Kind: captions
Language: en
the following is a conversation with
kevin systrom co-founder and long-time
ceo of instagram including for six years
after facebook's acquisition of
instagram
this is the lex friedman podcast to
support it please check out our sponsors
in the description
and now here's my conversation with
kevin
systrom
at the risk of uh asking the rolling
stones to play satisfaction let me ask
you about the origin story of instagram
sure
so maybe some context you
like we were talking about offline grew
up in massachusetts
learned computer programming there liked
to play doom two
uh worked at a vinyl record store then
you went to stanford
turned down mr uh
mark zuckerberg and facebook
went to florence to study photography
those are just some random beautiful
impossibly brief glimpses into a life so
let me ask again can you take me through
the origin story of instagram giving
that context you basically set it up um
all right so uh we have a fair amount of
time so i'll go into some detail but
basically what i'll say is
um
instagram started out of a company
actually called bourbon
and it was spelled b-u-r-b-n
and uh
a couple things were happening at the
time so if we zoom back to 2010 not a
lot of people remember what was
happening in the dot-com world then uh
but
check-in apps were all the rage
so what's this checking out uh gowalla
four square hot potato
so i'm at a place i'm gonna tell the
world that i'm at this place that's
right what's what's the idea behind this
kind of app by the way you know what i'm
gonna answer that but through
what instagram became and why i believe
instagram replaced them
so the whole idea was to share with the
world what you were doing specifically
with your friends right
um
but they were all the rage and
foursquare was getting all the press and
i remember sitting around saying hey i
want to build something but i don't know
what i want to build what if i built a
better
version of foursquare
and i asked myself
why don't i like foursquare or how could
it be improved
um
and basically i sat down and i said i
think that if you have a few extra
features it might be enough one of which
happened to be posting a photo of where
you were
there were some others it turns out that
wasn't enough my co-founder joined we
were going to attack uh you know
foursquare and the likes and and try to
build something interesting
um and no one used it no one cared
because it wasn't enough it wasn't it
wasn't different enough right
so one day we were sitting down and we
asked ourselves okay let's come to jesus
moment
are we going to do this startup
and if we're going to we can't do what
we're currently doing we have to switch
it up so what do people love the most so
we sat down and we wrote out
three things that we thought people
uniquely loved about our product that
weren't in other products
photos happened to be the top one so
sharing a photo of what you were doing
where you were at the moment was not
something
products let you do really facebook was
like post an album of your vacation from
two weeks ago right twitter allowed you
to post a photo but their feed was
primarily text and they didn't show the
photo in line or at least i don't think
they did at the time
so
even though it seems totally
stupid and obvious to us now at the
moment then
posting a photo of what you were doing
at the moment was like not a thing
so
we decided to go after that because we
noticed that people who used our service
the one thing they happened to like the
most was posting a photo
so that was the beginning of instagram
and yes like we went through and we
added filters and there's a bunch of
stories around that but the origin of
this was that we were trying to be a
checking app realized that no one wanted
another checking app
it became a photo sharing app but one
that was much more about what you're
doing and where you are
and that's why when i say i think we've
replaced checkin apps
it became a check-in via a photo rather
than
saying your location and then optionally
adding a photo when you were thinking
about what people like
from where did you get a sense that this
is what people like you you said we sat
down we wrote some stuff down on paper
where is that intuition that seems
fundamental to the success
of
an app like instagram what is that idea
where's that list of three things come
from exactly
only after having studied machine
learning now for a couple of years
i like i have a you have understood
yourself
i've started to make connections like we
can go into this later but
obviously the the um
the connections between
machine learning and the human brain i
think are stretched sometimes right
at the same time
being able to backprop and being able to
like look at the world try something
figure out how you're wrong how wrong
you are
and then nudge your company in the right
direction
based on how wrong you are is like a
fascinating concept right and i don't we
didn't know we were doing it at the time
but that's basically what we were doing
right
we put it out to call it a hundred
people
and you would look at their data you
would say what are they sharing
what like what resonates what doesn't
resonate we think they're going to
resonate with x but turns out they
resonate with y
okay shift the company towards y
and it turns out if you do that enough
quickly enough you can get to a solution
that has product market fit most
companies fail
because they sit there and they don't
either their learning rate's too slow
they sit there and they're just they're
adamant that they're right even though
the data is telling them they're not
right
or
they their learning rate's too high and
they wildly chase different ideas and
they never actually set on on one where
where they don't groove right and i
think when we sat down and we wrote out
those three ideas what we were saying is
what are the three possible
whether they're local or global maxima
in our world right
that users are telling us they like
because they're using the product that
way
it was clear people liked the photos
because that was the thing they were
doing
and we just said okay like what if we
just cut out most of the other stuff and
focus on that thing
um and then it happened to be a
multi-billion dollar business and
it's that easy by the way yeah i guess
so um well nobody ever writes about
neural networks that
miserably failed so this this particular
neural network succeeded this is the
sound all the time right yeah but nobody
right default state is failing yes
um when you said the way people are
using the app
is that the lost function for this
neural network or is it also self-report
like do you ever ask people what they
like or do you have to
track exactly what they're doing not
what they're saying
i once made a thanksgiving dinner okay
and
uh it was for
relatives and i like to cook a lot okay
and i worked really hard on picking the
specific
uh dishes and
and i was really proud because i had
planned it out using a gantt chart and
like it was ready on time and everything
was hot nice like i don't know if you're
a big thanksgiving guy but like the
worst thing about thanksgiving
is when the turkey is cold and some
things are hot and something anyway you
got a gantt chart you actually have a
chart oh yeah yeah omni plan fairly
expensive like gantt chart thing that i
think maybe 10 people have purchased
in the world but i'm one of them and i
use it for recipe planning only around
big holidays that's brilliant by the way
do people do this kind of uh over
engineering it's not overdue it's just
engineering it's planning thanksgiving
is a complicated uh
set of events with some uncertainty with
a lot of things going on you should be
able you should be planning in this way
there should be a chart it's not over i
mean so what's funny is um
brief aside yes uh brilliant i love
cooking i love food i love coffee and
i've spent some time with some chefs who
like know their stuff and
they always just take out a piece of
paper and just work backwards in rough
order like it's never perfect but rough
order it's just like oh that makes sense
why not just work backwards from from
the end goal right and put in some
buffer time and
so i probably overspecified it a bit
using a gantt chart but the fact that
you can do it it's what professional
kitchens roughly do
they just don't call it a gantt chart or
at least i don't think they do um anyway
i was telling a story about thanksgiving
so here's uh here's the thing
i'm sitting down we have the meal and
then i'm
you know i got to know ray dalio
fairly well over maybe the last year of
instagram um
and one thing that he kept saying was
like feedback is really hard to get
honestly from people
and i sat down at
after dinner i said guys i want feedback
what was good and what was bad yes and
what's funny is like literally everyone
just said everything was great
and i like personally knew i had screwed
up a handful of things
um but no one would say it
and can you imagine now
not something as high stakes as
thanksgiving dinner okay thanksgiving
dinner it's not that high stakes
but you're trying to build a product and
everyone knows you left your job for it
you're trying to build it out and you're
trying to make something wonderful
and it's yours right you designed it
now try asking for feedback
and know that you're giving this to your
friends and your family
people have trouble giving hard feedback
people have trouble saying
i don't like this or this isn't great or
this is how it's failed me
in fact
um
you usually have two classes of people
people who just won't say bad things you
can literally say to them please tell me
what you hate most about this and they
won't do it they'll try but they won't
and then the other class of people are
just negative period about everything
and it's hard to parse out
like what is true and what isn't
so my rule of thumb with this
is you should always ask people
but at the end of the day it's amazing
what data will tell you
and that's why with whatever project i
work on even now
collecting data from the beginning on
usage patterns so engagement how many
days of the week do they use it
how many i don't know if we were to go
back to instagram how many impressions
per day
right is that growing is that shrinking
and don't be like overly
scientific about it right because maybe
you have 50 beta users or something
but what's fascinating is that
data doesn't lie
people are very
uh defensive about their time
they'll say oh i'm so busy i'm sorry i
didn't get to use the app like i'm just
you know um
but i don't know you're posting on
instagram the whole time
so
i don't know at the end of the day like
at facebook there was
you know before time spent became
kind of this loaded
term there
the idea that people
people's currency in their lives is time
and they only have a certain amount of
time to give things whether it's friends
or family or apps or tv shows or
whatever it's
there's no way of inventing more of it
at least not that i know of
if they don't use it it's because it's
not great
so
the moral of the story is you can ask
all you want but you just have to look
at the data
and
data doesn't lie right i mean there's
metrics there's uh
data can obscure the key insight if
you're not careful so
so time spent in the app that's ones
there's so many metrics you can put at
this and they will give you totally
different insights
especially when you're trying to create
something that doesn't obviously exist
yet
so
you know
measuring maybe
why you left the app or measuring
special moments
of happiness that will make sure you
return to the app or moments of
happiness that are long lasting versus
like dopamine short term
all of those things but i think i
suppose in the beginning
you can just get away with just asking
the question
which features are used a lot let's do
more of that
and how hard was the decision
and uh i mean maybe you can tell me what
instagram looked in the beginning but
how hard was it to make
pictures the first class citizen that's
a revolutionary idea
like um at whatever point instagram
became this feed
of
photos
that's quite brilliant
plus
i also don't know when this happened but
they're all shaped the same
it's like uh i have to tell you why
that's the interesting part
why is that so a couple of things one is
data data like you're right you can over
interpret data like imagine trying to
fly a plane
by staring at
i don't know a single metric like
airspeed
you don't know if you're going up or
down i mean it correlates with up or
down but you don't actually know it will
never help you land the plane
so don't stare at one metric like it
turns out you have to synthesize a bunch
of metrics to know where to go
um but it doesn't lie like if your air
speed is zero unless it's not working
right if if it's zero
you're probably going to fall out of the
sky so
generally you look around and you have
the scan going yes
and you're just asking yourself is this
working or is this not working
um
but people have trouble explaining
how they actually feel
so just it's about synthesizing both of
them so then instagram right uh we were
talking about revolutionary moment where
where the feed became
square photos basically and photos first
and then square footage yeah um
it was clear to me that the biggest so
i believe the biggest companies are
founded
when
enormous technical shifts happen and the
biggest technical shift that happened
right before instagram was founded was
the advent of a phone that didn't suck
the iphone right like in retrospect
we're like oh my god the first iphone
that almost had
like it wasn't that good
but compared to everything else at the
time it was
amazing
and by the way the first
phone that had an incredible camera
that could that could like do as well as
the point and shoot you might carry
around
was the iphone 4 and that was right when
instagram launched and we looked around
and we said what will change because
everyone has a camera in their pocket
and it was so clear to me that the the
world of
social networks before
it was based in the desktop and sitting
there and having a link you could share
right
and that wasn't going to be the case so
the question is what would you share if
you were out and about in the world
if not only did you have a camera that's
in your pocket but by the way that
camera had a network attached to it that
allowed you to share instantly
that seemed revolutionary and a bunch of
people saw it at the same time it wasn't
just instagram there were a bunch of
competitors
the thing we did
i think was not only well we focused on
two things so we wrote down those things
we circled photos and we said i think we
should invest in this
but then we said what sucks about photos
one they look like crap right they just
at least back then now
my phone takes pretty great photos right
um back then they were blurry not so
great compressed right
two
uh it was really slow
like really slow to upload a photo
and i'll tell a fun story about that and
explain to you why they're all the same
size and square as well
um and three
man if you wanted to share a photo on
different networks you had to go to each
of the individual apps and select all of
them and upload individually and
so we're like all right those are the
pain points we're gonna focus on that so
one instead of
because they weren't beautiful um we
were like why don't we lean into the
fact that they're not beautiful and i
remember studying in florence my
photography teacher gave me this whole
gay camera and i'm not sure everyone
knows what a whole gay camera is but
they're these old-school plastic cameras
i think they're produced in china at the
time
and they're i want to say the original
ones like from the 70s or the 80s or
something they're supposed to be like
three dollar cameras for the every
person
they took nice medium format films large
large
negatives
but they kind of blurred the the the
light and they kind of like light leaked
into the side and there was this whole
resurgence where people looked at that
and said oh my god this is a style right
and i remember using that in florence
and just saying well why don't we just
like lean into the fact that these
photos suck and make them suck more
but in an artistic way
and it turns out that had product market
fit people really liked that they were
willing to share their not so great
photos if they
looked
not so great on purpose okay
it's the second part
that's the where the filters come into
the picture yeah so computational
modification of photos to make them look
extra crappy to where it becomes art
yeah yeah and
i mean
add light leaks add like an overlay
filter make them more contrasty than
they should be uh the first
filter we ever produced was called x-pro
2. and i designed it while i was in this
small little bed and breakfast room in
total santos mexico i was trying to take
a break from the
the bourbon days and i i remember saying
to my co-founder i just need like a week
to reset
and that was
on that trip worked on the first filter
because i said you know i think i can do
this and i literally iterated
one by one over the rgb values in the
array that was the photo
and just slightly shifted basically
there was a function of our function of
g function of b
that just shifted them slightly it
wasn't rocket science um and it turns
out that actually made your photo look
pretty cool
it just mapped from one color space to
another color space
it was simple but it was really slow i
mean if you applied a filter
i think it used to take two or three
seconds to render
only eventually would i figure out how
to do it on the gpu and i'm not even
sure it was gpu but was using opengl but
anyway um i would eventually figure that
out and then it would be instant but it
used to be really slow by the way anyone
who's watching or listening
it's amazing what you can get away with
in a startup as long as the product
outcome is right for the user like you
can be slow you can be terrible you can
be
as long as you have product market fit
people will put up with a lot and then
the question is just about compressing
making it more performant over time so
that they get that product market fit
instantly
so fascinating because there's some
things where
those three seconds would
make or break the app
but some things you're saying not it's
hard to know when you know it's what
it's the problem spotify solved
making streaming like
work
and
like delays in listening to music is a
huge negative
even like slight delays
but here you're saying i mean how do you
know
when those three seconds are okay are
you just gonna have to
try it out because to me my intuition
would be
those three seconds would kill the app
like i would try to do the opengl thing
right
so i wish i were that smart at the time
um
i wasn't i just knew how to do what i
knew how to do right
and i decided okay like why don't i just
iterate over the values and change them
and what's interesting
is that um
compared to the alternatives no one else
used opengl
right so everyone else was doing that
the dumb way and in fact they were doing
it at a high resolution now comes in the
small resolution that we'll talk about
for a second
um
by choosing 512 pixels by 512 pixels
which i believe it was at the time
we iterated over a lot fewer pixels than
our competitors who were trying to do
these enormous output like images yeah
so instead of taking 20 seconds i mean
three seconds feels pretty good right
so on a relative basis we were winning
like a lot
okay so that's answer number one answer
number two is uh
we actually focused on latency in the
right places so we did this really
wonderful thing um
when you uploaded so uh the way it would
work is you know you'd take your phone
you'd take the photo and then you'd go
to the um
you'd go the edit screen where you would
caption it
and on that caption screen you start
typing you think okay like what's a
clever caption and and i said to mike
hey when i worked on the gmail team you
know what they did when you typed in
your username or your email address
even before you've entered in your
password like the chat probability once
you enter in your username that you're
going to actually sign in is extremely
high
so why not just start loading your
account in the background not not like
sending it down to the desktop that
would be a security uh uh
issue but like load it into memory on
the server like get it ready prepare it
i always thought that was so fascinating
and unintuitive i was like mike why
don't we just do that but like we'll
just upload the photo and like assume
you're gonna upload the photo and if you
don't
forget about it we'll delete it right
so what ended up happening
was people would caption their uh photo
they'd press done or upload
and you'd see this little progress bar
just go
it was lightning fast okay
we were no faster than anyone else at
the time but by choosing 512 by 512 and
doing in the background it almost
guaranteed that it was done by the time
you captioned
and everyone when they used it was like
how the hell is this thing so fast
but we were slow we just hid the the
slowness it wasn't like
these things are just like it's a shelly
game you're just hiding the latency
that
that mattered to people like a lot and i
think that so you were willing to put up
with a slow filter if it meant you could
share it immediately and of course we
added sharing options which let you
distribute it really quickly that was
the third part
um
so latency matters but relative to what
and then
there's some like tricks you can get
around to just hiding the latency um
like i don't know if spotify starts
downloading the next song eagerly i'm
assuming they do
there are a bunch of ideas here that are
not rocket science that
that really help
and all of that was
stuff you were explicitly having a
discussion about like
those designs and argument you were
having like arguments discussions
uh i'm sure it was arguments i mean
i'm not sure if you've met my co-founder
mike but he's a pretty nice guy and he's
very reasonable and uh
and we both just saw eye to eye and
we're like yeah just like
make this fast early seem fast
it'll be great i mean honestly i think
the most contentious thing and he would
say this too initially
was um i was on an iphone 3g so like the
the not so fast one and he had a brand
new iphone 4. i was cheap nice um
and his feed loaded super smoothly like
when he would scroll from photo to photo
buttery smooth right but on my phone
every time you got to a new photo it was
like
a chunk allocate memory like all this
stuff right
i was like mike that's unacceptable he's
like oh come on man just like upgrade
your phone basically you didn't actually
say that it's nicer than that um
but i could tell he wished like i would
just stop being cheap and just get a new
phone
but what's funny is we actually sat
there working on that little detail for
a few days before launch and
that polished experience plus the fact
that uploading seemed fast for all these
people who didn't have nice phones
i think meant a lot
because
far too often you see teams focus not on
performance
they focus on
what's the cool computer science problem
they can solve
right can we scale this thing to a
billion users and they've got like 100
right
yeah
you talked about loss function so i want
to come back to that
but like the loss function is like do
you provide a great happy magical
whatever experience for the consumer
and listen if it happens to involve
something complex and technical then
great
but it turns out i think most of the
time
those experiences are just sitting there
waiting to be built with like not that
complex solutions
uh but everyone is just like so stuck in
their own head that they have to over
engineer everything and then they forget
about the easy stuff
i mean also maybe to flip the lost
function there is you're trying to
minimize the number of times
you there's unpleasant experience right
like uh the one you mention where when
you go to the next photo it freezes for
a little bit so it's almost as opposed
to maximizing pleasure it's probably
easier to minimize
the number of like
the friction
yeah and as we all know you just you
just uh you just make the pleasure
negative and then minimize everything so
we're mapping this all back to neural
networks but actually can i say one
thing on that which is
i don't know a lot about machine
learning but i feel like i've i've tried
studying a bunch that whole idea of
reinforcement learning
and planning out more than the greedy
single experience i think is
is the closest you can get to
like ideal product design thinking
where you're not saying hey like can we
have a great experience just this one
time
but like what is the right way to
onboard someone what series of
experiences correlate most with them
hanging on long term right so not just
saying oh did the photo load slowly a
couple times or did they get a great
photo at the top of their feed
but like what are the things that are
going to make this person come back
over the next week over the next month
and as a product designer asking
yourself okay i want to optimize not
just minimize bad experiences in the
short run but like
how do i get someone to engage over the
next month
and i'm not going to claim at all that i
thought that way at all at the time
because i certainly didn't but if i were
going back and giving myself any advice
it would be thinking what are those what
are those second order effects that you
can create
and it turns out having your friends on
the service
it's an enormous win so starting with a
very small group of people that produce
content that you wanted to see which we
did we seeded the community very well i
think
ended up mattering and
so
yeah you said that community is one of
the most important things so it's from a
metrics perspective from uh maybe a
philosophy perspective
building a certain kind of community
within the app see i wasn't sure what
exactly you meant by that when when i've
heard you say that maybe you can
elaborate but as i understand now it's
can literally mean
get your friends onto the app
yeah
think of it this way
you can build an amazing restaurant or
bar or whatever right
but if you show up and you're the only
one there is it like does it matter how
good the food is
the drinks whatever um no um these are
inherently social experiences that we
were working on
so
the idea of having people there
like you needed to have that otherwise
it was just to filter out but by the way
part of the genius
i'm going to say genius even though i
wasn't really genius was starting to be
marauding as a filter app
was awesome
the fact that you could so we talk about
single player mode a lot which is like
can you play the game alone
and instagram you could totally play
alone you could filter your photos and a
lot of people would tell me i didn't
even realize that this thing was a
social network
until my friend showed up it totally
worked as a single player
game and then when your friend showed up
all of a sudden it was like oh
not only was this great alone but now i
actually have this trove of photos that
people can look at and start liking and
then i can like theirs and
so it was this bootstrap method of how
do you make the thing not suck when the
restaurant is empty
yeah but the thing is when you say
friends i mean we're not necessarily
referring to friends in the physical
space so you're not bringing your
physical friends with you you're also
making new friends so you're finding new
community so it's not immediately
obvious to me that
it's like it's almost like building any
kind of community
it was it was both and what we learned
very early on was what made instagram
special and the reason why you would
sign up for it versus say just sit on
facebook and look at your friends photos
of course we were live and of course it
was interesting to see what your friends
were doing now
but the fact that you could connect with
people who like took really beautiful
photos in a certain style all around the
world whether they were travelers it was
the beginning or beginning of the
influencer economy there's these people
who became professional instagramers way
back when right
um
but they took these amazing photos and
some of them were photographers right um
like professionally
and all of a sudden you had this moment
in the day when you could open up this
app and sure you could see what your
friends were doing but also it was like
oh my god that's a beautiful beautiful
waterfall or oh my god i didn't realize
there was that corner of england or like
really cool stuff
um
and the beauty about instagram early on
was that it was
international by default you didn't have
to speak english to use it right you
could just look at the photos
worked great
we did translate we had some pretty bad
translations but we did translate the
app
and uh
you know even if our translations were
pretty poor the the idea that you could
just connect with other people through
their images was pretty powerful
how much uh technical difficulties there
with the programming like what
programming language you were talking
about what was zero i'd like maybe it
was hard for us but um i mean we
there was nothing the only thing that
was complex about instagram at the
beginning
technically was making it scale and we
were
just plain old objective c for the
client uh so it was iphone only
yep as an android person i'm deeply
offended but go ahead again come on this
was 2010. oh sure sure sorry android's
getting a lot better yeah yeah so
um i take it back you're right if i were
to do something today i think it would
be very different in terms of launch
strategy right android's enormous too uh
but anyway um back to that moment it was
objective c uh and then
we were python based uh which is just
like this is before python was really
cool like now it's cool because it's all
these machine learning libraries like
support python and right
now it's super
now it's like cool to be by the back
then it was like oh google uses python
like maybe you should use python
facebook was php like
i had worked at a small startup of some
ex-googlers that used python so we used
it and we used a framework called django
uh still exists and people use for
basically
the back end and then you threw a couple
interesting things in there i mean we
used postgres which was kind of fun it
was a little bit like hipster database
at the time right
my sequel my sequel like everyone used
my sequel so like using postcards was
like an interesting decision right uh
but we used it because it had a bunch of
uh geo features built in because we
thought we were going to be a checking
out pretty much it's also super cool now
so you were into python before it was
cool and you were into postgres before
it was cool yeah we were basically like
not only hipster
hipster photo company hipster tech
company right uh we also adopted redis
early and like loved it i mean it solved
so many problems for us
and turns out that's still pretty cool
but the programming was very easy it was
like sign up a user have a feed there
was nothing
no machine learning at all zero can you
get some context how many users at each
of these stages are we talking about 100
users a thousand users so the stage i
just described i mean that technical
stack lasted through
probably 50 million users
um
i mean seriously like you can get away
with a lot with with a pretty basic
stack um like i think a lot of startups
try to over engineer their solutions
from the beginning to like really scale
and you can get away with a lot
that being said most of the first two
years of instagram was literally just
trying to make that stack scale and it
wasn't it was it was not a python
problem it was like
literally just like where do we put the
data like it's all coming in too fast
like how do we store it how do we make
sure to be up how do we like
how do we make sure we're on the right
side of boxes that they have enough
memory um those were the issues but can
you speak to the choices you make at
that stage when you're growing so
quickly
do you use something like somebody
else's computer infrastructure or do you
build in-house
i'm only laughing because we when we
launched we had a single
computer that we had rented
in some colo space in la i don't even
remember what it was called
because i thought that's what you did
when i worked at a company called odio
that became twitter i remember visiting
our space in san francisco you walked in
you had to wear the ear things and it
was
cold and fans everywhere right
and we had to you know plug one out
replace one and i was the intern so i
just like held things
but i thought to myself oh this is how
it goes and then i remember being in a
vc's office
i think it was benchmark's office and i
think we ran into another entrepreneur
and they were like oh how are things
going we're like uh you know trying to
scale this thing
and they were like well i mean can't you
just add more instances and i was like
what do you mean
and they're like instances on amazon i
was like what are those
and it was this moment where we realized
how deep in it we were because we had no
idea that aw aws existed nor should we
be using it anyway
that night we went back to the office
and we got on aws but we we did this
really dumb thing we're
i'm so sorry to people listening but um
we brought up an instance which was our
our database it's going to be a
replacement for our database
but we had it talking over the public
internet to our little box in la that
was our app server very nice yeah um
that's how sophisticated we were and
obviously that was very very slow
didn't work at all i mean it worked but
didn't work did we only like later that
night did we realize we had to have it
all together
but at least like if you're listening
right now and you're thinking you know i
have no chance i'm going to start to
start i have no chance
i don't know we did it and we made a
bunch of really dumb mistakes initially
i think the question is how quickly do
you learn that you're making a mistake
and do you do the right thing
immediately right after so you didn't
pay for those mistakes by you know by
failure so
yeah how quickly did you fix it
i guess there's a lot of ways to sneak
up to this question of how the hell do
you scale the thing
other startups if you have an idea how
do you scale the thing is this is just
aws
and uh
you try to write the kind of code that's
easy to spread across a large number of
instances and then
the rest is just put money into it
basically i would say a couple things
first off
don't even ask the question just find
product market fit
duct tape it together right like if you
have to i think there's a big caveat
here which i want to get to
but generally all that matters is
product market fit that's all that
matters if people like your product
do not worry about when 50 000 people
use your product because you will be
happy that you have that problem when
you get there i actually can't name
many
startups
where they go from
nothing to something overnight and they
can't figure out how to scale it there
are some
but i think nowadays it's a when i say a
solved problem like
there are ways of solving it
the base case is typically that startups
worry way too much about
scaling way too early and forget that
they actually have to make something
that people like that's the that's the
default mistake case
but what i'll say is um
once you start scaling
i mean hiring quickly people who have
seen the game before and just know how
to do it it it becomes um
it becomes a bit of like yeah just throw
instances of the problem right
but the last thing i'll say on this that
i think did save us um
we were pretty rigorous about writing
tests uh from the beginning
that helped us
move very very quickly when we wanted to
rewrite parts of the product
and know that we weren't breaking
something else
tests are one of those things where it's
like you go slow to go fast
and they suck when you have to write
them because you have to figure it out
and
they're always those ones that break
when you don't want them to break and
they're annoying and it feels like you
spent all this time but
looking back i think that like long-term
optimal even with the team of four
it allowed us to move very very quickly
because anyone could touch any part of
the the the product and know that they
weren't going to bring down the site or
at least in general at which point do
you know product market fit how many
users would you say what is it all it
takes is like 10 people or is it a
thousand is it 50 000
i don't think it is
generally a question of absolute numbers
i think it's a question of cohorts and i
think it's a question of trends so
you know it depends how big
your business is trying to be right but
if i were signing up a thousand people a
week and they all retain like the
retention curves for those cohorts
looked good healthy
and even like
as you started getting more people on
the service maybe those earlier cohorts
started curving up again because now
there are network effects and their
friends are on the service or totally
depends what type of business you're in
but i'm talking purely social right
um
i don't think it's an absolute
number i think it is a i guess you could
call it a marginal number so i spend a
lot of time when i work with startups
asking them like okay have you looked at
that cohort versus this cohort whether
it's your clients or whether it's people
signing up for uh the service
but a lot of people think you just have
to hit some mark like 10 000 people or
50 000 people
but really
seven-ish billion people in the world
most people forever will not know about
your product there are always more
people out there to sign up it's just a
question of how you turn on the spigot
so
at that stage early stage
yourself but also by way of advice
should you worry about money at all how
this thing is going to make money or do
you just try to find product market fit
and get a lot of users to enjoy using
your thing
i think it totally depends and that's an
unsatisfying answer um i was talking
with a friend today
who
he was one of our earlier investors and
he was saying hey like have you been
doing any angel investing lately i said
not really i'm just like focused on what
i want to do next and
he said the number of financings have
just gone bonkers like just bonk
like people are throwing money
everywhere right now um
and
i think the question is
do you have an inkling of how you're
gonna make money
or are you really just like waving your
hands i would not like to be an
entrepreneur in the position of
well i have no idea how this will
eventually make money that's not fun um
if you are in an area like let's say you
wanted to start a social network right
not saying this is a good idea but if
you did
they're only a handful of ways they've
made money and really only one way
they've made money in the past and
that's ads
so
you know
if you
have a service that's amenable to that
and
then i wouldn't worry too much about
that because if you get to the scale you
can hire some smart people and figure
that out
i
do think that is really healthy for a
lot of startups these days especially
the ones doing like
enterprise software
slacks of the world etc to be worried
about money from the beginning but
mostly as a way of winning over clients
and having stickiness um
i think i like of course you need to be
worried about money but i'm going to
also say this again which is
it's like long-term profitability
if you have a roadmap to that then
that's great
but if you're just like i don't know
maybe never like we're working on this
meta first thing i think maybe someday
i don't know like that seems harder to
me um so you have to be as big as
facebook to like finance that bet right
do you think it's possible you said
you're not saying it's necessarily a
good idea to launch a social network
do you think it's possible
today
maybe you can put yourself in those
shoes
to launch a social network that achieves
the scale of a facebook
or a twitter or an instagram and maybe
even greater scale absolutely
how do you do it
asking for a friend yeah if i knew i i'd
probably be doing it right now and not
sitting here so
i mean there's a lot of ways to ask this
question one is create a totally new
product market fit create a new market
create something like instagram did
which is like create something kind of
new
or
literally out compete facebook at its
own thing or i'll compete twitter at its
own thing
the only way to compete now if you want
to build a large social network is to
look for the cracks look for the
openings
um
you know
no one competed
i mean no one competed with the core
business of google no one competed with
the core business of microsoft
you don't go at
the big guys doing exactly what they're
doing
instagram didn't win quote unquote
because it tried to be
a visual twitter like
we spotted things that either twitter
wasn't going to do or refused to do
images and feed for the longest time
right
or that facebook wasn't doing or not
paying attention to because they were
mostly desktop at the time and we were
purely mobile
purely visual
often there are opportunities sitting
there you just have to you have to
you have to figure out like uh i think
like there's a strategy book i can't
remember the name but talk about moats
and just like the best place to play
is where your competitor like literally
can't pivot because structurally they're
set up not to be there
and that's where you win um
and what's fascinating is like do you
know how many people are like images
facebook does that twitter does that i
mean how wrong were they really wrong
these are some of the smartest people in
silicon valley right
but now instagram exists for a while
how is it that snapchat could then exist
makes no sense
like plenty of people would say well
there's facebook no images okay okay i
mean instagram i'll give you that one
but wait now another image based social
network's gonna get really big
and then tick tock comes along
like
the prior so you asked me is it possible
the only answer and reason i'm answering
yes is because
my prior is that it's happened once
every i don't know three four or five
years
consistently and i can't imagine there's
anything structurally that would change
that
so that's why i answer that way not
because i know how i just
when you see a pattern you see a pattern
and there's no reason to believe that's
going to stop and it's subtle too
because like you said snapchat and tick
tock they're all doing the same space of
things but there's something
fundamentally different about
like a three second video and a five
second video and a 15 second video in a
one minute video and a one hour video
right like fundamentally different
fundamentally different i mean i think
one of the reasons snapchat exists is
because instagram was so focused on
posting great beautiful manicured
versions of yourself throughout time
and there was this enormous demand of
like hey i really like this behavior i
love using instagram but
man i just like wish i could share
something going on in my day like
do i really have to put it on my profile
do i really have to make it last forever
do i really
um and that opened up a door it created
a market right and then what's
fascinating is
instagram had an explore page for the
longest time it was image driven right
um
but there's absolutely a behavior where
you open up instagram and you sit on the
explore page all day that is effectively
tick tock but obviously focused on
videos and it's not like you could just
put the explore page in tik tok form and
it works it had to be video it had to
have music
these are the hard parts about product
development that are very hard to
predict
but um
they're all versions of the same thing
with varying
if you line them up in a bunch of
dimensions they're just like
kind of on
they're different values of the same
dimensions which is like i guess easy to
say in retrospect but like if i were an
entrepreneur going after that area i'd
ask myself like where's the opening
what needs to exist because tiktok
exists now
so i wonder how much
things that don't yet exist and can
exist is in the space of algorithms in
the space of recommender systems
so
in the space of how the feed is
generated so we kind of talk about the
actual elements of the um
the content that's what we've been
talking the difference between photos
between uh short videos longer videos i
wonder how much disruption is possible
in the way the algorithms work
because a lot of the criticism towards
social media is in the way the
algorithms work currently and it feels
like
first of all talking about
product market fit there's certainly a
hunger
for
um social media
algorithms that do something different i
don't think anyone everyone said
complaining this is not doing this is
this is hurting me and this is hurting
society but i keep doing it because i'm
addicted to it
and they say we want something different
but we don't know what it feels like a
uh
just different uh it feels like there's
a hunger for that
but that's in the space of algorithms i
wonder if it's possible to disrupt in
that space absolutely
um
i have this thesis that
the worst part about social networks is
that they're uh
is the people
it's it's
it's a line that sounds funny right
because like that's why you call it a
social network um but what does social
networks actually do for you like just
think you know like
imagine you were an alien and you landed
and someone says hey there's this site
it's a social network we're not going to
tell you what it is but just what does
it do and you have to explain it to them
it does two things one is that
people you know and have social ties
with
uh distribute updates through whether
it's uh you know photos or videos
about their lives so that you don't have
to physically be with them but you can
keep in touch with them that's one
that's like a big part of instagram
that's a big part of snap
it is not part of tick tock at all so
there's another big part which is
there's all this content out in the
world that's entertaining
whether you want to watch it or you want
to read it
um
and matchmaking between content that
exists in the world and
uh people that want that content turns
out to be like a really big business
right search and discovery would you
search and discovery but my point is it
could be video it could be text it could
be websites it could be i mean think
back to um
think back to like dig right or stumble
upon or
right
nice
yeah but like what did those do like
they basically distributed interesting
content to you right
um
i think the most interesting part or the
future of social networks is going to be
making them less social because i think
people are part of the root cause of the
problem so for instance
um often in recommender systems we talk
about two stages there's a candidate
generation step
which is just like of our vast trove of
stuff that you might want to see
what small subset
should we pick for you
okay
typically that is grabbed from things
your friends have shared
right
then there's a ranking step which says
okay now given these hundred 200 things
depends on the network right let's like
be really good about ranking them and
generally rank the things up higher that
get the most engagement right so what's
the problem with that
step one is we've limited everything you
could possibly see to things that your
friends have chosen to share
or maybe not friends but influencers
what things do people generally want to
share they want to share things that are
going to get likes that are going to
show up broadly
so they tend to be more emotionally
driven they tend to be more risque or
whatever so why do we have this problem
it's because
we show people things people have
decided to share and those things
self-select to being the things that are
most divisive
so how do you fix that
well
what if you just
imagine for a second that why do you
have to grab things from things your
friends have shared why not just like
grab things
that's really fascinating to me and
that's something i've been thinking a
lot about and just like
you know why is it that when you log on
to twitter
you're just sitting there looking at
things from accounts that you've
followed for whatever reason
and tick tock i think has done a
wonderful job here which is like you can
literally be anyone
and if you produce something
fascinating it'll go viral
but like
you don't have to be someone that anyone
knows you don't have to have built up a
giant following you don't have to have
paid for followers
you don't have to try to maintain those
followers you literally just have to
produce something interesting
that is i think the future of social
networking that's the that's the
direction things will head and i think
what you'll find is it's far less about
people manipulating distribution and far
more about what is like is this content
good
and good is obviously a vague definition
that we spend hours on but
different networks i think will decide
different value functions to decide what
is good and what isn't good and i i
think that's a fascinating direction so
that's almost like creating an internet
i mean that's what google did for web
pages
they did the
you know page rank search
so discovery you don't you don't follow
anybody on google when you use a search
engine you just discover web pages and
so what tick tock does
is saying let's start from scratch
let's like like start a new internet and
have people discover stuff on that new
internet within a particular kind of
pool of people well what's so
fascinating about this is like the
the um field of information retrieval
like i always talked about and as i was
studying this stuff they would always
use the word query and document so i was
like why are they saying query
undocuments like they're literally
imagine like
if you just stop thinking