Episode Transcript
Transcripts are displayed as originally observed. Some content, including advertisements may have changed.
Use Ctrl + F to search
0:00
Oh, markets got a nice bump today.
0:02
You get a good feeling about your room? Yes.
0:04
Can I get these vibes from them? You know? I want
0:06
your room like, wake up on his
0:09
twenty twenty one side of the bed one day.
0:11
Just be like, you know what? Times forget
0:13
back then. We caused cataclysmic damage
0:15
that needed to get unwound at some point, but
0:17
While it was up, it was all good. Did I really
0:19
need to keep bringing down the hammer out? Let's take
0:22
this FOMC meeting off. I'm
0:24
gonna post the story at Instagram just on
0:26
a beach with a drink.
0:29
Who got the truth? Is
0:32
it you? Is it you? Is it you? Who
0:34
got the truth now? Is
0:37
it you? Was it? You? Was it you? Sit
0:40
it down. Say it straight. Another
0:42
story on Welcome
0:46
to season eleven episode eight
0:48
of Acquired, The podcast about great
0:50
technology companies and the stories and playbooks
0:53
behind them. I'm Ben Gilbert, and I am the
0:55
co founder and managing director of Seattle based
0:57
Pioneer Square Labs, and our venture fund,
0:59
PSL Ventures. And I'm
1:01
David Rosenthal, and I'm an angel investor.
1:04
based in San Francisco.
1:06
And we are your hosts. Today,
1:08
we are telling the entire history
1:10
and strategy of stratigraphy. and
1:12
we are joined by the man himself, Ben
1:15
Thompson, on the eve of Stratecri's
1:17
ten year anniversary. Stratecri
1:20
is innocently billed on his website
1:22
as the business strategy and
1:24
impact of technology. Stratekri,
1:26
as Ben puts it, started as
1:28
some guy in Taiwan with no access.
1:31
Today, Stratek Re is a powerhouse, shaping
1:33
the thoughts and conversation of the entire
1:36
technology industry around the world
1:38
at the highest levels. Ben has
1:40
interviewed Mark Zuckerberg, Jensen Huang,
1:42
Satia Nadella, Sundar Pichai, Rich
1:44
Barton, Meredith Kopit Levien of The New York
1:46
Times, Pat Gelsinger of Intel, and John
1:49
Carlson, and many, many more,
1:51
many of which are regular readers.
1:53
He is the father of aggregation theory,
1:55
the most important business framework
1:58
developed in the last twenty years. And
2:00
Ben really pioneered the subscription
2:03
newsletter business model. We sat
2:05
down with him to talk about a bunch of his
2:07
recent changes going big into podcasting,
2:10
bringing on cohost and expanding the empire,
2:12
and even now launching a Strotectory property
2:15
that he does not appear on at
2:17
all. Regardless
2:17
of what your business model
2:19
is, regardless even almost of what your content
2:22
is, I don't think there's any creator.
2:24
Certainly not any business creator
2:26
out there today who doesn't
2:28
directly or indirectly look to
2:31
Ben and say, you inspired
2:33
me, you paved the way for what
2:35
I do. I mean, we certainly feel that way. I feel
2:37
that way. David, I even feel like you're in
2:39
my friendship developed early on from
2:41
reading Ben's pieces
2:43
and sharing them with each other with our thoughts. We
2:45
have these old email threads in twenty fourteen
2:47
of you and I discussing Ben's writing.
2:50
Totally. We might talk about that on the episode
2:52
here. Well,
2:53
for our presenting sponsor this episode, we are back
2:55
with Fundrise CEO, Ben
2:57
Miller, and we asked him to share with us what
2:59
his inspiration was for launching the
3:01
Fundrise Innovation Fund.
3:03
So technology will disrupt
3:06
every sector. including Venture capital.
3:08
I don't believe any VC would
3:10
argue their immune, happen
3:12
to commerce, happen to news, entertainment, public
3:14
markets, etcetera. but
3:17
just like every other incumbent, they
3:18
may have trouble imagining it. And
3:20
so the mission of our company, Fundrise,
3:24
is bring about that sea change.
3:26
So how do we do it? So in my experience,
3:28
what happens is
3:30
an industry ages
3:32
and
3:32
the stories it tells itself which
3:34
were once true, become
3:36
no longer true. Did you saw this
3:38
play out in real estate?
3:40
Yeah. So typically in a financial industry,
3:42
the essential story is
3:44
about alpha. The incumbents
3:46
raise money and create alpha. That alpha
3:48
is usually about deal picking, making
3:50
the best investments. And
3:52
in other financial markets, for
3:54
example, public markets, technology
3:57
made those markets more transparent and more
3:59
efficient over time and that alpha went away.
4:02
and Vanguard basically consistently
4:04
beat the best hedge funds in the world. So
4:07
that's going to have venture capital. It'll
4:09
start where I think we're starting, which is mid
4:11
to late stage companies where there is a lot more data.
4:14
And that's what we're doing. The Fundrise Innovation
4:16
Fund is a public mutual fund. It's
4:18
like a vanguard for the private technology
4:20
markets. And so
4:21
if I can paraphrase, it's less
4:23
about picking the winners to guarantee
4:26
alpha but more about increasing
4:28
transparency and access so
4:30
that more investors can just get the
4:32
benefit of beta. These correlated returns
4:35
where when the market goes up you're
4:37
invested in that asset class that you get to go
4:39
up with it. Right? If you'd indexed into
4:41
cloud from twenty twelve to twenty
4:43
seventeen, And you just
4:45
said I'm gonna find data that indicates
4:47
the top ten or twenty percent of companies
4:49
with product market fit, you would have
4:51
done great. I learned this in two thousand
4:54
eight, the cycle and the market
4:56
is way more powerful than individual.
4:58
You think you're adding alpha, but you're actually
5:00
riding these mega trends. I
5:02
mean, it's gonna happen to venture. For
5:04
sure, it's inevitable. It happens to
5:07
every sector. I just wanna be
5:09
the one helping to it. And
5:10
for Foundees, once you're at that stage,
5:13
the
5:13
actual natural best fit and form
5:15
of capital for you is the
5:17
equivalent of mutual fund public market
5:19
capital. and that's what you're bringing for the first
5:21
time to the industry.
5:23
Right. Exactly. Our
5:24
thanks to Fundrise. If you want to
5:27
join the over three hundred and fifty
5:29
thousand individuals investing with
5:31
Fundrise, you can click the link in the show notes.
5:33
And if you're a founder and you want to get in
5:35
touch, about having the innovation fund to participate
5:37
in your next funding round. Email
5:39
not v c, that's N0TVC
5:41
at fundrise dot com.
5:43
After this episode, come hang out with
5:45
the other thirteen thousand smart thoughtful
5:48
members of the acquired community at acquired dot f
5:50
m slash Slack. We'll all be talking about
5:52
it. Without further ado, this is
5:54
not investment advice. This show is
5:56
for informational and entertainment purposes
5:58
only, and now onto
5:59
our interview with Ben.
6:02
Ben Thompson, welcome to acquired.
6:04
Thank
6:04
you happy to be here. It's great to have you with
6:07
us. We've done surveys over the years of our audience
6:09
and we say, what is the number one topic
6:11
you'd wanna see us cover or a person we should have on the show
6:13
and your name comes up many, many times. So
6:15
excited to be doing this. You'll promise me I was
6:17
number one. No. You'd feel like there's a bit of a
6:19
left out going We asked who the number one was
6:21
and you were very high in the rankings. Well,
6:23
we have another number one though that is very
6:25
near and dear to our hearts, Ben, searching
6:28
his email today, and he sent me an
6:30
image over text that was very sweet. It
6:32
was the first email between
6:34
us, and it was about contradictory. Okay.
6:36
Fine. I'll take that. David
6:38
and I, I think, got drinks and we're talking about
6:41
Uber, and it was your don't blame Uber piece
6:43
from twenty fourteen. Yeah.
6:45
Well, it's kind of interesting and retrospect
6:47
to think about that. I think that was about healthcare
6:50
and rights and workers and
6:52
things on those lines. it's pretty
6:54
interesting. You could probably write a similar piece
6:56
about easy money and the macroeconomic
6:58
environment and lots of pieces. So
7:00
mean, it's funny. I haven't thought about that piece in
7:02
years and years. I usually have
7:04
a very good memory of everything that I've written,
7:07
but every now and then, I will encounter
7:09
piece, whether someone linked to it, or
7:11
I'll come up when I'm searching for I mean, I'm the number
7:13
one user. I'm just trying to research. built
7:17
that for yourself. I completely forgot
7:19
that. I wrote about that. Usually, I would
7:21
go back and read it back, oh, that wasn't bad. Sometimes I
7:23
go back, like, I don't know I don't know about that
7:25
one. But, yeah, it is what it is.
7:27
That is what happens. You've written thousands of
7:29
posts over approaching ten years
7:31
now. It's just prolific. It's interesting. I
7:33
went back and just read the early days preparing
7:35
for this, you've posted you imported from a
7:37
tumbler in two thousand and nine that are,
7:39
like, all the way back in the back catalog. Yeah.
7:42
I've started multiple blogs through the
7:44
years. I thought they were relevant and interesting.
7:46
And also, I think it's useful
7:48
to have a back
7:50
catalog when you launch. And something
7:52
that I do when I watch podcasts now
7:54
is it's important to sort of establish
7:56
the first time someone encounters you
7:59
that
7:59
this isn't just a flash in the pan sort of thing.
8:02
It's something that is interesting. You mean
8:04
like doing thirty episodes of dithering before you
8:06
released a single one? you know, in
8:08
that case, we really wanted to
8:10
figure out what that podcast should
8:12
be. And then also working with
8:14
Johnny can be a bit of perfect is, so it took us
8:16
that long in the website looking to sell you wanted it
8:18
too. Guys, I don't know anybody else like that.
8:20
There was a technical component that went to that
8:22
as well. But I think it's good for both sides.
8:24
It's good for the creator
8:26
because if you can't
8:28
have the discipline and sort of
8:30
stamina to build up a back catalog,
8:32
you're probably not gonna have the discipline and stamina to
8:34
keep going for a long time. So it's
8:37
almost good internally to sort of
8:39
get that done. But then also when
8:41
someone encounters it and then they feel,
8:43
wow, there's an excitement, not just in
8:45
anticipation of new
8:47
episodes that'll be down the road,
8:49
But, wow, there's already a whole bunch of stuff here for me to listen
8:51
to. And if you like the first one, he goes into a second
8:53
one. And I think that sort of second one
8:55
is super important. This is the point I
8:57
made about writing in general,
9:00
anyone can come up with one really good post or
9:02
one really good podcast. I don't mean to say that
9:04
dismissively. I think there's a lot of people out there through
9:06
smart insights I think a very
9:08
distinct skill and
9:10
capability to come
9:12
up with interesting things consistently.
9:14
And the sooner you can demonstrate that
9:17
to someone, the sooner
9:19
they are going to take
9:21
advantage of whatever means you have to follow
9:23
or just Scribe or whatever it might be
9:25
because it's not just a promise
9:27
of consistency, but it's evidence of
9:29
that. And I think that's really important
9:31
to online businesses in
9:33
general. I think
9:34
about that element a lot.
9:36
I also think about every
9:37
piece of content you create is a
9:40
churn opportunity. David and I
9:42
regularly look at each other mid episode and we're
9:44
like, okay, is this of acquired quality?
9:46
Or is it actually net negative for us
9:48
to release this episode? Because we've now
9:50
reduced the average quality of
9:52
a thing that someone
9:54
comes to expect from us. Yeah. I mean,
9:56
well, thanks for putting the pressure on me right
9:58
up there. I better make sure this is
9:59
interesting. But interesting because I
10:02
would push back against that a little
10:04
bit. And this
10:04
is why the consistency part is important. I mean,
10:06
there's an ongoing discussion. I think there's
10:08
that new social network post news or something. And I
10:10
just saw a thing on Twitter. I didn't actually fall through, but they
10:12
wanna do micro payments for articles, which
10:14
I think are terrible. And
10:16
there's
10:17
a tendency and a lot of things to get
10:19
overly indexed on
10:21
the consumer, which kinda sounds wrong.
10:23
Like, why wouldn't you want to give the consumer exactly
10:25
what they want? But just to take
10:27
the most obvious example, if a creator
10:29
is not making money, well, the consumer
10:32
may get what they want, which is free content, but they're
10:34
not gonna get it for very long. there is a
10:36
short term perspective versus a long term
10:38
perspective, which well know it by charging for
10:40
this, I can do this over the long run. But from
10:42
a micro transaction perspective, I think one of
10:44
the issues is to
10:46
create the piece of content is very time
10:48
consuming. And then, obviously, you get
10:50
the free distribution on the back end. You have zero marginal
10:52
cost on the back end. But the problem is
10:54
you have a timing mismatch with micro
10:56
transactions between when the payments made and
10:58
when the investment is made. And so
11:00
you wanna put a lot of time into an article and
11:02
then hope it gets traction. So I
11:04
can not just make money, but also
11:06
repay what I'm doing from there. And I think
11:08
that the way to think about
11:10
publishing online is
11:12
what are you selling? I think when you get in
11:14
the trap of thinking you're selling single
11:16
episode or selling a single article, that's
11:18
actually getting the incentives wrong. What I'm
11:20
selling to my subscribers is
11:23
consistency and yes,
11:25
certainly a quality bar. When I write something that
11:27
I'm not happy with, I'm miserable for the next
11:29
twenty four hours. Or if it's the end of the week, I have to
11:31
wait till next week, and that really sucks. because
11:33
I certainly have the drive and
11:35
compulsion to make sure that what I put out is
11:37
high quality. But at
11:39
the same time, there's an aspect of
11:41
what I'm actually selling to
11:43
my subscribers. And this is more of an
11:45
implicit promise, but I think it's the reality is
11:48
the consistency and the
11:50
regularity and the knowing that something happens, you're
11:52
gonna be able to get my take on it.
11:53
And I think that's where subscription
11:56
pricing does make much more sense for
11:58
content production because in this case, you're
11:59
actually getting the money upfront. The money is
12:02
funding the work as opposed to the
12:04
work being a speculative bid
12:06
for the money. And one thing
12:08
that I've pushed over time
12:10
and I'm very happy about is
12:13
trying to get people to sort of annual subscription. So I
12:15
think my audience at this point is like seventy percent
12:17
annual subscriptions at a hundred and
12:19
fifty bucks a year. Is that right?
12:21
It's
12:21
twelve dollars a month, a hundred and twenty dollars a year. So
12:23
you do get a twenty four dollar discount
12:25
by being annual. You save on the straight fees.
12:27
Yeah. It's like a sixteen dollar difference.
12:29
So that certainly makes a big difference. But
12:31
then also, I just
12:33
think it's good for me and it's good
12:35
for my customers to know that,
12:37
look, I have a chance to make it up to you. your
12:40
Internet media business
12:42
about the business of technology
12:45
is a very
12:45
different business than our Internet media business
12:48
about the business. technology across
12:50
a number of dimensions. Right? There's obviously the
12:52
business model, your primary business
12:54
model versus our primary business model, you
12:56
being subscriptions, us being advertising,
12:58
but
12:58
there's also the periodic nature of
13:01
content. I think one of the reasons
13:03
we feel such pressure
13:05
on each episode is We're only
13:07
dropping one, maybe two episodes a
13:09
month. How many pieces of content are you doing a week
13:11
at this point? Well, less than
13:13
I once did believe it or not. usually
13:15
three written pieces, one interview,
13:17
two sharp tech episodes, and two dithering
13:19
episodes. So I guess that
13:21
is eight pieces of content. when
13:23
I first started the daily update, I was doing
13:26
two free articles, five paid
13:28
articles, and a podcast. So that was eight. So I guess
13:30
I'm doing more now, but revealing my secrets,
13:32
podcasts are easier to do than correcting,
13:34
so it feels easier today than it was
13:36
back then. It certainly
13:38
is a lot I don't think that
13:40
will always be the shape of it forever. But
13:42
yeah, there definitely is a lot more than two
13:44
per month, so it's a fair pushback. Alright.
13:46
So because this is acquired and
13:49
I'm playing format
13:50
police David, I'm gonna take us back
13:52
to the beginning, and we're gonna try not
13:54
to dwell too long because Ben, I know you've told your story elsewhere,
13:56
but I think it's important for context. So you
13:58
get this inkling out of an idea that you
13:59
can do this and
14:01
you have idea for a way to cover the
14:04
intersection of strategy and technology in a
14:06
differentiated way than has
14:08
been done before. and you decide that you
14:10
wanna start doing it. I think you are
14:12
still doing it while you're at
14:14
Microsoft in twenty fourteen. Is that right, Ben?
14:16
I had started tons of blogs through the
14:18
years. Ice in some respects
14:20
felt I had missed out because
14:22
I'm the same age as folks like
14:24
as requiring and Matthew Glasius and,
14:26
you know, I was right here for the school newspaper
14:29
and, you know, I started a blog and I'm
14:31
like, man, if I had sort of kept with it,
14:33
you know, I could doing what they do. And
14:35
then John Gruber started doing fireball around two
14:37
thousand three. And so there was some
14:39
aspect of, like, well, it's a bummer I missed out on that. it's
14:41
like the famous Mark entries and you show up in
14:43
Silicon Valley and you feel like, oh darn, it's already
14:45
all happened without question. So,
14:47
you know, suddenly people push me, you should write again, you
14:49
should do it. I think I did those Tumblr posts
14:51
while I was at business school. And, yeah,
14:53
I mean, it was definitely something that I
14:55
always wanted to do just from a
14:58
personal self interest perspective.
15:00
But also, it's not enough to
15:02
just want to do something. There needs to be a
15:04
market. There needs to be an opportunity. The way
15:06
I always put it is there's lots
15:08
of sites writing about the products.
15:11
And Wall Street is writing
15:13
about the financial results, but there's a big
15:15
gap in the middle there. You know, like, what
15:17
is the strategy that goes into the products,
15:19
like, what are the margins and the business models
15:21
that undergird those products that drive to the
15:23
financial results? How does
15:25
company culture impact decision
15:27
making? why do companies do stupid
15:29
things? This is a common question
15:31
in business school and you would
15:33
do these interviews, interview prep you're
15:35
the first year doing second year students, then you're helping the
15:37
first year students. And my favorite
15:39
question was always, what's a
15:41
decision or a product company has
15:43
made that you think is a bad idea or you disagree
15:45
with. And everyone always have an answer to
15:47
that, but the more interesting question is the follow-up
15:49
question, which is why did they
15:51
do that? And what you would see is a lot
15:53
of people, their answer
15:55
really was because they're stupid.
15:57
And the reality is no one at these
15:59
companies is
15:59
stupid. A very popular choice back then
16:02
is Microsoft making the Zoom players. Like,
16:04
that's a dumb idea blah blah blah. Well, why did
16:06
Microsoft do that? It's not just
16:08
useful for critique but
16:10
it's also useful for saying
16:12
where companies should go. So I
16:14
think probably Microsoft. Either
16:16
Microsoft or Facebook are probably my two
16:18
biggest successes as far
16:20
as an analyst. You know, I think particularly
16:22
Microsoft where I think
16:24
I very clearly
16:27
articulated what their issues
16:29
were. And also by understanding
16:31
those issues, the sort of organization they
16:33
were, the sort of capabilities they were, and
16:35
not just the weaknesses that entailed, but
16:37
also the strengths that entailed, where
16:39
they should go, and the sort of
16:41
strategy they should adopt. And that's
16:43
basically the strategy they have adopted and obviously
16:45
it's tremendously successful. But there's
16:47
no one writing that sort of stuff back in
16:49
twenty thirteen. one of the things we
16:51
thought about a lot was we
16:53
are not journalists, and we came
16:55
at this from having worked
16:58
in the industries that
17:00
we're talking about and obviously you did too having
17:02
been at Microsoft Apple and plenty of other
17:04
places. Was that in your mind like
17:06
everybody you mentioned As recline, Grupo is different.
17:08
We'll talk about Grupo later, but they were
17:10
journalists. They came out of the media,
17:12
whereas we were all coming out of
17:14
the industry. Yeah. It's
17:16
interesting. I wouldn't classify
17:18
either quinorglazingist as journalist. I mean, they
17:20
were bloggers. That's what they
17:22
were. And they ended
17:24
up as somewhat journalist.
17:26
Although, I would say both of them, their
17:28
best work has always been
17:30
more analysis than
17:32
journalism. I mean, maybe it's a stupid classification. But
17:34
I think about journalists as sort of
17:36
writing about what is happening at
17:38
best, uncovering facts. Analysis
17:41
is explaining why it
17:43
happened. And opinion is some
17:45
aspect of saying what you think should
17:47
happen. those have certainly gotten
17:50
conflated over the years. That's probably a
17:52
different discussion. But
17:53
analysis and opinion, I
17:56
think, do
17:56
go together to some extent, but it's actually a
17:58
line that I'm actually pretty careful about.
18:00
I really try to not break news.
18:02
That's actually an explicit
18:04
choice on my side. I mean, it was an easy
18:07
decision to make at the beginning because I knew no
18:09
one. I like foreign Twitter followers when I
18:11
started. Everything that I
18:13
wrote was self generated. Right? It was just
18:15
my own insight, my own view of the situation. You
18:17
didn't have any particular access. No
18:19
access. A negative access. I sort of
18:21
felt like it's hard because I couldn't write about Microsoft because I
18:23
was some on there. I think
18:25
that of my work experience, the
18:27
Microsoft bit, was certainly the most
18:29
impactful, not just because of what I
18:31
learned about Microsoft, but just what you
18:33
learned about big companies, I think, in general. And
18:35
it's
18:35
very easy on the outside to
18:39
anthropomorphize
18:39
these companies. And, like, a single entity
18:41
sort of making this
18:43
and doing what you want doing what you don't. And the reality is,
18:45
hits thousands and thousands of people. There's
18:47
massive coordination problems. Everyone has a different take,
18:50
different opinion. which is why the
18:52
culture stuff is so interesting, which
18:54
is why the broad
18:56
shared understanding of what a company is
18:58
and is not capable of. is
19:00
so important to understanding why companies do what they
19:02
do because actually tends to be much
19:04
more impactful than single individuals and
19:08
definitely seeing individuals below, like, the
19:10
CEO level. And even then, what
19:12
constraints the CEO's actions? Is a pretty
19:14
significant thing?
19:14
but they had to build a team to get
19:17
apps for Windows eight. And so
19:19
I was responsible for
19:22
social media,
19:22
other than Facebook and Twitter.
19:25
So, I guess,
19:28
blogging. I was responsible
19:30
for publishing So all
19:32
the book publishers, Amazon, Kendall, things on those lines,
19:34
I was responsible for
19:37
lifestyle So I really got to know the publishing
19:39
industry very well both on, like, the magazine
19:41
side, the newspaper side, book publishing
19:44
side. And all that was
19:46
certainly useful learning
19:48
for writing about those industries,
19:50
but
19:50
also very useful for
19:53
understanding apps or dynamics, developer
19:55
dynamics, it was super fun,
19:57
very busy, very high trying to get stuff out
19:59
the door, working with, you know, half
20:01
build SDKs and flying all over and
20:03
working with developers. And then sucked after the
20:06
fact. But it was about up until then. So I
20:08
was at
20:08
Microsoft, and I remember
20:10
my wife and I
20:11
knew it was our first episode of Hawaii.
20:14
I think this was the winter of
20:16
two thousand twelve to thirteen and
20:19
sat down and talked about it. It's like, yeah, it's probably gonna
20:21
be time to move on
20:23
pretty soon. I mean, anyhow, Microsoft
20:25
was always a bit of an upset. I was always sort
20:27
of an Apple person. That was my obsession.
20:30
I wanted to work at Apple. the good fortune of being
20:32
able to intern there.
20:34
And I was at Apple University when
20:36
Apple University had just started. And the original
20:38
vision for Apple University was that it's gonna
20:40
be really focused on Apple has this notion
20:42
of like the top two hundred, although I think it's actually today
20:44
more like top three hundred or something. They go
20:46
to this big company Retreat. They're the ones that
20:48
know all the future plans and are that's going on, and
20:50
that's really sort of the decision makers at Apple.
20:52
And it's not necessarily by management
20:55
position. Like, there's ICs that are, you know,
20:57
we have no management responsibility, but are, you know, sort
21:00
of part of that core. That was an
21:02
amazing experience. And
21:05
really, trying to figure out, like, what makes Apple Apple sort of
21:07
defining the culture. And
21:09
it did seem problematic that
21:11
it's very hard to write culture down
21:13
because then that's kind of a recipe
21:15
I think for sort of ossification and getting sort of
21:17
walked into something. What is interesting
21:19
is it was written a lot about as this
21:21
sort of Steve Jobs initiative. And,
21:23
you know, I'm sure he ok'd it, but he had basically no
21:25
involvement at all. Tim Cook was very
21:27
heavily involved. Like, the whole Tim Cook doctrine
21:29
thing that folks talked about during during
21:31
the earnings call, that was at Apple
21:34
University first. Did you coin the Cook
21:36
doctrine? No. I think
21:38
Horace Deju did from a CIMCO. Oh,
21:40
yeah. The reality is his culture comes
21:42
from doing and it doesn't come
21:44
from saying. And there's
21:46
a issue, like, if you've made decisions based on
21:48
data. Right? It's inherently backwards looking because the
21:50
data has to be generated first. And
21:52
I think there's analogy to culture
21:54
along those
21:55
lines that and the concern
21:57
about writing it down and getting focused
21:59
on that is you're sort of locking
22:02
yourself in. Like, cultures are very, very
22:04
powerful. It's the
22:04
way that you coordinate a
22:07
massive company and keep
22:08
it going in the broadly same direction because
22:10
there's so many things that you can articulate.
22:12
You can articulate that this is the way we do things.
22:14
If you did, you would be bogged down like
22:16
the transaction cost. of communicating every
22:19
little detail are massive. And so
22:21
culture keeps you going in the right
22:23
direction, but culture is also very
22:25
dangerous because if you have to
22:27
change direction, suddenly you realize you're in a straight
22:29
jacket. And I do
22:31
wonder if writing
22:33
stuff down and saying this is the way we
22:35
do things is actually, like, sewing the
22:37
striped jacket. I'm not sure how that pays off.
22:39
Right. Culture and process accomplish the same
22:41
goal in many cases. And so either it
22:43
can be done in a process lightweight because
22:45
the culture sort of facilitates it getting
22:47
done, but then you have no
22:49
processes to change something if the culture is no longer
22:51
facilitating what you wanna do. Yeah.
22:53
I think that that's an interesting way to put it.
22:55
Culture is probably the processes that
22:57
can't be written down to some extent
22:59
because it's just an understanding of this is how
23:01
we do it. I left that
23:03
summer feeling like this is not the
23:05
right thing for me. You get to business school.
23:07
You have to get a job. Everyone else is getting
23:09
jobs and you feel you know, and it was pretty tough for me.
23:11
Can you feel like what you do immediately
23:13
after business school? Like, have there such weight
23:15
to that career decision? And then you just, like,
23:17
yeah, whatever is, like, a dumb Well, I
23:19
was hell bent on being in tech.
23:21
And I had lots of opportunities
23:23
and offers. Obviously, consulting was an
23:25
option. It's interesting. It's not a very non
23:27
traditional background. an English teacher in
23:29
Taiwan, you know, I built a sort of
23:31
distributed presentation system for a
23:33
a group of schools here. So there was like some sort of
23:35
like tech angle per se. But, really,
23:37
I was just, like, gonna go teach from Taiwan. We, you know,
23:39
back to sort of the Esrecline mavaglesias,
23:41
you would kinda cut from that cloth before
23:43
getting into tech. Right? Like, you were polycyne
23:45
major. Right. You worked in politics?
23:47
Yep. And honestly, a lot of that was
23:49
my background, and I grew up definitely
23:51
blue collar, but my parents were
23:53
Minnis and you're in a small town of
23:56
two thousand, like, the pastor is, like,
23:58
the most educated person around by
23:59
and large. So there's an
24:01
aspect, well, alternative to
24:04
working in the factory around the farm
24:06
is you go to college and you'd
24:08
be like a professor or something. There just
24:10
wasn't even any awareness
24:13
any sort of opportunities. And
24:15
I think it's very not well
24:18
understood by elites in general
24:20
and people on the coast where
24:22
it's not a lack of opportunity. It's a lack
24:24
of even knowledge that there are opportunities
24:27
elsewhere. And
24:28
so for me, going to
24:30
their University of Wisconsin was like this big act
24:32
of rebellion and like, oh, I'm not gonna go to a
24:34
bible college or I'm not gonna go to
24:37
whatever. And in retrospect,
24:40
had the grades and test scores
24:42
to go somewhere, you
24:44
know, I mean, South is a great school, but I probably
24:46
could have got to an idly scores. I didn't
24:48
even apply. that just wasn't on your
24:50
radar screen. Not even on my radar
24:52
screen, not on the radar screen of anyone around
24:54
me. And I think this is a perspective
24:56
that is probably not very well
24:58
understood by a lot of our audience, I think, to
25:00
say the least. And so I
25:02
actually love I was very
25:04
intuitive. I like I was the first person I knew around
25:06
me to get an email address to sort of be
25:08
online. Did you get the
25:10
Internet at home before you went to college or was college
25:12
your first experience No. I pestered
25:14
and pestered my parents to get
25:16
Internet. I mean, I think we started out with like a
25:18
Juno account, which is just email
25:20
only. Oh, wait. Was you know the one that was a Desha spin
25:22
out? Yes. That was an incubation
25:24
at Desha when Bezos was there
25:27
before Amazon? Yep. It was like a stand
25:29
alone client that was email only and then
25:31
emerged with net zero. Yep.
25:33
So even in college, you know, I was
25:35
super online, me and my group of friends, in the
25:37
dorms an extra year just because of the broadband
25:39
connection. Broadband is used very loosely
25:41
here. I think it's like two megabytes up and down. because I
25:43
worked in the suit in his paper, which I loved.
25:45
One of the things that
25:46
did at the newspaper was
25:49
I took over
25:49
the editorial section in
25:52
my senior year so
25:54
we sat down and we had five things that
25:56
we wanted to see happen. Well, first, we decided
25:58
number one, editorials will only ever be
26:00
about things that affect students. And what we
26:02
did was look, We're gonna write an editorial
26:04
every single day, Monday through Thursday. Friday, the
26:07
whole page will be given away to guest columns. We're
26:09
gonna write an editorial every single day, matter
26:11
what and we're gonna write about these five issues.
26:13
And did that mean you writing an editorial
26:15
every single day? Well, so we had editorial
26:18
board, but basically ended up me writing editorial every single day. It's
26:20
what it came down to. Not every day with
26:22
the vast majority. And it's
26:24
interesting because there is certainly there's
26:26
an what ended up doing with my career. Right? This
26:28
sort of daily hitting
26:31
points, having to generate content.
26:33
But it was, number one,
26:35
just from a practical perspective, it was
26:37
fun and interesting. It's something I definitely
26:40
enjoyed. And number two, it worked of our
26:42
five sort of things we
26:44
had goals. We accomplished four of
26:46
them.
26:46
though So I wanna go forward to
26:49
you leaving Microsoft Give
26:51
me the emotional moment where you're
26:54
like, it feels like it could
26:56
work. Like, I think there's
26:58
something here. I think I have content
27:00
market fit. Well,
27:01
so to go to the back to the wide trip, there
27:03
was an aspect about going to Microsoft.
27:05
It's like, well, I should go to a place
27:07
that is directly oppositional to everything I
27:09
believe in. Right? And can I make changes
27:12
at Microsoft? You know, very sort of
27:14
arrogant. Really, I learned a lot from Microsoft
27:16
as opposed to the other way around. I sort
27:18
of signed up for knowing this is probably not
27:20
where I was gonna be long term. And
27:22
after that trip
27:23
is when I did start stratigraphy, I
27:27
wanted to figure out a way to back some point. My wife's from
27:29
Taiwan, and we enjoyed living
27:31
here. Also, again, I just felt
27:33
like there was this hole and
27:35
opportunity. And I sort of looked at Gruber as
27:37
someone that was like, well, he did
27:39
it, but it was also pretty
27:41
clear that an advertising based
27:43
model was probably not going to work. I mean,
27:45
that was obviously a a regular theme of mind about
27:47
the centralization of advertising under Google and
27:49
Facebook and that world of starting a
27:51
blog and throwing out some Google ads was not
27:53
going to be a viable one. What
27:55
made you think that? To my mind at
27:57
least, there is a very
28:00
viable and robust alternative history
28:02
where Sterectomy is an
28:04
advertising business model. Well, I
28:06
think that I would have to have started
28:08
much earlier. And I
28:10
think Gruber's model is the right one where I
28:12
think it is actually underrated that. I think he
28:14
actually kind of invented native advertising
28:16
because you read piece of content that he
28:18
writes, and one of them just happens to basically be an
28:20
admin. It's the same format, same content as what's
28:23
there. And I would bet the
28:25
rate of consumption to encountering
28:27
it for his ads are like higher than
28:29
almost anything else out there. And
28:31
obviously, Facebook does that at
28:34
scale. and it's all automated, self serve, and
28:36
you scroll through, and there's different pieces of content you're interested
28:38
in, and then some of them are ads. But you didn't
28:40
think you could do the same thing that
28:42
John does. Well, I thought that
28:44
was a possibility, a potential leg of
28:46
the stool. But number
28:48
one, I didn't know that I could do the
28:50
same foreman. I mean, what works
28:52
very well is because most of his stuff are
28:54
short snippets and wink out. It's very easy to
28:56
check-in and read everything. I
28:58
ended up moving more towards
29:00
longer pieces. I mean, I did and
29:02
play around with that format.
29:04
And it didn't seem right to me.
29:06
And then also, just strategically,
29:09
you could see where things were
29:11
going. this idea that
29:14
users are going to Google, going to social
29:16
media. And that's the
29:18
best and most obvious place to put
29:20
advertising. That's where the users are, understanding about them, all
29:22
those sorts of things. You know, advertising is
29:24
an ROI measurement. It's
29:26
not just what's your return, but also how
29:28
much work do you have do to get it. And
29:30
why would someone wanna go to a small blog and
29:32
put the effort of putting an ad there?
29:34
It
29:34
didn't seem like a very scalable possibility.
29:39
And meanwhile, At
29:39
the same time, there been this new company that was something
29:41
called Stripe. And it seemed
29:43
viable to me that if you thought through the idea, look,
29:45
if I produced super highly differentiated
29:48
content, that is something
29:51
that is not gonna appeal to the whole
29:53
world, but to the people who like it, they'll
29:55
really like it. And if you have that
29:57
sort of
29:58
audience, you
29:59
should maximize your revenue per user,
30:02
and the way to maximize your revenue per user
30:04
is to charge them a subscription. and
30:06
the tools were then becoming available to
30:08
sort of do that just to validate the
30:10
point real quick. And, David, this is my
30:12
pushback on there being an alternative history where
30:14
Strotectory could have been ad driven. for
30:16
anybody who doesn't read during fireball and doesn't know who
30:18
John Gruber is that we're talking about quick lesson
30:20
when Ben says that he invented the infeed
30:22
advertisement, you go to daring fireball
30:24
dot net. yes dot net. And you see the same exact
30:26
website that's been there for ten, close to
30:28
twenty years at this point, and most of
30:30
it is links out to other things.
30:32
some of its well thought through pieces, and
30:34
some of it is sponsored post. And
30:36
I consumed just like many other
30:39
people the entire feed in an RSS reader
30:41
and the sponsored post would show up just like any
30:43
other post in an RSS reader. Now
30:45
I think the set of people who are likely
30:47
to use an RSS reader and who are likely
30:49
to go directly to dairy fireball dot net.
30:51
Does the place where they're intentionally typing in
30:53
URL or clicking a bookmarks bar
30:55
and going to it? that set
30:57
of apple nerds is far more likely
30:59
to do that behavior than anybody else in the world.
31:01
The world is shifting away from going
31:04
directly to pages and a lot of
31:06
your readers are like, I'm just getting my news from Twitter and wherever else
31:08
at this point, and I'm not doing the RSS thing.
31:11
Yeah. I think the more niche you
31:13
are, interestingly, that does pay off
31:15
subscriptions. It can also work well for advertising. I
31:17
think Gruber was at the right place, the right
31:19
time for Apple for sure, not
31:21
just in terms of Apple's growth,
31:23
but all sold the explosion in the App Store. There was a
31:25
period where you would have lots of apps as a
31:27
sort of featured things. And so it
31:29
works great for him. again,
31:31
I did have sponsored posts a
31:33
little bit when I did go independent because
31:35
I figured I'd have multiple revenue
31:38
streams. I didn't remember that. How
31:40
long did you do that for? Oh, like six
31:42
months. Again, the ROI wasn't there for me either.
31:44
I couldn't charge that much. and it was a
31:46
big hassle and getting it arranged. And then
31:49
also, there's a lot of complicated
31:51
factors, like pushing people to email was not a
31:53
good fit for that. And so I
31:56
started Streckery pretty clear that
31:58
subscriptions would be the core model. I was
31:59
gonna try lots of things to monetize. Clear
32:02
to you though, but you pioneered that. I mean,
32:04
there was no sub stack. I think in
32:06
many ways, substack was I don't know the exact
32:08
history, but probably modeled after
32:10
you. Yeah. No. Substacks, CDEC,
32:12
says Sierczakarina Box. That was
32:15
their So this was a big innovation. Yeah.
32:17
I mean, I think like the subscriptions did exist.
32:19
I think on Wall Street in particular,
32:21
but they were generally like twenty thousand
32:23
dollars. and you would get all the hedge
32:25
funds to subscribe and the banks if that model
32:27
still exists. And it's it's actually a a very very
32:29
profitable one. But what I do think
32:32
strategically innovated was sort of
32:34
subscriptions at scale where
32:35
you're charging a low price
32:37
relative to twenty five thousand dollars or
32:39
ten thousand dollars and you're
32:40
doing on a self serve basis. People sign up with a
32:42
credit card. And again Stripe was a really important part
32:45
of making this possible, but there were
32:47
no real subscription tools. Like, I had
32:49
to actually hacked it all
32:51
together when I first started. There was, like,
32:53
Wordpress subscription tools, then there was Stripe
32:55
and you had to glue them together, and
32:57
it was sort of a big mess when I started. A year after
32:59
I started a company called MemberFUL then
33:01
launched, which I switched to them. That big
33:03
thing is definitely much easier.
33:06
And then, obviously, I I have my own system now. You've
33:08
worked with an outside development agency to custom
33:10
build software for which you are the
33:12
only user to publish Protect right
33:15
now. Right? Yeah. Well, I mean, at the beginning, it
33:17
was all me. Like, I built my own page and
33:19
I did all that integration myself. And I'm
33:21
not really a developer, so it was very
33:23
happy. Actually, one
33:25
of my nightmares was when I watched
33:27
the paid product. So, sergeant
33:29
Stackery, I messed around format a little bit
33:31
at the beginning. Actually, the real pioneer
33:33
who I should mention was Andrew
33:35
Sullivan. And
33:36
he had
33:37
this manic posting schedule, like,
33:39
tens of posts a day, and you did have a team to help
33:42
him on the daily dish. And
33:45
he switched to a subscription when I think
33:47
he left the Atlantic. was
33:49
successful. Like, he quickly was
33:51
doing a million dollars in revenue a
33:53
year, but he sort of maintained
33:55
that manic posted schedule that kind
33:57
of made sense in an advertising driven world where
33:59
you want people always coming back and getting lots of
34:02
impressions. And the paywall
34:04
was super loose. It was like thirty posts and
34:06
then you hit a paywall or something. And the reality is what
34:08
happens he burned out. I think he had some health
34:10
issues and he ended up bleeding. And it was funny
34:12
because when
34:14
he left Everyone's like blogging's dad, it's finished, was a
34:16
failure. And I looked at him, I thought, like, no, I think this
34:18
was actually a huge success. One of the
34:20
problems was he
34:22
had this posting schedule
34:24
and system and then you try to drop a paywall on
34:26
top of it. And so when I thought about trajectory,
34:28
I wanna do the opposite. where
34:30
I wanted a subscription to not
34:32
feel like from a customer perspective that
34:34
I was taking stuff away, but that
34:37
I was giving them more. just
34:39
from a psychological perspective. And so once
34:41
I realize I definitely wanna do a
34:43
subscription model, I started becoming super disciplined
34:46
about never writing more than twice
34:48
a week. this was back in twenty thirteen when I joined Security. my
34:50
plan was for subscribers, if
34:52
they subscribed, they'd get more.
34:54
And it would be like an
34:56
exciting purchase as opposed to automating a paywall. This sucks. And your audience
34:59
that you were building up before putting the
35:01
paywall in would still keep
35:03
getting the same. That's right. And so
35:05
my initial model was still daring fireball like where I figured, oh,
35:08
I'd have big articles.
35:10
And then if you
35:12
subscribe, you're get a whole feed of a
35:14
bunch of little stuff, my commentary on news
35:16
articles, I built a new
35:18
website again as a not
35:20
very good web developer. And
35:22
the metric I was paying attention to
35:24
was people who visited
35:26
the home page on days I didn't post.
35:29
because to me those were people
35:31
that were hoping I had written and
35:33
they wanted more from me. And I
35:35
figured the classic, if I could XYZ
35:37
of y. But I thought if I get ten percent to this
35:39
audience, that would be something that would be
35:41
reasonable and that metric was pretty high.
35:43
I mean, I didn't appreciate than just how powerful social
35:45
media was and publishers bimonth
35:48
social media, but those are publishers with
35:51
old business models. If you have a new business
35:54
media is a godsend because you basically get
35:56
free marketing. What a lot of people
35:58
don't what about you don't understand understand
36:00
is I think a lot of writers give away way too much content on Twitter and
36:02
they're out there posting. It's like, why should I read your site
36:04
when you're just telling me everything you think on
36:06
Twitter? And The reality the
36:08
power of social media is not that it gives you
36:10
another platform. I have a platform at strecker dot
36:12
com. What it does do is
36:14
give all my readers
36:16
a platform to tell other
36:18
people, wow, this site's really great. Look how
36:20
smart I am for having a take on this thing that
36:22
Ben wrote, which you can read about on his website.
36:24
Right. If you share something
36:26
that's great, you get the likes and you get the retweets and you get
36:28
the status of people feeling like, wow, this person
36:30
is sharing good stuff. And
36:32
that's really powerful. This
36:33
is probably an angle you
36:36
mentioned the marketing and recent quote when I thought I was too late. In this case,
36:38
it turned out I was early. I
36:40
was the only person doing this
36:43
sort of model out there And
36:45
there was still an era of twenty
36:47
thirteen Twitter, twenty fourteen Twitter where
36:50
people would share links
36:52
to something and it would break through if it was successfully good. This
36:54
is also where the principle of
36:56
having the second article be
36:58
good
36:58
is super important.
37:00
know, and this is I think unique to the
37:02
social media era where people follow link to a
37:04
site and they have pay no attention to
37:06
what the site is. They don't even know what it is.
37:09
basically a Twitter article for all types of purposes. So
37:11
I did want the site to be sort of visually
37:13
distinct. So I did like a custom font, which
37:15
back then was very rare. That was sort of
37:17
a new thing. I had the orange,
37:19
you know, there weren't very many orange ice back then, and I did a lot of these sort
37:21
of hand drawings, which were very
37:24
visually distinct.
37:26
the Ben Thompson iPad hand drying. You're a good artist. The
37:28
boats, even in the first post,
37:30
the sailboats, was quite good. Yeah.
37:33
Take goodness for iPads and the ability to sort of
37:36
edit that nicely. But the reason for
37:38
that is what I was really thinking
37:40
about was, oh, hey, they found an article
37:42
like, oh, that's a good article. a day later,
37:43
a week later, a month later, they follow another
37:46
wink. And like, wow, that's
37:47
gone. Wait. I've been on this site
37:49
before. It's triggering my memory
37:51
And that's really, I think, the key moment. And it's like,
37:53
oh, yeah, I'm gonna follow this guy on Twitter or I'm
37:55
gonna put this site in my bookmark. So I'm gonna put
37:57
my RSS reader or whatever it
38:00
might be. And that's sort of the key thing that I
38:02
think you want to accomplish. This
38:04
is one of my big criticisms of
38:06
sub stack
38:08
I mean, it's interesting because social security, I think, is achieving some degree of
38:10
network effects where everyone's familiar with it, that
38:12
your payment's already on file. All that
38:15
stuff is true. but nothing visually about SubSec is
38:17
memorable. Every single site looks the same. And,
38:20
also, they'll be reading a post
38:22
after, like, bring down the address
38:24
bar number. Who am I reading again? The
38:26
kerning's also weird. This is a weird
38:28
typography in it, but I'm like, it always looks slightly
38:30
wrong to me when I'm reading the
38:32
body text. Yeah.
38:32
I agree. Does trajectory even the name? I know in
38:34
many respects it's a bad name for
38:37
a word-of-mouth sight to be a name that
38:39
no one
38:39
knows how to pronounce. not
38:42
audio word of mouth to sight. It's a --
38:44
Word of tweets. -- it's super
38:46
memorable for good or bad
38:48
reasons. It's one of those No
38:50
exposure is bad exposure sort of
38:52
things. More importantly, the URL existed,
38:54
the Twitter handle existed. It didn't
38:56
exist in Google Search. So there was plus
38:58
or minus When did you come up with
39:00
it? The name that I've always been the most
39:02
jealous of is a Simcoe, again, to go back to
39:04
Horace Deju. It has all
39:06
those qualities. with the advantage of being pretty noticeable and
39:08
easily available. So in
39:10
retrospect, I always make fun of the
39:12
trajectory name because it's easy to make
39:14
fun of But
39:16
I do think there's a lot of positive qualities to it that
39:19
are underappreciated from this
39:21
memorability aspect. But there is an
39:23
aspect where it is harder for
39:25
adults. And people are afraid to say it because they're worried they'll say it wrong. And if
39:27
they say it, they don't wanna spell it, so
39:29
it's not perfect by
39:32
any means. It is based on the
39:34
strategic movie bit. Right? Where it's supposed to be strategicory?
39:36
Yeah. So it's strategy and
39:37
tech.
39:39
And, yes, there was a strategic bet. And so then
39:41
I'm like, well, I should pronounce a strategicory. That was a terrible idea. I think I did that
39:43
for the month. Didn't even
39:46
say on the website it's
39:48
pronounced strategicory Yeah.
39:50
At the very beginning. Yeah. And then I quickly realized that was a very bad
39:52
idea because it is strategy and taxes should
39:54
be trajectory. Yeah. Lots of poor decisions. It's very
39:56
easy to look back and see all the good ones.
39:59
But this is how you build brands. The only way is
40:01
to, like, do this stuff. I feel like every
40:03
single time we talk to someone who
40:06
built something in their own
40:08
image from first principles with their ideas
40:10
in a super opinionated way about how they
40:12
wanted to sort of like birth their
40:14
Internet child into the world. It has all
40:16
these really rough edges for a
40:18
while. Yeah. And that applied to lots of pieces. I
40:20
mean, I had my people visiting the home
40:22
page of days I didn't post. So in the meantime,
40:24
what's up? I was really looking for a
40:26
job at Microsoft where I could still blog.
40:28
Isn't that all jobs at Microsoft? You just kinda chill
40:30
and do plenty of time to blog? But I couldn't
40:32
write about Microsoft soft. And also, it
40:34
was getting a lot of attention. Again, I was
40:36
surprised at how quickly it
40:38
blew
40:38
up. I mean, this is where the other Gruber
40:40
story comes in where and
40:42
appendicitis, so I'm laying in a hospital bed, getting ready for surgery, and
40:44
I'm like waiting around for a few
40:45
hours. I had
40:47
waited to
40:50
email John because for the same principles at the beginning, I wanted to have a
40:52
body of content, not just like an introductory
40:54
poll say, oh, check out my new blog. How many emails does he
40:56
get about
40:58
that? Right? And
40:59
so I waited a couple months or a month or so when I emailed
41:01
and say, hey, been a big fan. He had just been
41:03
on a podcast
41:03
on his origin story, and he also worked the
41:06
newspaper in XYZ
41:08
So I shared that, you know, like, a rule related to this. I regret
41:10
not doing what you did back in two thousand
41:13
three, but I'm giving it a shot now. Here's
41:15
a few articles I'd be interested. and
41:18
I heard nothing back. I didn't notice spots. And so,
41:21
you know, I had my surgery. He
41:23
was very successful. No problems. couple
41:26
later, I'm at Microsoft, and I get
41:28
an email from him
41:30
pointing out
41:30
that I made a grammatical error. I
41:32
think it was really a Jive Jive
41:35
Jive, like, I used the
41:36
wrong one. And he's like, I make this mistake
41:39
all the time. He's like, tell me the etymology
41:41
of each one, and that's all
41:43
the emails said. No closing.
41:46
Like, I love your work or just like, no, full
41:48
stop. It was just explaining that I used the wrong
41:50
word and what different words meant, what I probably meant
41:54
to use. So I thought, well, he's probably gonna post a link to the site. So I was
41:56
very excited. So I'm, like, sitting on the analytics page, and I'm
41:58
sitting on Danny Fireball, seeing what he
41:59
will post. And I was
42:02
hoping for just a short link. Right? He has
42:04
all those short snippets. Instead, he
42:05
it's a full article and
42:07
he starts out saying, all my
42:09
readers should be reading this new site I've discovered.
42:11
I've discovered, of course, college
42:14
trajectory. It's the best new site I've
42:16
encountered in years. and they list, like,
42:18
three articles that I've discovered with
42:20
the help of its author. Right. He's
42:22
like, you said, this article this
42:24
article, it's like, five hundred words of
42:26
just the most glowing praise
42:28
imaginable. And then he gets to the
42:30
end. He's like, but for the first time, I
42:32
disagree with Thompson. And then he spends another
42:34
thousand words saying why I was wrong about something,
42:36
which was fine. I mean, I think he was wrong. I know it's
42:38
the start of a beautiful friendship. Do
42:40
you remember numerically
42:40
what happened to your traffic after
42:43
that? So the
42:43
thing I always was indexed on was
42:45
Twitter followers because that was a very
42:47
clear metric of people who had
42:50
affirmatively decided that they wanted to know
42:52
what I had to say. And I
42:54
don't think it's the case anymore, but for a long
42:56
time, my percentage of subscribers
42:58
actually followed was a fixed percentage
43:00
of my
43:02
Twitter followers. paid
43:02
subscribers were up fixed percentage. Like, people who subscribed
43:04
to the daily update? Yeah. Which meant
43:06
I
43:06
knew my limiting factor on subscribers was
43:08
awareness of who I was. So
43:11
I had five hundred Twitter followers when
43:13
he posted that. And within
43:15
twelve hours, I had fifty hundred
43:17
Twitter followers. And That sounds very small. I mean, I
43:19
have two hundred and thousand some followers. Now, tripling
43:22
out hour over hour is a great rate to
43:24
compound that.
43:26
That's right. And also, it was a lot of people that
43:28
were also bloggers or writers
43:30
or influential. There was
43:33
two step changes, instructors growth. That
43:35
was really the first one for sure. The next day, I get reached out by
43:37
the
43:37
head of strategy at Microsoft or a week later. He's
43:39
like, hey, can I we have a meeting? I'm like, gosh, shit.
43:41
I'm in trouble. he
43:43
gets summoned to the principal's office. At least it's not the head of PR.
43:46
That would have been way worse. So
43:48
he ended up
43:48
being with him. He wanted to hire me. He's like, well,
43:50
I'll he'll bump you up to level sixty five.
43:53
you know, I kinda wanna look at This is Kurt Delbeni. Who was the
43:55
head of the strategy at the time? No. His name
43:57
is Charlie Songhurst. Oh, yeah. So
44:00
and I kinda
44:00
sipped about it. I knew he was probably leaving soon.
44:02
He believing shortly after he's like, no. Go live in Taiwan. Just fly back once
44:04
a month. It'll be fine. But fortunately,
44:06
I turned him down. But All
44:09
this is a circuitous way to say, I did feel I needed to
44:12
leave Microsoft. Like, this thing was taking it off.
44:14
It wasn't sustainable. Number one, I
44:16
quit right
44:18
about Microsoft. and it was stuff that people at the highest levels cared It
44:20
wasn't, like, writing about the finer
44:22
details of VPNs or something
44:25
along those lines. And so I'm looking for a job and up
44:27
in automatic, and that's where I met Matt. So you weren't
44:30
looking for a job in your, like, heart
44:32
of hearts. Were you hoping that
44:34
you could someday not have a job? Without
44:36
question, no. She definitely was my plan. You
44:38
just felt like you couldn't do it yet. Yeah. I
44:40
mean,
44:40
I had a family. Like, you know, I had business school
44:42
debt. I needed to have a job that paid the bills. I wanted
44:44
to find a job that would let me
44:45
continue to do this. So,
44:48
automatic
44:49
worked out it actually
44:51
worked out even better because I was, like, hired
44:53
as, like, a growth engineer or whatever, which
44:55
I didn't really know anything about. But
44:58
then the team I was
45:00
on shortly after I was there to building, like, a new back
45:02
end of WordPress. And so
45:04
there was nothing for me to do. We didn't have a
45:06
product out But
45:08
fortunately, there was a little bit of pushback internally because
45:10
I was becoming this very visible blogger. It's like,
45:12
what the hell like, this guy's working for us? I thought
45:14
we're just to have because back then,
45:16
Automatic was one of the few distributed companies who live in
45:18
where he wanted to. So I as soon as I got the job, we
45:20
moved back to Taiwan, and you
45:22
weren't supposed to have a second job for
45:24
sort of obvious reasons. And I think Matt did
45:26
run some interference for me where
45:28
he's like, no. Don't worry. He's good. And the
45:30
one thing I did do
45:32
for automatic like is, I went
45:34
through a lot of the company what they were doing, and
45:36
I wrote this big strategy document
45:38
about where they're at, what they should do. It
45:40
had some impact, I think, on where they ended
45:42
up going. you guys should buy Tumbler? Yeah. That was at the time when they were it
45:44
had been a very small company. And at
45:46
around that time, they decided to raise a lot of money
45:48
and grow into being sort of a much
45:50
bigger one. again, I don't know how
45:52
much impact I had. Matt says I had a big
45:54
impact. He may be being generous. I don't know.
45:56
But I I kinda actually end up working mainly
45:58
directly with Matt. And he mostly
46:00
left me alone, and then he would have a
46:02
question or whatever, and then I would think about it and write
46:04
something up and come back, well, nominally being the
46:06
growth engineer for this team that did not have a product
46:08
in the market. So I owe automatic a lot, but it was also very
46:10
stressful. I felt really guilty. I'd say I'm
46:12
like, I'm really devoting all my
46:14
energy and time to this trajectory and I'm
46:16
getting fade.
46:18
by automatic. Just that was internally very stressful to me.
46:20
And, you know, that wasn't the idea, like,
46:23
I mean, I sound cliche, but I, you know,
46:25
braze in a Midwest blue collar
46:28
tone. Like, I'm getting paid far more than I've ever made any time my life
46:30
and I don't feel like I'm devoting
46:32
all my time and energy to this company that
46:34
was very internally stressing.
46:36
And so
46:37
though I
46:38
got this job to be like a consultant for this company
46:40
that wanted to expand Asia Pacific. And
46:43
I'm like, well, I can do this consulting
46:45
job, and then I can
46:48
make sure that you're gonna pay things. I couldn't do that once that automatic again. You couldn't
46:50
have a separate pay job while being there. And
46:52
so you left automatic to go do this
46:55
sort of potential consulting thing. Well, no. First,
46:57
I left automatic because I'm like, I
46:59
need to build this subscription
47:01
bit and this new
47:03
website. consulting job. So I left automatic. I'm
47:05
working on building this. What happened was that
47:08
construction shop fell through it. It just didn't
47:10
materialize. Oh,
47:12
no way. you burned the boat and then there was no
47:14
car. So I watched the Checkatrade and
47:16
the idea was you would log in, you get this much
47:18
fuller experience.
47:20
and I watched it. And number one, I had something wrong with
47:22
security certificate. So the first twenty four hours no
47:24
one can make a purchase. And so
47:27
I'm like up thirty
47:28
six hours straight. I barely slept for the week before because I'm, like, finalizing
47:30
the site. I stupidly had, like, reached
47:33
out to, like, Cara Swisher at Rico,
47:35
and so she had prescheduled post that
47:37
was gonna downsize going paid. And so I got this stupid artificial
47:39
deadline that made zero difference other than
47:41
stroking my ego. And
47:43
so it was awful.
47:44
It was a horrible experience. But the worst
47:46
thing was
47:46
that the product sucked. It was super
47:49
confusing. I didn't achieve my
47:51
goal of having a good experience for nonpaying
47:53
customers that was ideally lure them in. And for
47:55
paying customers, it didn't work well. It was
47:57
very janky, and it sucked. It was
47:59
really,
47:59
really bad. So over
48:01
the weekend, I'm just miserable. And I know
48:03
I
48:03
screwed up and I burned my bridges and all this
48:06
sort of thing. And I realized
48:07
that, look, I messed up. the
48:10
business model's right. The product's wrong.
48:12
So over the weekend, I tore the whole thing
48:14
out. I went back to the
48:16
old website. and I told the people that subscribed, I'm like,
48:18
look, I appreciate you subscribing. The
48:20
formats all wrong. What I'm going to
48:22
do is
48:24
for your subscription, I'm just gonna email you. I'll
48:26
email you once a day with the extra
48:28
stuff that I promised you. And
48:31
so I felt completely and utterly ass backwards
48:33
in the email. I thought it would all be
48:35
on the website. I've been thinking this whole time. I'm like, we
48:37
have not talked about email once. Yeah. And
48:39
by mentality is always the chef's website first and
48:42
foremost, but it's like, look, I'll deliver you this extra
48:44
stuff via email. And it will also be on the
48:46
website, on the sidebar, and you'll it'll be
48:48
archived there. but you're gonna get an email. And it worked out
48:50
in the long
48:50
run. In the short run though, I had
48:52
a one day
48:53
goal for new subscribers well,
48:55
I guess a two day goal because the site wasn't operable for the first
48:58
day. I had a one
48:59
day goal, a one week goal, and
49:01
a one month goal. and I
49:03
failed to reach all of them and failed pretty significantly, honestly, to reach
49:05
all of them. And so I'm there. I've
49:07
given up this six figure job
49:09
living in
49:11
Taiwan, business school debt
49:13
and an idea that I thought could work but did
49:15
not seem to be working very well.
49:17
And it was very,
49:19
very stressful. I mean, I stopped paying all my credit cards because
49:22
especially back then Taiwan was very cash centric, so I
49:24
needed to preserve cash.
49:26
I was, you know, not sleeping. I also
49:28
watched a podcast at the time, which was
49:30
exponent with James Allworth. So I am
49:32
writing, as I mentioned
49:34
before, seven times a week, plus doing a
49:36
podcast. I'm doing all the editing for the podcast,
49:38
all the all this sort of
49:40
stuff. And I thought I'm gonna have to, like, go teach
49:42
English again to, like, pay the bills at the same
49:44
time. And then I
49:44
think, like, the first month when
49:46
I tore all that out, I screwed up the subscription, so I double charged the whole
49:48
bunch of people, set up process all these refunds,
49:51
all those customer support. And it
49:53
was
49:53
pretty tough. I watched it
49:56
in April. This is April
49:58
twenty fourteen, fifteen. Twenty
50:00
fourteen. So next year, it'll be
50:01
ten years directory, but nine years of
50:04
it being my job. So fast
50:06
forward to the summer. We're back in the
50:08
states. My wife and I had planned a
50:10
long standing trip. We're gonna go
50:12
to Paris. and we the kids with my parents. You fast forward there a little bit, but you
50:14
must have some thoughts in those intervening months where you're
50:16
like, well, shoot, I'm just gonna have to give up on all
50:18
this. Right? Yeah. But, like, there was certainly a
50:20
prime factor.
50:22
I couldn't bail on it. Like, I mean, no one thought I would succeed. I had lots of
50:24
people reach out, be like, hey, bad, I love you,
50:26
but subscriptions are not a thing on the
50:28
internet. It's not
50:30
gonna work. and a lot of people
50:32
intact, especially VCs, which is weird because they always know exactly
50:34
what's right. No. They're a very conventional wisdom
50:37
driven, I think, So
50:40
I can still picture it. I was sitting at the kitchen table
50:42
at my parents house in Wisconsin
50:44
and you looked
50:45
at the numbers and
50:46
numbers and subscribers
50:48
were
50:48
more than April, June was more
50:50
than May, and July was on pace to be more than June,
50:52
and we're still talking like a couple hundred,
50:55
not very many. oh, and I'd also have, like, multiple levels. So some
50:57
of everybody bought, like, a three and hour level where I was supposed to, like,
50:59
chat with them and have subscriber calls and all this
51:02
sort of stuff. Yeah. Everybody starts that way
51:04
and then quickly gets rid of
51:06
it. And I'm also trying to do, like, sponsored
51:08
posts, really just an overwhelming amount of
51:10
stuff. But I looked at those numbers, I'm
51:12
like, I think it's going
51:14
to work. And so
51:16
I paid off all the credit cards that had been
51:18
running out for the last few months. My wife's like, should
51:20
we cancel the trip? Like, no. We're not canceling this trip.
51:22
you know, she was very mad at me. She was like, you threw away this six figure job. And so
51:24
I didn't wanna tell her all poorly things were going.
51:26
You didn't give her access to the
51:30
analytics. No. Of course not. Like, no. We're not cats ago. Triptics are going great.
51:32
But so we flew to Paris, ran the credit cards
51:34
all back up again, of course. And
51:36
in Paris, I remember I
51:39
would wake up every morning at four:zero a.
51:41
m. sit in the bathroom and write a daily update.
51:43
And then my wife would wake up and we would sort
51:45
of go for the day. and that ended
51:47
up being truth. It just sort of kept increasing. I mentioned there was two big step
51:49
changes, the Gruber one being the first one.
51:52
In November,
51:55
I actually ended up reaching a thousand, which is my
51:57
one year goal. So I hit my one year
51:59
goal
51:59
much
51:59
earlier than I expected, one year, a hundred dollars a
52:02
year, a hundred thousand dollar run rate. thing about exponential
52:04
growth. You missed the early goals, but you nailed
52:06
the later ones. Yeah. I don't think exponential
52:08
growth really applies to my business, but maybe more
52:10
so at the beginning. There's more of an exponential curve.
52:13
And the problem with a word-of-mouth business and
52:15
that special growth is people run out of
52:17
people to talk to, and so that's the limiting
52:19
factor. Right? It's only the new
52:21
subscribers give you the exponential. Right. There's a little bit
52:23
of exponential with every new subscribers. They will tell
52:26
new people, but networks
52:28
get exhausted. And so it's more than linear, but it's if
52:30
you zoom out, it's really linear is the way this
52:32
sort of growth works. But so what happened was
52:36
in November,
52:37
november I
52:38
put a little post out. First, I simplify my model, so no more of
52:40
this three hundred level cheap level, one
52:42
price and one product. That's all
52:46
you're getting. And then number two, I'm like, hey, I got a lot of subscribers, business
52:48
model works. You know, this is definitely gonna
52:50
be my job sort of
52:52
going forward. And
52:53
in the next twenty four hours, you got two
52:55
hundred fifty new subscribers. By far, the
52:57
biggest step change in my subscriber growth.
52:59
Did all the VCs email you again and
53:01
say, Ben, you're a genius. Will you please
53:03
come talk to our portfolio companies? I was right about
53:05
my metric of people visiting the home page and
53:07
days I didn't subscribe. What
53:10
I was wrong about is that the vast majority of people thought I
53:12
would fail and go out of business. They didn't wanna lose
53:14
their money. And so
53:17
they didn't subscribe. But once it was clear, I would
53:19
be an ongoing entity that, like, oh,
53:22
okay. I guess, my money will be safe,
53:24
and so I will
53:26
now subscribe. I was
53:28
right about the market.
53:28
I just I didn't understand that psychological
53:30
aspect. One of the things I think
53:32
is great for sort of the I went back to
53:34
the subject, make it easy to pay.
53:36
I think the real great thing for SubStack and SubStack writers today,
53:38
it's not that it's easy to pay,
53:40
but people aren't scared to give their
53:42
money to some random writer out there and
53:45
they previously and now they're not. And I think
53:48
that's really great. It's something I'm very proud of.
53:50
Wait. So what was it about years where they
53:52
suddenly became
53:54
not afraid? because I had a thousand subscribers. I was making a hundred thousand dollars a year. It was
53:56
that they thought that before they knew that
53:58
you were secure, that you might just
54:00
stop writing, and then they would have paid you, and then they
54:02
would get nothing
54:04
for it. right. I think that that was really the case. And so
54:06
then since then, you've never had a growth
54:08
explosion like that since that twenty five percent
54:10
growth in
54:12
one day. but it's been sort of slow and steady since
54:14
that point. That accidental
54:16
email hack, how do you think about it?
54:18
Now you say you still think
54:20
about Stuttgart, primarily as a website. But how do you think about email now?
54:22
Like, obviously, it's a big part of what
54:24
you do? Yeah. Well, it
54:26
is, but It's
54:28
not. I mean, I think if you see the stuff that I've
54:30
done, for example,
54:32
now you can consume the
54:33
daily update via
54:36
a podcast. or my articles via
54:38
podcast. I think that's actually very
54:40
much in line with my
54:42
vision and view of trajectory,
54:44
which is This is a publication that I write, and I will make it
54:46
as easy for you to consume in the way you
54:48
wanna consume as possible.
54:50
And back in twenty fourteen, that meant
54:54
sending emails. And certainly, there's all the advantage of emails that
54:56
everyone's talked about. It's a fee that everyone checks
54:58
daily. You don't need permission to get
55:00
into there. There's
55:02
no algorithm. you know, it's funny
55:04
because I think once that came
55:06
along, it became overwhelming and people started
55:08
setting up rules to send all
55:09
their emails to a folder, which they never
55:11
checked. your packets are the same problem, a self
55:13
imposed algorithm, which is where podcast is
55:15
great. It's the same idea. It's a feed
55:17
that people check that, I
55:20
mean, it works through polling, but to people, it
55:22
feels like push. And that's certainly
55:23
a great thing from an independent
55:25
publisher perspective. And I think in line
55:27
with my vision. I mean, one of
55:29
my criticisms of Substack was I think they were too
55:31
email centric to start. Again, I think the web
55:34
experience is important. It's super important
55:36
for growth. because that's where people find out about stuff is often via social
55:38
media and sharing links.
55:40
And email is a tactic. It's not
55:42
a strategy. And
55:44
at the end of the day, the strategy is about
55:46
differentiation. It's about consistency.
55:48
And I want
55:49
it to be
55:50
easy to use and fit in your life
55:53
and email was a way to do that. Again,
55:54
sort of backed into now podcast and
55:57
way to do that. You can
55:58
get structured articles on all the SMS, which gives
56:00
you a link and you can read on the website
56:02
obviously have always had RSS. And I just wanna make it as easy for you
56:04
to access the content that you're paying for as
56:06
possible because that sort of makes
56:09
old and going. you know, you
56:12
about this model is
56:14
number one from a writer perspective. All
56:16
I really need to worry about is
56:20
keeping
56:20
my subscribers happy. And so, like, you know, should this
56:22
be
56:22
a free article? Should it be paid? Well, you know
56:24
what? If it should have been free, but it made
56:27
it paid, That's fine. subscribers feel like they got a great article that they paid
56:29
for and they feel good about that. And that
56:31
recurring revenue is really really powerful.
56:34
And then also, that's my
56:36
marketing channel. them telling other people about that this is good and you should sort
56:38
of check it out. And one of the things we do with
56:40
Passport is a big wave
56:42
of stuff spreads is people
56:44
forwarding emails. and Passport is
56:46
the technology infrastructure you've built
56:48
to enable. Other people
56:50
to build stratigraphy like experiences if they're
56:52
whitelisted by you, how does that work? No.
56:54
It's just my stuff for now. I mean, obviously, we would love to make it broadly available, but
56:56
it's not all finished yet and there's
56:59
customer support issues and things that
57:01
would entail that. But something
57:03
that I would like to do. But one of the things
57:05
we did with that is I winked to myself a
57:07
lot, which people make fun of and deserve really
57:09
so because it's kind of a a writing joke between me
57:11
and my readers. But a way
57:13
I think about trajectory is it's a live thing. It's an ongoing
57:16
sort of journal of my
57:18
attempt to understand the world, to
57:20
understand technology. And
57:22
sometimes I was right about something, and I was always fun to
57:24
point back and say, I was right. Sometimes I was wrong. And
57:26
it's like, why was that wrong? What mistakes that
57:30
I make? sometimes there's a trend and it's like, well, this happened back then and
57:32
this happened here. And then, again, I think
57:34
about it as being sort of a live thing and so
57:36
waking back to myself is a way to
57:38
do that. it's also a
57:40
great way to trigger that
57:42
second article sensation where you read an
57:44
article and then there's a link right there and you go
57:46
read another article, well, that was also really good and
57:48
then there's YZ. So one of the things that I wanna do with Passport was how
57:50
can I really leverage email forwarding to
57:52
accomplish this, especially when I'm looking to
57:56
paint articles? So, like, one of
57:58
the things we do there is every link to myself
58:00
in a
58:02
trajectory
58:02
email is
58:03
tokenized, and that token goes one link deep.
58:06
So even if the emails forwarded to you,
58:08
you can go read old stuff. So you get
58:10
that second article sensation, if you're in that second
58:12
article when you click another link, now you're gonna paywall.
58:14
And then you're gonna realize, oh, I have to pay to get
58:16
this. But there's
58:17
all these little bits
58:20
pieces that
58:20
the subscription model makes possible,
58:22
which is really leaning into just serving your customers,
58:24
trying to make it as good of an experience
58:28
for them as it can be. And the other really important
58:30
about subscriptions is there is
58:32
always debates about, oh, people are just right for
58:34
clicks and blah
58:36
blah blah. And then the publishers or the editors that come
58:38
back say, no. We don't even show our
58:40
writers their quick numbers. Like, they don't know. All all
58:42
we want them to do is
58:44
write content. And I think that's
58:46
so foolish because everyone wants feedback. They want
58:48
affirmation. They wanna know, did I do
58:51
a good job? And if you don't even get access to your quick
58:53
numbers, where do you go? You to Twitter. Like, what are people saying about
58:55
me? Right? You will find a way to figure out
58:57
if you've resonated
59:00
or not. Right? And and who's on Twitter? The
59:02
biggest loud mails. Right? I'm one of the things I
59:04
I word on Twitter about Twitter
59:06
very early
59:08
on was I would be very engaged on Twitter.
59:10
I'd be responding to people. And it's only that occurred to me. I'm responding to the same ten
59:12
people or thirty people or whatever it is.
59:14
Right? So true. Yeah. you're
59:17
engaging with point two percent of your
59:19
audience. Right. And so that's a fixed
59:22
number. And meanwhile,
59:22
I have a feedback mechanism which
59:24
is my subscriber number which is going up, which
59:26
means I know
59:27
there's a huge number of people
59:29
that like and appreciate and are sharing.
59:31
And it's all dark matter. I don't know who they are. I don't
59:33
know where they are, but they like
59:36
what I'm doing. and I need
59:38
to anchor on that. Not worry about
59:40
what the Wild House of Twitter are saying. And so
59:42
subscriptions give this
59:44
feedback mechanism as feedback mechanism for people that will never email you, that will
59:46
never contact you, but they will pull out their credit
59:48
card and they'll give you money. And
59:50
I think that's
59:52
so valuable and so
59:54
positive from an incentive structure. Now there's
59:56
a downside where subscriptions
59:58
are niche. Right? And I do
1:00:00
think you tend up. People will
1:00:02
if they dislike you, they could stop paying. And there a worry and a
1:00:04
challenge that you're not like tendering to your audience
1:00:07
or whatever it might be. I
1:00:09
think I'm kind of lucky because I was so
1:00:12
early, because I was the only person in the
1:00:14
space doing this sort
1:00:16
of stuff. I got
1:00:17
to a large enough audience size where I could not care.
1:00:19
And if someone wants to cancel
1:00:20
the description, I don't care. In fact, if someone
1:00:23
comes back and if they are
1:00:26
just over the top rude or disrespectful or insinuating things
1:00:28
about me, I will not just cancel their subscription. I
1:00:30
will refund them their entire purchase and say,
1:00:32
please just never come isn't
1:00:34
the nice thing about a approachable price point. You're actually, your dollars
1:00:37
are insignificant to me because my number
1:00:39
of subscribers is so much
1:00:42
larger it's very different than, like, if Apple decides to stop advertising on Twitter, it's
1:00:44
like, oh, crap. 0000 I'm so sorry. I'm
1:00:46
so sorry to come back. Yep. That's exactly
1:00:49
right. And again, would I be able to
1:00:51
do that when I had a very small number of subscribers? No. But I mean, it was easier
1:00:53
back then because, I
1:00:54
mean, I started in twenty
1:00:56
thirteen and the
1:00:59
entire opinion of the tech press and Wall Street is that Apple was doomed because, of
1:01:01
course, we'll grab the windows mac blah blah blah. So
1:01:03
I had the lowest hanging fruit in the
1:01:05
world to sort of thick.
1:01:07
Right? It'd be like, no, I don't think Apple's do them, to be
1:01:10
honest. You know, and that's why I wrote that white paint
1:01:12
Christian was wrong because all the disruption people were
1:01:14
like, yo, apples for
1:01:16
sure screwed. unless they go down market. I'm like, no. I don't think they need to go down
1:01:18
market. And so that was great
1:01:20
because if Dalko is writing anything
1:01:22
that controversial, And
1:01:24
so I didn't even run into any issues where people would
1:01:26
try to hold over me. And by the
1:01:29
time I got into issues where
1:01:31
that mattered, I would large enough
1:01:33
that it didn't matter. I was fortunate in that regard.
1:01:35
For our second sponsor of
1:01:36
this episode and all of
1:01:39
season eleven, thank you for
1:01:42
a great season. We have one of our very favorite
1:01:44
longtime members of the acquired
1:01:47
family, pilot dot com. dot com
1:01:49
pilot sets
1:01:49
up and operates the entire
1:01:52
financial stack for startups and growing
1:01:54
companies. That means your finance,
1:01:56
your accounting,
1:01:58
your tax, even higher level, CFO like services, like investor
1:01:59
reporting, all of which,
1:02:02
as we've talked about, all season, you
1:02:04
would either
1:02:06
be hiring a big four, not a big five accounting firm
1:02:08
to do for you or building out your own finance
1:02:10
team or doing it yourself
1:02:13
as a founder. none of which are
1:02:15
good decisions or good leverage on your time and resources as
1:02:17
a startup. Pilot is a much
1:02:20
better decision they take literally all
1:02:22
the headache
1:02:24
of finance and accounting off of your plate as a founder
1:02:26
and a startup team, and they provide
1:02:28
you with their own team of experts
1:02:31
actual financial accounting experts, and
1:02:34
technology to set
1:02:34
up your financial infrastructure and run it
1:02:37
right. Now we've talked about this a lot on
1:02:39
this show, but it really is worth talking
1:02:41
about this defining characteristic of pilot. They integrate with the APIs
1:02:43
of all the modern start
1:02:45
upgrade platforms like
1:02:48
Stripe and Brex Gusto, Shopify Square, etcetera. But
1:02:50
most importantly, the next five
1:02:52
that come after that, they know
1:02:54
how
1:02:55
to make that work seamlessly
1:02:57
with your company's books so you can always use the best tools
1:02:59
available to you and trust that
1:03:01
your accounting bookkeeping
1:03:04
provider will have the ability to keep
1:03:06
up. Stripe itself is an investor in the company alongside
1:03:08
Sequoia Index and Bezos Expeditions.
1:03:13
There are no other accounting firms in the
1:03:15
world that can say that. So if
1:03:17
you are still using QuickBooks on your own
1:03:19
as a founder or paying
1:03:22
someone to Integrate
1:03:22
a bunch of stuff on your behalf or paying an old school accounting firm,
1:03:24
head on over to pilot dot com
1:03:26
slash acquired, make your life easier,
1:03:30
Thanks
1:03:30
to the pilot co founders, Waseem, Jessica, and Jeff,
1:03:33
for providing this awesome twenty percent
1:03:35
discount for the first six
1:03:37
months of service. if you go
1:03:39
to pilot dot com slash acquired. Thanks,
1:03:41
pilot. Thank you, pilot. Can you
1:03:42
share, like, what does the shape of the
1:03:45
curve look like in terms of
1:03:47
if you graph it on the x axis is time
1:03:50
and on the y axis is number of
1:03:52
subscribers from the start up
1:03:54
Stryker today. We talked about going
1:03:56
up to a thousand and then step
1:03:58
Even if it's like relatively linear,
1:03:59
has it sort of tapered off over
1:04:02
time? Like, are
1:04:04
you saturating the
1:04:06
world of tech people
1:04:08
who care about business strategy?
1:04:10
Yeah. It's interesting
1:04:10
because there's been different periods where it actually
1:04:12
I thought it was started taper and that actually accelerated.
1:04:15
And this is probably twenty eighteen, twenty nineteen around there. Over the
1:04:17
last couple of years, it has tapered off.
1:04:19
It's not quite what
1:04:22
it was. which is fine in terms of the growth. It's tapered
1:04:24
off. I paid subscription revenue. If it stays
1:04:26
flat, you're in great shape. Yeah. Well, and
1:04:28
then also I
1:04:30
raise prices a couple years ago, which
1:04:32
is the only time I've raised prices. And I still kind of mildly regret it. It's funny because
1:04:34
everyone's like always your prices too low,
1:04:36
you should raise prices. that's
1:04:40
definitely the absolute easiest way to increase
1:04:42
revenue. I
1:04:42
think the easiest thing for
1:04:45
me to do would be to just keep doing
1:04:47
what I'm doing and just raise the price
1:04:49
a dollar a year every year. And for sure, my revenue
1:04:51
increase would very much
1:04:53
outweigh
1:04:53
my churn. I honestly,
1:04:55
the biggest change that happened when I did raise prices
1:04:58
was it didn't affect growth at all,
1:05:00
except that I had a lot more people switched to annual
1:05:02
subscriptions because I, like, preannounce is, like, hey, if you wanna
1:05:04
get a year at the old price, which is great because I
1:05:06
want people and I know subscriptions as we sort of
1:05:08
discussed at the beginning. So I
1:05:09
could do that. I think
1:05:11
my view on it
1:05:13
has always been I think on the Internet in
1:05:15
general, the markets are much larger than people think. I thought it was much larger than all the folks
1:05:17
that are critical of my model thought, and it's been larger
1:05:20
than I
1:05:22
thought. all along. Right? I thought it would take five years to build or to watch as my business.
1:05:24
It took a year. And I would
1:05:26
rather explore the edges of that
1:05:28
market how big it is as
1:05:30
opposed to sort of maximizing my revenue too soon. And so
1:05:32
that's sort of number
1:05:33
one. Number two is
1:05:35
I do wanna have an impact. I
1:05:37
do want people to read
1:05:40
by deaf admission, the larger your subscriber base, the
1:05:42
more your sort of marginal customer
1:05:44
is not aware of you, is not immersed in the
1:05:46
space. Their willingness to pay is going to
1:05:48
be lower. And so when
1:05:50
you start raising prices, you're gonna hit a
1:05:52
wall very fast, I think, as far as subscriber growth,
1:05:54
sort of by definition. That's sort of
1:05:55
number two. the sort
1:05:57
of shift over the last
1:05:59
couple years has been, well,
1:06:01
is this it? I've reached my natural base or whatever
1:06:03
it might be, which is again, massively
1:06:06
larger than I ever expected. I just
1:06:08
wanted to pay the bills. Yeah.
1:06:10
Like tens of thousands of people
1:06:12
that pay you to read
1:06:14
your work and millions of people that, I'm sure,
1:06:16
come to your website and read the free product. Like,
1:06:18
there is a real market
1:06:20
for this. definitely push back on the
1:06:22
millions. I think it's interesting
1:06:24
because I do think this is an area
1:06:26
where I do feel I got there in time. I think the
1:06:28
ability to spread on social
1:06:30
media on Twitter has really diminished. You know, Twitter famously really started
1:06:32
devaluing tweets with links
1:06:34
a few years ago. which
1:06:37
I think has made the product much worse. I mean, that's how we end up
1:06:39
with all these threads because those spread
1:06:42
more. threads. Oh my god. Yeah.
1:06:44
Which all should be blog posts So I think that
1:06:46
has hurt the other thing that has hurt my
1:06:48
growth, frankly, is releases
1:06:50
off as a podcast. So
1:06:52
my subscribers love it. half
1:06:54
of my subscribers now listen instead of read, which is That's how I consume
1:06:57
every single post. Yeah. And then the
1:06:59
other good thing is people would churn
1:07:01
because emails would build up. and
1:07:03
they look at their inbox, they'd be like twenty
1:07:05
hundred emails, and
1:07:06
they're like, why am I paying
1:07:08
for this? Whereas an eight minute podcast
1:07:11
just feels I can knock this off real quick.
1:07:11
So it's been very good for
1:07:14
churn, but the problem is people don't
1:07:16
share podcasts.
1:07:18
Right? You listen to a podcast. You go to the next podcast. You do XYZ
1:07:20
Whereas if you're reading something, it's super easy
1:07:22
to tweet it or to forward the email. Click forward.
1:07:24
Yeah. Yeah. So I do think that
1:07:27
might be the biggest issue on why my growth slowed down, to be
1:07:29
honest, which is fine. It is what it is. And you're
1:07:32
obviously leaning more into podcasts. Right? I
1:07:34
did face a decision this year,
1:07:36
which is growth is somewhat
1:07:38
leveled off. Do I raise prices? And
1:07:40
just sort of be happy with the base that I
1:07:42
have? And I
1:07:42
figured it'd be
1:07:43
worth a shot to see if
1:07:45
I can restart growth. in this
1:07:47
case, by sort of working to increase and
1:07:50
broaden the value
1:07:51
of a secondary subscription and
1:07:53
not just my own
1:07:56
content. So watch Dithering with Groover. Obviously, he's now my podcast partner a
1:07:58
couple years ago. That was like an add on, so you
1:07:59
pay extra and you get dithering. And that
1:08:02
was really on the let me make more money
1:08:04
for my best subscribers. That's a way to raise prices
1:08:06
on my best subscribers without explicitly raising
1:08:08
prices. Right? Who's
1:08:10
lunching, dithering, business decision
1:08:12
or just a this is gonna be a fun thing with
1:08:14
a good friend or both? Oh, both. And then also I I
1:08:16
had this technology for pay podcast.
1:08:18
That was a way to do it. And so,
1:08:20
yeah, we watched it. It it's been successful. I mean, it
1:08:22
has you know, I think, like, from day one, we have, like, ten thousand subscribers.
1:08:26
And then it's sort of level test. It's it's kinda like where we stuck. We didn't churn at
1:08:29
all. But, like, how do you grow paid podcast? It's
1:08:31
we haven't done anything around free
1:08:34
episodes and marketing. We will but that was a de facto price raise of my
1:08:36
best customers. And when I wanted
1:08:38
to
1:08:38
sort of explore this space, it's
1:08:40
like, well, is my market as
1:08:42
big as it can be? And
1:08:45
Is there a way to sort of
1:08:47
go broader? I figured out the newsletter thing. Like, I think subscriptions make tons
1:08:50
of sense for podcasts.
1:08:52
because people get so attached to a podcast, they get so attached to
1:08:54
a particular host, and it becomes such a part of their routine to
1:08:58
a much greater extent than
1:09:00
reading, which is sort of more of an affirmative choice. I have
1:09:02
to go read this where it's like, oh, this is my podcast player. I guess I'll listen to this next. And so I think it's
1:09:05
a great product
1:09:08
for subscription. but it's
1:09:10
totally unclear how do you actually build a podcast, particularly a paid podcast. It's really hard
1:09:12
to get people to
1:09:14
start to listen to it.
1:09:17
And so in general, I think writers tend to
1:09:19
succeed with podcasts because writing is a good marketing medium. Again, it
1:09:21
spreads easily. You just pass
1:09:24
the URL people
1:09:26
become familiar, like, oh, yeah, I'll try your podcast because
1:09:28
I like you as a writer. But if you're just
1:09:30
a pure podcast, how do you sort of grow?
1:09:32
And so what
1:09:33
I've decided to do is gathering is no longer gonna be an add on. It's gonna
1:09:35
be a part of a stratospheric subscription. So now again, your
1:09:38
stratospheric is worth more.
1:09:41
you have this sort of extra piece and you got
1:09:43
it for quote unquote free. So you don't sell it separately at all. Well, it is available separately. Again, it's cold with John
1:09:45
and still don't know if your subscribers maybe don't like
1:09:47
me because they wanna Drive
1:09:51
they can. But it's no longer as an add on Shrekori. If
1:09:53
you are Shrek transcriber, you get dithering. It's
1:09:55
part of your subscription.
1:09:56
Watch the new
1:09:57
podcast called Shrek
1:09:59
Tech
1:09:59
that, again, there's
1:10:00
both free episodes and paid episodes. It's available to
1:10:02
all structure subscribers. So your
1:10:03
subscription is worth more. And so
1:10:05
people are
1:10:06
sort of opting in just to
1:10:08
make the point clear, because I've listened to every single
1:10:11
episode of Dithering, including the backhaul, like, before you launched. I'd started resale listening to Sharp Tech and,
1:10:13
of course, Sharp China, which you're
1:10:15
about to talk about. it seems
1:10:17
to me that Sharp Tech and Dithering is sort of like, well, I
1:10:19
get Ben and I get Ben's thoughts and I get a lot
1:10:21
of the same thoughts and I basically am
1:10:23
picking whether I like Andrew
1:10:26
or John Better, and I'm also picking
1:10:28
the format and duration of the
1:10:30
episode. Yeah. Right now, the
1:10:33
bundle is kind of a joke because it's
1:10:35
mostly just me. Right? And different flavors have been. I only have
1:10:37
so many opinions. I have a lot of opinions, but I only have
1:10:39
so many. So it's not in
1:10:43
its long term place now for sure. And it's
1:10:45
very tricky
1:10:46
because there is an extent
1:10:48
to your point about
1:10:50
if you release a bad episode of Acquired, you
1:10:52
know, do you worry about churn?
1:10:54
Am I diluting what trajectory is to
1:10:56
someone? It's no longer just an email. And
1:10:58
now I'm reintroducing the possibility that people feel overwhelmed with content. And then they also feel like, oh, Ben's repeating
1:11:02
himself
1:11:02
all the time.
1:11:05
I think there's a lot of underrated
1:11:07
risks in doing what I'm doing. But number one, I think WatchGuard China was
1:11:09
a good signal
1:11:12
to my audience of
1:11:14
what I do want to sort of accomplish
1:11:16
going forward. Well, that's a collaboration with Bill Bishop who writes Sinosysm
1:11:18
and it's the first introductory product that I'm not on. But
1:11:23
I think it's a important topic. Understanding China. I
1:11:25
think it's of a
1:11:27
similar tone and quality level
1:11:29
of what you should expect
1:11:32
from trajectory. Do I expect all strategic drivers
1:11:34
listed? No. Of course not. Some people like me specifically, and I'm not on there, so they're not gonna be interested. But
1:11:36
I think for
1:11:39
a lot of folks, it's like, well, I
1:11:41
keep hearing lots of news from China. Where do I go? I'm
1:11:43
already a secondary subscriber. There's this other product here.
1:11:46
And if they become a regular I've now decreased my
1:11:49
churn that much more for that sort of
1:11:51
person. And I think, again, a big part
1:11:53
of subscription is just churn management. It's
1:11:55
like, I wanna make sure
1:11:57
this is a one time subscriber and you get that recurring revenue over time.
1:11:59
As you might have imagined, there are plans for other podcasts
1:12:01
sort of be added in
1:12:03
the long run. And
1:12:06
we'll see what happens. I mean, growth has, I think,
1:12:08
picked up over the last couple of
1:12:10
months, so early indicators are that
1:12:13
it's working. where I'm increasing the value and will make
1:12:15
up for the cost. Would I make as much if I just raise the
1:12:17
prices and didn't bring on the additional
1:12:19
cost of paying for Andrew
1:12:23
and other broadcasters and he's on those lines? Probably not. Again, the
1:12:25
revenue maximizing thing for me would be
1:12:27
stayed with them before and just write
1:12:29
your check right now. And it's nice
1:12:31
and simple. Yeah. But number one,
1:12:33
I've always been worried about getting stale. I know I have critics I hear about on Twitter who think I already
1:12:35
am, but they're point two percent
1:12:39
of your audience. Well, who knows? They're very hard
1:12:41
on doing it. Right? Number two, one of the things I'm so proud about was Shookery. I always have most of them
1:12:43
known for Shookery, but I'm so proud of
1:12:46
the business model and that it exists.
1:12:48
And the
1:12:50
fact that sub stack is out there,
1:12:52
and there's hundreds or thousands of ordinary people making their living, doing this.
1:12:54
And I feel responsibility for that in a very positive way.
1:12:59
And I really wanna get this working
1:13:01
for podcasts. I think there's a
1:13:03
similar opportunity. And so helping
1:13:05
to help pioneer that is fun.
1:13:07
building software is fun even though it's cost me a lot
1:13:09
of money as you might have imagined. Producing
1:13:11
new podcasts is fun. Figure
1:13:14
out this bundle thing is fun. how do you cross link stuff? Like, you if
1:13:16
you listen to the Checkatrade episode because we
1:13:18
know who you are because it's a unique
1:13:20
feed to you, you
1:13:22
could just add another without having
1:13:24
to sign in or do anything,
1:13:26
like, really smoothing out this experience of cross promotion. It's
1:13:30
fun. And I think I'm fortunate enough that trajectory has
1:13:32
been successful enough, that I can make
1:13:34
choices that at least in the short
1:13:38
term are more optimal for figuring stuff
1:13:40
out and having fun even if
1:13:42
it's not revenue optimal. And obviously,
1:13:46
there's upside maybe it'll end up being a huge thing and and I'll
1:13:48
make much more money than I ever would have just
1:13:50
doing trajectory. But I do feel very fortunate that
1:13:53
I can make that choice in the short to medium
1:13:55
term. nice thing about having no outside shareholders. You're your own little Zuckerberg
1:13:57
over there ruling by Fiat on whatever
1:13:59
sounds most fun to you.
1:14:01
You're no activist shareholder. No.
1:14:04
For sure. I actually thought
1:14:06
about building a sub stack a few years before they watched, actually had a team together, and it ended up falling apart.
1:14:09
But my
1:14:12
big hesitation was always
1:14:14
I just wasn't sure if it would be venture scale. And from my perspective, if
1:14:19
you have the stamina capability to
1:14:22
produce regularly. Subscriptions are amazing ability to sort of bootstrap. And,
1:14:24
again, you're asking people to
1:14:26
pay for the regular pressure content.
1:14:29
It's actually a very straightforward transaction and can be
1:14:31
sort of very sustainable. How many of those are venture
1:14:34
scale businesses? I think
1:14:37
maybe not as many. And you're right. It comes with a bunch of
1:14:39
compromises. It comes with trade offs. You have to optimize in certain ways. And I think venture
1:14:41
is a phenomenal thing. I think
1:14:44
I'm not antiventure
1:14:47
at all, just anti venture for me first. I
1:14:49
don't think it would make much sense. I
1:14:51
like the lifestyle, so lifestyle business
1:14:53
for me. You were talking to
1:14:55
two venture capitalists, who have a business
1:14:57
which is structured the same as yours. So I love that we've had this part of the conversation because I was really
1:14:59
curious, I mean, selfishly, for
1:15:04
us, I'm curious your
1:15:06
thoughts on there are corporations that have large numbers of people that
1:15:08
are make
1:15:09
products that
1:15:12
are scalable. And then at
1:15:14
the opposite end of the spectrum, there's being talent. a solo business, but business is being other people's
1:15:20
productions. but the internet has enabled this
1:15:22
whole new class of stuff, like stratigraphy, like acquired, like not boring, like, what
1:15:24
have you? How big do you think this
1:15:26
third class of whatever it is we are
1:15:28
can be?
1:15:30
Is that part of what this experiment is? For sure.
1:15:33
Like, I'm not passionate about like
1:15:35
meta's business prospect. I definitely have
1:15:37
takes on it. I think that's the other
1:15:39
company I've mostly been pretty right about over
1:15:41
the years along with Microsoft or probably my two
1:15:43
long standing best calls. but I'm
1:15:45
not, like, waking up in the morning, killing myself, like, with their stock goes up or down, I
1:15:47
don't really care. You're not gonna go work it better.
1:15:50
I'm not gonna work it
1:15:52
better. I think that's a pretty safe
1:15:54
I mean, if that wants to buy me for a hundred million dollars, I guess, I'll earn out, but You'll rest Yeah. I
1:15:58
will rest
1:15:59
invest. Yes. I
1:16:01
mean, Mark, you're listening. But what does get me super excited
1:16:03
I am really passionate about is this completely
1:16:07
new arena of possibilities and I
1:16:09
think new jobs that are made possible by the Internet. And actually one of the reasons I actually
1:16:12
am in
1:16:16
general favorable towards meta is I think
1:16:18
they're an essential component in the broader ecosystem of niche businesses. Because
1:16:20
if the entire
1:16:23
world is your audience, how
1:16:25
does the world find out about you? I'm lucky because
1:16:27
I produce content that people want to talk about. And so they share it
1:16:30
for free on Twitter.
1:16:32
Right? if someone makes, like, a really cool
1:16:34
new piece of clothing or accessory, like, people aren't gonna talk about that on Twitter, like, you need a way
1:16:36
to advertise. And I think that's what Facebook
1:16:38
advertising has always been the best at.
1:16:42
and I think is really valuable and so
1:16:44
I am in general a big defender
1:16:46
of that because it actually is in
1:16:48
line with what I am personally very
1:16:50
passionate about which are the economic opportunities made possible
1:16:52
by the Internet. What the Internet does
1:16:55
do an industry after industry and
1:16:57
this applies to creative talent as
1:16:59
much as anything else, is number one,
1:17:01
there's always a barbell effect. You're either very large or very small. You're very large, you
1:17:03
get scale, you get aggregation effects, and
1:17:07
then you can make massive investments or acquisitions. No. You can immediately
1:17:09
feed that to hundreds of millions or billions
1:17:11
of people and
1:17:14
get a payoff. or you take advantage of the Internet and open
1:17:16
source software or wherever it might be
1:17:18
or different platforms and your cost structures are
1:17:20
basically zero and you only need a thousand
1:17:22
true fans like the thousand true fans was
1:17:25
one hundred percent true. That was by guiding vision and ended up
1:17:27
being the case. And so if you're stuck in the middle, which all the
1:17:29
old publications were stuck in
1:17:31
the middle, they weren't big
1:17:34
enough to get aggregation effects. Their cost structures were way too large to handle a fractured audience. They were all in
1:17:36
trouble and got a lot of traction about
1:17:38
saying they were doomed and then they were
1:17:40
doomed. Like
1:17:43
you're saying about, that's why they hate social media. But if you're on the low
1:17:45
end of the barbell, like, oh, it's like, you love social
1:17:47
media. No, that's right. It's an amazing asset.
1:17:49
So that's number one. The owner
1:17:51
has its effects. Number two, the Internet
1:17:54
has winner take all effects in specific markets. So it's
1:17:56
gonna
1:17:57
be hard for someone to
1:17:59
be a
1:17:59
generalist tech and media
1:18:02
analyst and do what
1:18:04
I do. It's possible. I mean,
1:18:06
there's people that are very good. but
1:18:08
just because I got there first. But
1:18:10
that doesn't mean there's
1:18:11
only one analyst role available.
1:18:13
There's a guy
1:18:15
Neil Seibert, I think. Well, it's like an research department in a bank. Yeah. You
1:18:17
can write about Apple. And he very much
1:18:20
mimics his business after me. He was
1:18:22
pretty early too, so he actually did a
1:18:24
lot stuff himself. He's done a really great
1:18:26
job. He's been an independent Apple analyst for eight years or something like that,
1:18:28
which is fine.
1:18:31
He read most Apple every single day.
1:18:33
Believe me, there are people that want to hear about
1:18:35
Apple every single day. And so that's an example where it
1:18:37
seems like we're doing
1:18:39
the same thing but we're in different
1:18:41
markets. This is what the internet makes possible. Is the key to success on
1:18:43
the internet is you wanna be
1:18:45
the biggest fish in the
1:18:48
pond, but The success
1:18:50
metric is not competing with other fish. It's finding your own pond. The opportunity is broad. It's not
1:18:55
deep. Like in each pond, there's probably
1:18:58
only be one or two fisces that survive. Right? But there's an infinite number of potential ponds.
1:19:00
You can define
1:19:03
your pond in very specific
1:19:06
parameters. I'm still surprised there is not someone that runs from Amazon every day. Like, it's a massive sprawling business. There's so
1:19:08
many things to write about. I
1:19:10
think you'd probably send you all Google
1:19:14
Casey Newton basically writes about social media every single day.
1:19:16
It's been Christmas for him over there because of
1:19:18
the Twitter stuff, but that's fine. I
1:19:21
feel guilty about Twitter too often. It's very
1:19:23
exhausting I cover lots of other stuff. There's other things to talk about. And so I'm writing
1:19:25
about it more than I want to, particularly yesterday.
1:19:27
I really didn't want
1:19:30
it on Twitter. Obviously, Twitter take on Apple is gonna be everyone's
1:19:32
been anticipating it. Right? It's the super
1:19:34
bowl of Ben Thompson fans. I kinda
1:19:37
have to write about it. Right? Right. But
1:19:39
Casey, like, that's the expectation is that he read about social
1:19:41
media. So he's read about Twitter every single day for,
1:19:43
like, a month, and that's exactly what people
1:19:45
want. And I'm glad he exists from
1:19:47
my perspective. Hey, if you want Twitter
1:19:49
coverage every single day, go re case of you. That's a great thing for my perspective. And I think that sort of
1:19:51
defines this market. And so,
1:19:54
number
1:19:55
one, I'm really
1:19:56
one i'm really proud
1:19:58
that this business model for newsletters exists. I think it should exist for podcasts.
1:19:59
And I think
1:20:02
part of that is
1:20:04
just figuring out the mechanics
1:20:06
of it. Like, how do you market? How do you grow stuff? I haven't done a good job marketing my paid podcast, I would say,
1:20:09
other than
1:20:12
leveraging Checkatrade. but that's the
1:20:14
number one we want to work on. Obviously, there needs to be should there be some video stuff, like, there should be, you what's
1:20:16
the ratio of free stuff to
1:20:18
clips, to snippets, things on those lines?
1:20:23
we're actually experimenting that more with Start China and Start Tech, where
1:20:25
even if you're on the free feed or you're getting
1:20:27
more clips and snippets. Also,
1:20:30
I think that the technological aspect is important. It's funny,
1:20:32
even free podcasts. It's like, oh, we're gonna have
1:20:34
a guest on. Go as to his podcast. Go
1:20:36
to your podcast player, search for
1:20:38
XYZ and it's like,
1:20:41
offloading this huge number of steps to the user, hoping they follow through and
1:20:43
do it, whereas we've built something where go
1:20:46
to your show notes, click
1:20:49
the link. Boom. It's your podcast player. Like, I
1:20:51
think that's something that's gonna be important to sort of, yo, the sharing stuff is really tricky
1:20:53
because people don't share podcasts
1:20:56
naturally. So I
1:20:58
wanna figure that stuff out. And it's gratifying.
1:21:01
Again, not just because it's fun and
1:21:03
new and different than my day
1:21:05
job, but also that is my passion.
1:21:07
My passion is these sort of businesses, the Internet not the
1:21:10
internet
1:21:11
just destroying
1:21:13
old business models, but making new ones possible, and not just the
1:21:15
big guys, but like for individuals as well. And so
1:21:19
at very fortunate that
1:21:20
I sort of get to
1:21:22
do both.
1:21:22
Yeah. I wanna switch
1:21:24
gears and ask
1:21:25
you about aggregation theory. There's
1:21:27
a multipronged quest in
1:21:30
here. So the business that you're running is far, far,
1:21:31
far superior to a thing that I'm sure you've been asked a
1:21:33
zillion times about why don't you
1:21:35
write a book? You
1:21:38
make a hundred and fifty dollars a year per subscription, at least
1:21:40
that's what I pay for the full bundle, I think.
1:21:42
No. You got a refund and a discount to a
1:21:44
hundred and twenty. Alright. Then I pay a hundred
1:21:46
and twenty dollars a year. Yeah. Dithering is now wrapped
1:21:48
into a trajectory subscription. So, yes, I should be clear for doing
1:21:50
listening. You got a credit on your subscription. we
1:21:54
did not issue refunds, but you did get a
1:21:56
credit. But that is to say, I have no idea
1:21:58
what I'm paying, and there's some segment of your
1:22:00
listeners that are so wildly price insensitive about
1:22:02
what you do that it doesn't So it's probably hard
1:22:04
to price discriminate on them. Yeah. No. It should look to me for pricing
1:22:06
strategies because I know it's not optimal. But anyway, my quick math was
1:22:08
if you did a hundred and fifty dollars a year and
1:22:11
I know it's a hundred and twenty and
1:22:13
you assume some five year lifetime on customers, and you compare that to like what you
1:22:15
would make selling somebody a book once, which is
1:22:18
what? Twenty dollars times
1:22:21
seven percent goes to the author. It's literally
1:22:23
five hundred times more revenue to you to do what you do versus writing
1:22:25
a book. And on
1:22:27
the other hand, there's something
1:22:29
that you've become known for in aggregation theory and some other topics around the edges
1:22:31
of it that sort of
1:22:34
deserve a canonical work It's
1:22:37
sort of the modern Porter's Five Forces.
1:22:39
It deserves a canonical than something that's been edited
1:22:42
and revised and attached and
1:22:46
rethought through. So have you thought about
1:22:49
what form canonical Ben
1:22:51
Thompson topics would
1:22:53
take? Yeah. Ariation theory is the obvious one.
1:22:55
I time have book be seventeen. mean, I was
1:22:58
writing about the ideas of
1:23:02
variation theory from the very beginning. It's
1:23:04
hard to imagine, but back in twenty
1:23:06
thirteen, again, like, back then, people thought
1:23:08
Apple was doomed. people thought the Internet
1:23:10
was inherently decentralizing. Oh, it was not long after the Facebook IPO, and it was a
1:23:12
disaster. Yeah. Facebook was dope. Well,
1:23:14
then Microsoft was the other thing.
1:23:18
I was able to come out of the gate with really four takes
1:23:20
that were probably, I totally forgot
1:23:22
that. But one, Apple's not doomed.
1:23:25
They're actually gonna be doing very, very well.
1:23:27
Number two, the Internet is centralizing. It's not decentralizing.
1:23:29
Everyone's understanding the dynamics are
1:23:31
completely wrong. Again, today,
1:23:33
everyone understands that, but
1:23:36
it was believe it or not very controversial a
1:23:38
decade ago. Number three,
1:23:38
this is what Microsoft should do. Like, there actually is a
1:23:39
clear path forward. They're
1:23:42
not doomed to your relevance.
1:23:44
And
1:23:44
then number four, Facebook is way more dominant and valuable than
1:23:47
anyone
1:23:47
thinks. And so that was a again, like I
1:23:49
said, lots of low hanging fruit
1:23:51
to sort of pick. it's
1:23:53
easy. You're being self depreciated here, obviously, but I think a big part of the reason this is behind Ben's
1:23:55
question to you of, you know, it
1:23:59
deserves kinocoric. The reason
1:24:01
these are accepted clues and realities now is in large part due to your
1:24:03
work. It is a bit of humbling to be a
1:24:05
writer because I think people would understand
1:24:07
all those things Had
1:24:10
I not written it? I think at best, they understand
1:24:13
that maybe a few months or at best
1:24:15
a year or two before
1:24:17
it becomes common knowledge. and that's just the reality of
1:24:19
the game. The Edge you can provide is usually
1:24:21
measured in sort of months or years or a
1:24:24
very low number of years. I
1:24:26
do think irrigation theory should be a book. It was probably again more pertinent at the height of all these sorts of Maybe it's
1:24:28
still pertinent
1:24:32
now, but Number
1:24:33
one, now there's just a logistical issue I write every day into your point and make way more
1:24:35
money every day than I would writing a
1:24:38
book. So that's number one.
1:24:41
Number two, there's a fear
1:24:43
factor, which is people think I'm very
1:24:43
productive, but I in
1:24:46
reality have
1:24:47
daily deadlines that sperm
1:24:50
that productivity. In the absence of those
1:24:52
daily deadlines would terrify it be in terms of
1:24:54
a book. And number three, there's a second fear
1:24:56
factor, which is a book is frozen in
1:24:58
time. And if I were to have written an aggregation theory book a couple years ago,
1:25:01
I probably would have centered on Netflix. And
1:25:03
that ended up being wrong. It was
1:25:05
wrong in
1:25:06
another way that I wrote about
1:25:08
where Content
1:25:08
is super important. That's how you
1:25:10
break away from aggregators having high diversion content. And I had the balance wrong
1:25:13
between Netflix's
1:25:16
aggregation, effects versus the power
1:25:18
of content. And so I'm very fortunate to write that book. Whereas now, I was still wrong,
1:25:20
but I have the medium and
1:25:22
ability to go back and say, well,
1:25:26
I was wrong. This is where I was wrong. What does this
1:25:28
mean for Discovery Time Warner? What does this mean for
1:25:30
Disney? What does this mean for the other platforms?
1:25:32
So that
1:25:33
is a fear factor. And
1:25:35
then number four, I
1:25:36
think there's an aspect, yeah, maybe there's
1:25:38
a treatment, but trajectory is very much of the Internet. I think a nature of
1:25:41
the Internet is it's
1:25:43
transient. It's not permanent. Again, another thing
1:25:45
people got wrong. I think actually one of the biggest mistakes Twitter made, and it's understandable no one
1:25:47
could have seen it at the time, but Twitter should
1:25:50
have had disappearing tweets from day one. It
1:25:52
should have always
1:25:54
been just in the moment sort of social network. I
1:25:56
think it would be a much better product. I don't
1:25:58
see that they should still do it now. People
1:26:01
experience it and think of it as in the moment
1:26:03
product and then are stuck with this
1:26:06
archive that induces fear,
1:26:08
induces
1:26:08
ruin,
1:26:10
and it
1:26:11
reduces the facade of what it's like to be on Twitter.
1:26:13
That's a big reason why Twitter's not what it used to be
1:26:15
because now people are scared. Well, to
1:26:17
the aggregation theory point, I think I've heard you make
1:26:19
this argument before, but articulating it the way you have
1:26:21
over the years and revising it on
1:26:24
stratigraphy is
1:26:26
actually a better product. than if you were to, like, have written a book five
1:26:28
years ago. Oh, I so disagree, David. It makes it
1:26:30
really difficult to explain what it is to people. I'm
1:26:32
like, okay. If I'm on the board of
1:26:34
a company and I'm trying to, like,
1:26:36
explain to them that they need to reframe their
1:26:38
thinking and think about aggregation theory. If someone's like, what's that? I'm like, I actually don't
1:26:41
know what to
1:26:44
send you. Like, read these things in this order to
1:26:46
watch this guy contradict himself and say when he got wrong and then sort of take away what you think
1:26:48
the modern interpretation is.
1:26:50
I think you're both right.
1:26:52
there should be set in stone trees. Honestly, there's just
1:26:54
a matter of priority. I mean, I think, again, it
1:26:57
would be good
1:27:00
to share and point you this thing and go read this. To
1:27:02
be clear, I'm in the middle of massive self franchised right now for explaining why this is the case. So take everything
1:27:06
I'm saying with a grain of salt. And the other thing is I'm just having more fun
1:27:08
and more interested in building software
1:27:11
and figuring out podcasts and I
1:27:15
would like to think, oh, a book is just really about
1:27:17
my ego and putting it out there. So I'm not
1:27:19
that sort of person. I
1:27:21
just wanna give back a creative community. That's why I said,
1:27:23
I'm definitely in rationalization mode. You're definitely
1:27:25
rationalizing. But, no, I should write
1:27:27
a book. I just haven't. I don't know
1:27:30
if I ever will. but I agree that I
1:27:32
should have, by now, written a book. But
1:27:34
a last one last quick question
1:27:37
on aggregation theory. Obviously, you were building up
1:27:39
towards it even as you said at the time, I think.
1:27:41
Was it one or two years you were doing stratigraphy before
1:27:44
you wrote the first post? this
1:27:46
is actually an example of the
1:27:48
power of branding. I wrote articles that were basically
1:27:50
allegation theory, well before I wrote allegation
1:27:51
theory. But giving
1:27:54
it a term and pointing it is what
1:27:57
made it stick. And actually, some of
1:27:58
those articles that I wrote
1:27:59
before, there's one I think it was
1:28:02
called economic power in the age of abundance.
1:28:04
doesn't have the same ring to it. I think it's
1:28:06
a better articulation of our aviation theory in many respects. I wrote it in twenty
1:28:11
fourteen, I think. good ring to it? Number two
1:28:13
is one of my least read articles up to that
1:28:15
date. Just no one got it
1:28:17
or understood what I
1:28:19
was talking about. And it's a weird thing as
1:28:21
a writer. Like, there's things that I'm thinking about now or that I know I will write
1:28:24
about, but it's not the right time.
1:28:26
And there's something you just learn over time
1:28:28
where you
1:28:30
can be too early as a writer too. This kinda
1:28:32
goes back to, like, I didn't know how to communicate.
1:28:34
Right? Like, I remember I was in there's
1:28:37
some meeting at Microsoft, and we walked out and
1:28:39
my manager's like, Ben, you're the only person in
1:28:41
the room that actually understood the issue and what
1:28:43
we should do. And absolutely
1:28:46
everyone, no one understands talking about and people are kind of annoying to you.
1:28:48
And he's like, the problem is we want to
1:28:50
get to h, but everyone in that room
1:28:52
is
1:28:52
on a. And you cannot
1:28:54
talk about h. You have to
1:28:57
talk about b and then assist
1:28:59
JRD. It is JR. Yes. And it's funny because that applies to
1:29:03
writing online too. Number one, sometimes
1:29:05
I'm wrong, so it's good to wait or to be sort of sure. But number two, there's
1:29:07
an aspect where there's just the right time in place
1:29:10
for things. There's always stuff I'm sort of
1:29:12
thinking about And
1:29:15
one thing I've learned is it's definitely better to write about
1:29:17
something a current event, which is bad from
1:29:19
a book perspective because
1:29:21
I'll write an article that actually has some key insights,
1:29:23
but it's talking about some event that happened by twenty
1:29:25
fifteen, which no one remembers or cares about. But
1:29:27
in the nature of my business,
1:29:29
that is actually much better for helping people rock it, get it, spread
1:29:32
it, share it, those sorts of things. So there
1:29:34
isn't a bit where the incentives of my business
1:29:36
do work against timeless
1:29:38
pieces in a certain respect.
1:29:40
Okay.
1:29:40
So when you
1:29:41
hit publish though on aggregation theory, did you think it could
1:29:44
be what it became? I did
1:29:46
know. You know, I thought about the
1:29:48
name. obviously,
1:29:50
I was inspired by Clay Christensen disruption
1:29:53
theory. That was sort of one of the
1:29:55
things. But I had written a
1:29:57
number of articles going up to that that
1:29:59
we're clearly building to that point. So
1:30:02
there's one about Airbnb. There's one
1:30:04
about Netflix. There's one
1:30:06
about just websites and publications.
1:30:08
And then our Asian theory was sort of short and
1:30:10
sweet because it was basically distilling what was in those previous articles into
1:30:15
one thing. I felt very confident I had a thing and I wanted to
1:30:17
have a definitive piece that was doing what
1:30:19
it was and needed to have
1:30:21
a name.
1:30:22
So that's what that was. For
1:30:24
our last sponsor this episode, we are thinking our good
1:30:26
friends at tiny. Tiny as you know is the Berkshire Hathaway of the
1:30:29
Internet and has built and acquired
1:30:31
the world's premier collection of truly
1:30:34
wonderful internet businesses. By now,
1:30:37
you all know their story, Andrew
1:30:39
Wilkinson, started the really premier
1:30:41
design agency in the entire world, metal lab from his home in Victoria,
1:30:43
British Columbia, beautiful.
1:30:47
Victoria, British Columbia, and
1:30:50
they built the UIs for Slack, Coinbase,
1:30:52
Tinder, Headspace, Patreon. You name it
1:30:54
the best of the best.
1:30:57
And then he and his partners became
1:30:59
completely obsessed just like us with Berkshire Hathaway and Warren and Charlie and they
1:31:02
realized, wait a minute,
1:31:05
There are a whole bunch of companies out
1:31:07
there that are Internet companies who would be perfect Berkshire type but Warren and
1:31:09
Charlie aren't gonna buy
1:31:12
them because we spent
1:31:14
ten hours talking about, they don't get the but tiny can. So went they companies like dribble,
1:31:16
or Pixel Union, Creative
1:31:18
Market, eighty twenty, Girl Boss,
1:31:23
ARO Press among others. Businesses doing five million
1:31:25
or more in recurring revenue with thirty
1:31:27
to forty percent
1:31:30
operating margins but that otherwise were sort of stranded assets in
1:31:33
the technology industry because they either
1:31:35
couldn't raise venture capital
1:31:37
or even worse they did raise venture capital, and then it turned
1:31:40
out that there weren't M and A buyers for
1:31:42
them, and they couldn't get to a scale required
1:31:45
for an IPO. Yeah. We're
1:31:46
definitely in a situation now that the market
1:31:49
has shifted where ninety percent of
1:31:50
businesses already shouldn't raise venture capital,
1:31:53
but ninety percent of the businesses raised
1:31:55
venture capital probably shouldn't have either.
1:31:57
So what do you do
1:31:59
from here? Well, tiny has realized
1:32:01
they can fix this situation for
1:32:03
everybody If you sell the company to them, the VCs can get their
1:32:05
money back or whatever the appropriate amount of
1:32:07
their money back is,
1:32:10
founders can get to take control back and set up an incentive structure
1:32:12
with tiny that makes sense to keep growing
1:32:14
the business, to keep pouring their energy
1:32:17
into their life's work and still have enough control over
1:32:19
it with the right capital structure that actually
1:32:22
makes sense for the type of
1:32:24
business that they're in. They've
1:32:26
done this with a bunch of businesses already year. Venture companies over million thirty to percent operating
1:32:28
margins, or the potential for
1:32:30
that to quickly become the case.
1:32:35
And the most important thing is it removes the immense pressure on
1:32:37
the business to get to an exit.
1:32:39
An exit doesn't matter anymore.
1:32:41
Tiny is happy to own
1:32:44
these businesses in perpetuity. And, man, I
1:32:46
can tell you in my previous life as a professional venture capitalist and board
1:32:48
member of many
1:32:51
VC backed companies. There
1:32:53
are a lot of companies in VC
1:32:55
portfolios that fit this bill, and it's just heart wrenching
1:32:56
what to do with these companies
1:32:59
because they're making real products they
1:33:02
have customers, they have employees, but they're just
1:33:04
not going to get to an outcome that
1:33:06
makes sense for
1:33:07
a venture capital firm. And
1:33:09
so having an off ramp
1:33:11
like this is a
1:33:12
godsend. Well, if you're the founder of one of these
1:33:14
businesses, if you're the VC board member of one of these businesses, don't let
1:33:16
sunk cost
1:33:17
fallacy get the better
1:33:20
of everyone recognize what the business potential actually
1:33:22
is now that you have seen it. If it could be a profitable going concern,
1:33:26
shoot
1:33:27
a note. to hi at tiny dot Tell them Ben and David
1:33:29
sent you, and thanks so much to our
1:33:31
friends at tiny. Well, Ben, I have
1:33:33
one more section that I wanna
1:33:35
do, which is you
1:33:37
do all this analysis on companies and you have this
1:33:39
very enviable position of getting to comment on and critique their strategies after
1:33:41
they announce them or an event
1:33:43
after they do them. And
1:33:46
I wanna play a game where I give you a
1:33:48
company. And I'm curious, either if you were the
1:33:50
CEO or you were giving advice to
1:33:53
the CEO, what would you do strategically? And I think
1:33:55
an interesting
1:33:55
place to start because you talked about it on a bunch on this
1:33:57
episode is meta. You are the new
1:33:59
CEO
1:33:59
of meta. What do you
1:34:02
do over the next five years?
1:34:04
Well, I think beta is
1:34:07
actually doing a fair number of things that they should. Now,
1:34:10
now
1:34:11
the Apple changes were very
1:34:13
devastating, and they're structurally devastating. And so they deserved a significant haircut
1:34:15
on their valuation because
1:34:19
of that. but I also think they were mode enhancing in the long run,
1:34:21
where I question whether any other company is
1:34:23
ever gonna build a top of the
1:34:25
funnel advertising product that will be
1:34:27
competitive with Beto. assuming
1:34:29
they can keep their audience. It appears that they've done a good job limiting the sort
1:34:31
of the TikTok threat. growth has sort of
1:34:33
flattened out over the last
1:34:36
few years. So
1:34:38
I think they're actually in good shape.
1:34:40
What I do worry about is this is a company
1:34:42
that's always been growing. It's really hard for those companies
1:34:45
to shift to more of scarcity mindset.
1:34:47
I mean, they weighed off eleven thousand
1:34:49
people, which only brings them back like nine
1:34:51
months. They need a lot smaller, but it's
1:34:53
also very destructive to company culture and morale.
1:34:55
do they still have people they need to pull that off? Those are
1:34:58
probably some of the bigger questions. And obviously,
1:35:00
any aggregator is dependent on having
1:35:02
that audience. But I do think that
1:35:05
the network effects of their products are
1:35:07
still underrated. I mean, everyone has their head that they're shrinking and actually product they have is still
1:35:10
growing, which I think is underappreciated.
1:35:12
So Number
1:35:15
one, I think they're kind of broadly on the right track. Number two,
1:35:17
I would acquire Shopify and I
1:35:20
would take the FTC adjustment
1:35:22
to court when they soon stop
1:35:24
it. I think they need to close the
1:35:26
loop on e commerce and advertising. And, again, I think that'd probably
1:35:28
be bad generally, but I think
1:35:30
it'd be very good for meta.
1:35:33
So I would do that. As far
1:35:34
as the Metiverse stuff, I think it's a bad idea. It's very hard to talk about the Metiverse because it
1:35:37
was
1:35:37
like, what are
1:35:40
the prospects? versus, is it, you know,
1:35:42
XYZI do think that it's not just a bad idea because it's taking so many resources
1:35:44
and attention, I think,
1:35:46
from architect word, but also
1:35:50
I just don't think innovation is
1:35:52
necessarily born of mass expenditures
1:35:54
in large companies. We sort
1:35:56
of skipped over a period
1:35:58
of experimentation and ecosystem that in the
1:36:00
long run, perhaps, would be consolidated into a couple companies.
1:36:03
But instead, it's just one monolith sort of
1:36:05
trying to brute force this
1:36:07
sort of bit. And the reality
1:36:10
is this Facebook is a services company, it's a social network, and everything about their strategy against
1:36:12
that. To succeed, they not
1:36:14
only need to sell headsets, They
1:36:19
use our headsets at sufficient scale and into friend
1:36:21
networks such that people can social network
1:36:23
on the headsets. And
1:36:25
so they're Place of winning is
1:36:28
even further away than I think people think.
1:36:30
And so I don't think strategically it's the
1:36:32
best thing for them to be doing.
1:36:34
And I've been pretty anti them doing from day one. They bought
1:36:36
oculus, I said it was bad. And again,
1:36:38
it's very hard to distinguish between, well,
1:36:41
what are the
1:36:43
prospects of versus zooming out? And should they even
1:36:45
be doing this? But I think the market is overly down on meta in part because the
1:36:48
branding was
1:36:50
too successful. this is still a powerhouse in social media and
1:36:52
advertising, and the bout of
1:36:54
money they're spending on the
1:36:57
better versus is all things considered not that
1:36:59
much. They're still, you know, have a five billion
1:37:02
dollar profit a quarter. So, yeah, I would
1:37:04
double down on what they are,
1:37:06
and I would consolidate, not acquire Shopify, and
1:37:08
spend three years fighting out in court because I
1:37:10
think the payoff would be worth it. Alright. Second
1:37:12
one,
1:37:12
rather than going with another
1:37:15
big American tech company, I wanna go
1:37:17
with one that's an acquired, fan favorite, and one that's very
1:37:19
close to home for you, TSMC.
1:37:23
Any change to what TSMC
1:37:25
is doing now if you became CEO tomorrow. Not really.
1:37:27
It's a very complicated situation
1:37:28
to say the
1:37:30
least for lots of reasons, not
1:37:32
just political, technological,
1:37:34
I think probably doing what they're doing. Broadly, it makes sense. Yeah. I don't know. I've ever written a ton about them
1:37:37
and where they
1:37:40
are, and you
1:37:42
could talk about maybe some of the pricing stuff. They
1:37:44
were pretty slow to raise prices even
1:37:46
in the face of shortages. They've been
1:37:49
much more aggressive about that over
1:37:51
the last which seems reasonable to me.
1:37:53
It is opening the door for when and if a competitor
1:37:55
comes along to undercut
1:37:58
them in that regard. I think they're probably blessed by the weakness
1:38:00
of their competition. But the long
1:38:02
term risk is obviously number one
1:38:05
just the geopolitical risk. The
1:38:07
reality is is TSMC a
1:38:09
core to their model, essentially
1:38:11
to model being so of incorporating new
1:38:15
equipment, of accommodating hundreds
1:38:17
and hundreds or thousands of customers is all their engineering's in one
1:38:19
And that
1:38:20
the order engineers in one place and
1:38:22
that place is Taiwan, and
1:38:25
that has geopolitical risk. And from TSMC's perspective, I
1:38:27
think that's just a reality you can't
1:38:28
really hedge against
1:38:31
that. And so if
1:38:33
anything, yes, they're building these fabs
1:38:34
in the US, I think, larger for political reasons. And sure, do that if,
1:38:36
you know, the Chinese government feels
1:38:38
that's what needs to be done.
1:38:42
to keep the US on board, could be as happy, build a
1:38:44
couple in Japan because Japan is a future
1:38:46
ally. But I think you do have to sort
1:38:48
of double down in Taiwan and roll the dice
1:38:50
that nothing happens. I don't think it's a hedgeable risk, the China risk. The
1:38:53
other risk is the, you know,
1:38:55
Moore's loss are running out.
1:38:58
I mean, EUV has another five, six years,
1:39:00
and then it's not super clear
1:39:02
what's after that. How can you
1:39:05
get smaller than one, Ben? just have
1:39:07
one nanometer in there. There is visibility to under one,
1:39:09
but isn't it like two molecules
1:39:11
wide or two
1:39:14
atoms? It's very small. I'm not sure if it's quite
1:39:16
that small. It's ridiculously small. So obviously,
1:39:18
I I would imagine they're investing heavily in
1:39:21
they've been doing this, like, advanced packaging, like, multiple chips on the chip
1:39:23
that sort of approach. AMD is doing that. Intel is not doing that.
1:39:28
And getting really good at that stuff
1:39:30
is gonna be super important. I mean, actually, one thing I do like about their Japan investment is their trailing
1:39:32
edge. I think their twenty
1:39:34
eight nanometer fabs, which is
1:39:36
which is really where
1:39:39
China is making a lot of progress
1:39:41
because there's no
1:39:42
real economic reason to build TriNet's fabs, the
1:39:44
whole idea is you build it one
1:39:46
it's long since depreciated and you're still making
1:39:48
money selling cheap chips out of that
1:39:50
fab. But I don't know. Maybe there's
1:39:52
something in the drawing edge. More things need chips
1:39:54
and they don't all need the leading edge. That's right. The
1:39:56
economics of it are very difficult. I mean, the reality is that's why China
1:39:58
is filling the gap because that's all they can really make economically. You
1:40:01
don't just jump
1:40:03
to the waiting you have to sort of
1:40:05
build your way up. And so China has a motivation and an economic rationale. And a lot of
1:40:07
the chips that they're manufacturing
1:40:11
go right into products
1:40:12
that are already made in China. And so they're filling
1:40:14
in that gap. You know, that's something that would be beneficial.
1:40:18
But at the end of day, CsmC, Taiwan. And that's gonna always be the biggest
1:40:20
risk, and I'm not sure there's really much they
1:40:22
can do about that. Okay. Last one, Amazon,
1:40:25
David and I did seven hours of amazon
1:40:28
dot com and then a big AWS episode. And
1:40:30
our I mean, at least my big takeaway
1:40:32
was, it is day
1:40:34
two. And Day two is about
1:40:36
becoming a profitable company where the
1:40:38
big cell story of tomorrow is
1:40:40
realized today. And you gotta lean into that,
1:40:42
and that means lots of changes in how they organize, how
1:40:44
they innovate, and what markets they choose to
1:40:46
enter. And I'm sort of curious for
1:40:50
your take on Is Amazon a day two company and then
1:40:52
the same question if you were to become CEO?
1:40:54
What would you do? I think there's an
1:40:57
analogy to, like, real life when you're young, you're like, I'm
1:40:59
never gonna become an old bogey like those old
1:41:02
people. And then you get old like
1:41:04
me, and the people around you that are
1:41:06
trying to still be young are just kind
1:41:08
of pathetic. And actually, being old
1:41:10
is kinda great. My kids are fairly independent. They can take care of themselves.
1:41:12
I can going
1:41:16
out with friends. Obviously, I have more means
1:41:18
than I did when I was younger. And I'm a big advocate of
1:41:20
just
1:41:20
in general living in the present
1:41:22
and embracing who you are
1:41:25
where you are your life stage. And I think that's good personally, and I think you're spot on
1:41:27
if that's also true from company perspective. You can't
1:41:30
be a startup forever. And
1:41:34
it's bad too. I wrote about from day
1:41:37
two to one day or something
1:41:39
about Amazon when Bezos sort
1:41:41
of like re asserted control a few years
1:41:43
ago. I was like, no, we're gonna do one day
1:41:45
delivery. And there's been too much time spent trying to
1:41:47
squeeze our suppliers profits
1:41:49
and margins and know we need to
1:41:51
get back to this. And in retrospect, that
1:41:53
was the seed of Amazon's kinda disastrous last
1:41:56
few years. which was
1:41:58
they dramatically over invested in their logistics network. Their cost structure got
1:42:00
completely out of
1:42:03
control. They over hired And
1:42:05
it's probably a good example and maybe it was a precursor of what we saw
1:42:07
with Jeff Bezos personally of
1:42:12
sort of losing track of
1:42:14
where you are in life and what actually makes sense and trying to be young forever.
1:42:17
So I agree
1:42:20
with you broadly, with
1:42:22
that bit
1:42:22
about Amazon, do you think that applies to AWS too? Or is that different?
1:42:24
Realistically, the challenge
1:42:27
for AWS is Number
1:42:31
one, what's gonna happen with the real
1:42:33
dry up in startup formation and
1:42:35
easy money in that space? A
1:42:37
lot of which went to Amazon they're
1:42:39
sort of the
1:42:39
default choice for startups. Number two, you know, Microsoft
1:42:42
sort of bread and
1:42:43
butter has been look,
1:42:45
you've been
1:42:46
working us for a long time.
1:42:48
we're gonna package it at like, suddenly you're not
1:42:50
just paying for on premise windows. It also includes Azure credits, and we're gonna be able to attribute that to
1:42:52
our numbers. And now you can sort of move pieces
1:42:54
over and we'll help you do it. what
1:42:58
I think is exactly what they should do. It's been a very
1:43:00
smart strategy, but
1:43:01
I think that the issue in
1:43:03
any market is, it's like
1:43:04
me. All the initial easy stuff is like
1:43:06
there's lots of low hanging fruit. but then the largest part market
1:43:08
is still the part of the market that
1:43:10
sort of always been there and
1:43:11
is initially fast moving. And Microsoft
1:43:13
has just really cleaned up
1:43:15
in that market. where
1:43:17
the known entity we can help
1:43:20
you move over. And I think from Amazon
1:43:22
building up the support capabilities and rationalizing their
1:43:24
offerings, mean,
1:43:27
it's weird because Amazon benefits because they have so many features.
1:43:29
Right? AWS is the Microsoft Word
1:43:31
of cloud providers. In
1:43:33
that in that. an absolute absurd
1:43:36
number of features, and the
1:43:38
interface is pretty terrible. But
1:43:40
every single customer is completely
1:43:42
dependent on one of those features. And
1:43:44
if that feature's not there, then they can't go
1:43:46
away. And that actually ends up being their
1:43:48
modes. We were researching the episode and
1:43:50
we were talking to longtime AWS veterans to sort of understand mental framework
1:43:53
to talk about
1:43:56
them today. and that became very
1:43:58
clear that AWS doesn't deprecate features. Amazon will kill stuff, kill a fire phone, kill
1:43:59
local delivery
1:44:04
for food, They'll call all kinds
1:44:06
of stuff that in AWS land, if a customer is depending on something, AWS's long term enterprise
1:44:08
value is
1:44:12
determined by customers believing that Amazon
1:44:14
will continue to support them forever. They don't deprecate services even when those services end
1:44:16
up flipping upside down on the
1:44:18
unit economics, and they have to, like,
1:44:22
maintain a costly service that they never figured
1:44:24
out how to optimize. Right.
1:44:26
And every time a customer does
1:44:28
something custom for AWS, it's a
1:44:30
walk in. Right? I mean, everyone fantasizes about this
1:44:32
world where, you know, like, every time it comes up, whether it be,
1:44:34
you know, containers or you have all these, you know, IBM's
1:44:37
is, like, talking about doing this when
1:44:39
they acquire Red Hat. we're
1:44:41
gonna make it so you can
1:44:43
be cloud agnostic. Right? Right. And then it just turns out that, well,
1:44:48
quadnostic, but this one little piece of
1:44:50
bead would be better than better than better service. Right? And then you wake up and, you know, of course, I want
1:44:52
a passport to
1:44:55
be cloud agnostic. passport's not moving off AWS.
1:44:57
I could be I could promise you that. And so, yeah, it's
1:44:59
the Microsoft strategy. Amazon has a
1:45:02
lot of old Metro people in
1:45:04
it. a ton. It's
1:45:06
really interesting to consider, like, the different cultures between Seattle and San Seattle just
1:45:08
is a platform
1:45:11
town. Like Microsoft Where's
1:45:14
the originator? Microsoft's always been
1:45:16
the best platform administrator. All that backwards
1:45:18
compatibility that you wanna make fun of from
1:45:20
a user perspective is essential to building this
1:45:22
foundation that people trust and that walks them in, and they're happy to be
1:45:25
locked in because they don't wanna
1:45:27
go change it anyway. Right? and
1:45:31
Amazon does with AWS. Obviously, Microsoft is doing that
1:45:33
with Azure. And the Silicon Valley companies are
1:45:35
just very, very bad at
1:45:37
that. Right? Like, no one trusts Google.
1:45:39
Right? No one trusts Facebook. Silicon Valley is much more consumer
1:45:41
focused. Even the SaaS companies, those
1:45:43
are consumerized enterprise technology.
1:45:46
The whole idea is don't
1:45:48
worry. You never need to pay for an upgrade. We're
1:45:50
upgrading on the back and all our self, but that means, like, they will remove stuff because it's one thing
1:45:54
to remove an API that piece
1:45:56
of software depends on versus removing a feature that EMA,
1:45:58
your customers are annoyed, but it's not
1:45:59
actually breaking
1:46:02
like what they operate on. And it's just pretty interesting to see those differences,
1:46:04
which I do think there's a geographic
1:46:06
aspect to it. I mean, geographies,
1:46:10
let's say, Microsoft because
1:46:12
When
1:46:13
Microsoft was in the dumps, if they were in
1:46:15
Silicon Valley, all their best talent would have left and not worked for other companies. But all their best talent
1:46:17
had kids, they
1:46:20
had families, didn't wanna work for
1:46:22
Amazon. Those people in maniacs. And so they stayed on Microsoft, and they were miserable, and they bitched, and they wrote
1:46:24
snarky blog posts
1:46:27
about the company, But then
1:46:30
when Nadella came in and sort of refocused the company, they had this foundation
1:46:36
of talent. that the HP's of the world,
1:46:38
the Yahoo's of the world had long since lost. Well, Ben, this has been awesome. Thank
1:46:40
you for joining us
1:46:43
and celebrating the Almost ten
1:46:45
year anniversary of stratigraphy with us. What is the easiest path for everyone
1:46:47
who right now is listening to this in a
1:46:49
variety of podcast players? You got Spotify
1:46:51
and Overcast and Apple
1:46:55
Podcasts. What is the easiest way for them to opt into the
1:46:57
stratigraphy universe? Well, I mean, if you go
1:46:59
to stratigraphy, you can click
1:47:01
any of the podcasts out of the side, you
1:47:03
subscribe, I mean, it's funny, like, the how does the funnels work? Right? Do you
1:47:05
just go a Checkatrade and read and follow? And I'll
1:47:07
I'll lure you in. Well, you'd go direct.
1:47:10
Maybe we'll put a link in the show
1:47:12
notes Yeah. Yeah. Yeah. We use the show
1:47:14
note link. Twelve dollars a month, hundred and twenty dollars a year, you not only get Stretectory, but
1:47:16
you also get the
1:47:18
update, you get Stretory interviews,
1:47:21
didn't talk about interviews. That's actually been
1:47:23
an interesting evolution of trajectory. Yeah. You're a primary source now. What's
1:47:27
going on? Well, that was a challenge. Right? because
1:47:29
I started out. I had no access. I didn't know anyone. And in some respects, I missed those days.
1:47:31
I still remember the first time I had a
1:47:34
company very angry at me and called me
1:47:36
in. and it
1:47:38
was Twitter actually because I had teased out their
1:47:40
results that their direct response program was failing. This was like years
1:47:42
ago and they got super mad in in my next
1:47:47
update. I'm like, well, it's possible this. I kind
1:47:49
of like walked it back a little bit. And
1:47:51
the next quarter out and they're like, yeah, we're
1:47:53
gonna need to I don't know if they go
1:47:55
write down, but it was a big thing and they said, like, it's not working out. And
1:47:57
I was totally right, and I was so mad
1:47:59
that I had walked it back a
1:48:01
little bit, but I'm glad it
1:48:04
happened because when I started out, no one cared, no
1:48:06
one paid any attention. Now people cared and sort of paid attention. And so I'm glad I was right on that
1:48:08
one because it sort of
1:48:10
gave me courage going forward. But
1:48:13
one that's interesting about feedback from companies is everyone
1:48:15
like COO down or VP down, everyone like from
1:48:18
philo down or vp down they
1:48:20
always push back. And they're like, no,
1:48:22
you got this wrong. You understand if was he, I never get pushed back from CEOs. Even
1:48:24
when I'm wrong. And what
1:48:26
I've come to
1:48:27
realize is CEOs
1:48:30
are surrounded by people telling them
1:48:32
what they wanna hear. That's their incentives. Right? The
1:48:34
reason why the VP is attacking me is
1:48:36
because he's worried I'm gonna make the
1:48:38
VP look bad in front CEO. Right?
1:48:40
That's the concern. Whereas the CEO is like,
1:48:43
they're so thirsty for any sort of
1:48:45
feedback that's outside
1:48:48
of that rotten incentive structure that's just inherent
1:48:50
in any corporation. And so sometimes they know they have more information that I do.
1:48:52
They're aware that there's things
1:48:54
that I don't know, but they're
1:48:57
grateful to have sort of the different point of view. And so over time,
1:48:59
I transfer having no access to having total Right? I can
1:49:02
reach out to basically
1:49:04
anyone And just for people
1:49:06
who haven't, let's just go through the lineup. You've had this year, Mark Zuckerberg gone. You've had Jensen from NVIDIA. You've had I'm
1:49:08
trying to think of some of your
1:49:10
early big ones. You've Rich Barton on from
1:49:14
and I think he credited you with you convince us to start
1:49:16
buying houses. You convince us in the open door. I
1:49:18
don't I don't know why to apologize for
1:49:23
that. So I was trying to figure out, like, what do I do with this
1:49:25
access? I didn't wanna become a reporter. And I
1:49:27
thought, you know, it's just important to me personally
1:49:29
for my own pride that my takes are
1:49:31
my takes. I'm not getting fed them
1:49:33
from someone else. And so what I ended up going
1:49:35
with was interviews. And part of
1:49:37
this was a product bit.
1:49:40
I had now you can
1:49:42
listen to Checkatrade via podcast. So if I had an interview, there's a good reason to try the product out to get it. So there a
1:49:44
product angle, but
1:49:47
also I'm like, My whole
1:49:50
take is, I'm not exclusive. I don't have exclusive information. What I'm selling is my
1:49:53
personal analysis.
1:49:56
And so if I'm talking
1:49:58
to a CEO, I don't want the sense to be Ben's just parodying what
1:50:00
a CEO says. So if a CEO wants to talk
1:50:02
to me, it has to be on the record.
1:50:06
and it's going to be the full transcript and the full
1:50:09
interview is gonna be available to my subscribers.
1:50:11
And I want my subscribers
1:50:13
to have all the same information I do
1:50:15
And so when I write, it's me. It's coming out of my brain. It's sort
1:50:17
of unique to me. And I think it's worked out
1:50:19
pretty well. It's a
1:50:22
change. Like, instead of be running four days a week. I'm basically running three days a
1:50:24
week plus interview. It's not just CEOs. I think
1:50:26
actually often the better interviews are with, like,
1:50:29
other analysts or people in different spaces. This also solves the
1:50:31
how do I cover startups because part of not
1:50:33
being a reporter and being independent
1:50:35
is public companies have to
1:50:37
disclose a lot of data and and they have earnings calls and they
1:50:39
have presentations and things on those lines. With a startup, you
1:50:41
don't have access to very much of that and you
1:50:44
don't know how much
1:50:46
is Boeing smoke, how much is not. So now I am doing, like, interviewing founders
1:50:48
where I say upfront, look,
1:50:50
this is completely subjective. I'm not
1:50:52
gonna verify what they say. I'm not gonna
1:50:54
XYZ It's their chance to
1:50:57
tell their story, what their business is gonna be XYZ And so that's a way
1:50:59
to sort of cover that space given the limitations
1:51:02
in the way Stackery operates.
1:51:07
And so that's one of the
1:51:07
products. It's not a stand alone thing,
1:51:10
which I think is better. It's still
1:51:12
potash or
1:51:12
techy. I thought about should
1:51:14
be a stand alone product, but
1:51:16
I think there's an aspect of I don't
1:51:18
want there to ever be any sense that that's a driver of my bottom line. I wanna be able
1:51:21
to say that I think the
1:51:22
meta versus probably a bad idea.
1:51:26
that's really tricky. Right? Like, I mean, though Barksuckerberg
1:51:28
won. Obviously, it's a huge get. I think
1:51:30
that his interviews with me are I'm
1:51:32
biased, but I think our what year's better
1:51:34
than he does anywhere else. You and Nila had the best interviews with him of this whole last site. Right. It it
1:51:36
bugs that he thinks I'm fair, whereas
1:51:38
he thinks a lot of other journalists
1:51:43
aren't. But what's the line between that
1:51:45
and being like you favor
1:51:47
Facebook? Because that's not the case
1:51:49
at all. Number one, it doesn't
1:51:52
drive subscriptions. I always make them free to be clear. Like,
1:51:54
no one's gonna have to pay to get access to that because you shouldn't subscribe to Secretary because
1:51:56
you want to reimburseuckerberg.
1:51:59
That's not a reason. but then
1:52:00
it was really annoying because Facebook had that bad
1:52:02
earnings call and their stock went down. And I was like, crap. I've got an idea to write article about
1:52:04
why this reaction is
1:52:06
unwarranted. They're actually fine. This
1:52:09
gonna be, like, three weeks after I just sent
1:52:11
a remark back over God. It's gonna be a
1:52:13
relatively positive article, but that's, like, where you build
1:52:15
up your reputation over
1:52:16
years. Right? And sometimes you can't always be
1:52:19
depositing. Sometimes you do a withdrawal. Me writing meta myths
1:52:21
was a
1:52:22
withdrawal on my reputation. Yes,
1:52:26
it did look great that I
1:52:28
just got an interview with Marcus Hookerberg, and
1:52:30
I wrote a positive article, a contrarian,
1:52:32
positive article about met up three weeks
1:52:35
later. but I hope I've built up enough sort of credibility with
1:52:37
my audience over the last nine years
1:52:39
that they believe that, no, that
1:52:41
I would have written an article
1:52:43
with or without that interview. You're
1:52:45
also unabashed about your incentives. You're like, look, everyone. My incentives are to have
1:52:47
as many people subscribed
1:52:50
for as long as
1:52:52
possible. my incentive is not I think
1:52:54
I'm gonna be Mark Zuckerberg's best friend, so I'm really hoping that he invites me on his p j and we get to hang out.
1:52:56
Yeah. Well, it's tough because like I
1:52:58
said, I'm super positive on meta's advertising
1:53:02
product because that's aligned with my passion. So
1:53:04
I do find the Zuckerberg one's
1:53:06
the most challenging just because
1:53:08
I generally have a contrarian
1:53:11
opinion about meta over all and it's generally
1:53:13
positive. So I'm super wary about that. I also feel like
1:53:15
I get more interesting stuff out
1:53:17
of him than most
1:53:19
other interviewers, so I wanna do that,
1:53:21
and sure it's a feather in my cap two interview. But I definitely don't want
1:53:23
that to ever perceived as really,
1:53:27
these interviews don't drive subscriptions. And so I do
1:53:29
it because, again, I wanna push
1:53:32
the tech product on people
1:53:34
who use my podcast product. And
1:53:36
then
1:53:36
then it's
1:53:37
a different kind of work. As you guys know, preparing for interviews is a lot of
1:53:39
work. It's also different than necessarily, like, writing
1:53:42
an article of for
1:53:44
sure. we could not
1:53:46
be writers. Yeah. You can't say the same forever. You can't be really like the startup forever. Right? I can't be the
1:53:48
guy on the outside forever,
1:53:50
which I was when I started
1:53:54
like with my chip on my shoulder and I'm just
1:53:56
some dude in Taiwan giving
1:53:58
his opinions about Apple, that
1:54:00
worked for me? and I'm not that
1:54:02
person anymore and it'd be sort of dishonest person that I am. The reality is I probably have the best access to anyone
1:54:04
in tech. I can literally contact basically
1:54:06
anyone and get access to them. and
1:54:11
it's
1:54:11
more honest to my product and what I
1:54:14
do to weed into that I think
1:54:16
than
1:54:16
to pretend it's
1:54:18
not the case. Alright. So listeners, click the link in the
1:54:20
show notes. Go experience the spectacular
1:54:22
cinematic universe, not just a guy
1:54:24
in Taiwan, giving his opinion on
1:54:26
Apple or coming hopefully. Love it.
1:54:28
With
1:54:29
that, thank you to Fundrise, Pilot, and Tiny.
1:54:31
After you've finished this episode, come hang out in the
1:54:33
Slack with thirteen thousand other
1:54:35
smart, kind, clever curious
1:54:39
members of the acquired community, acquired dot
1:54:41
f m slash Slack, get some
1:54:43
suite acquired merch. We
1:54:45
now officially have live.
1:54:48
The ACQT We've got market size
1:54:50
unconstrained with the original AWS logo. Oh, we gotta do a collab with Ben
1:54:52
on a acquired stratigraphy t
1:54:54
shirt for the merch store. Oh,
1:54:58
that's a good point. I have a stratigraphy t shirt. I
1:55:00
almost wore it for recording this episode. I
1:55:02
also have a framed aggregation theory. I don't
1:55:04
want a fanboy in front of Ben, but on my
1:55:07
desk, I have a framed aggregation theory illustration. So I
1:55:09
don't know,
1:55:09
check out Ben's
1:55:10
merch store too. We'll put a link
1:55:13
to that in the show notes. but the
1:55:15
other great acquired merch that dropped a couple of
1:55:17
weeks ago is the benchmark tea that in
1:55:19
the benchmark style of
1:55:21
course says there's always room at the
1:55:23
top. If you wanna listen to the acquired LP and get more acquired crammed in
1:55:25
before the holiday season. Got some good
1:55:28
stuff. Search acquired LP
1:55:30
show in the podcast player.
1:55:33
of your choice. And with at listeners, we hope you
1:55:35
have a wonderful holiday season. I believe we
1:55:37
still have one more
1:55:39
episode coming after No.
1:55:41
Yeah. We've got something special, Kevin. We
1:55:44
got a little fun in holiday acquired
1:55:46
present for everybody. Yes. And
1:55:47
with that, listeners. We'll see
1:55:49
you next time. We'll see you next
1:55:51
time. Who the Is Is it you? Is it
1:55:54
you? Who got the truth now?
Podchaser is the ultimate destination for podcast data, search, and discovery. Learn More