
47:22
I I do this as well and it's like if I've got 30 minutes I
47:26
can give you something that's more than just like words on a
47:28
page, visual, then it's it is more helpful.
47:32
Yeah, yeah. What's more signal per token
47:35
than code? Definitely not words.
47:40
Well, it's great chatting. Thank you for coming on.
47:43
Yeah, this is awesome, man. Thank you.
47:45
All right, see you all.