00:00:00
Feeling good? Yeah. You? I think so.
00:00:08
Hey, what's up, people of the internet? Welcome back to a bonus episode of the Waveform Podcast. In this episode, I'm
00:00:14
going to be sitting down and interviewing one-on-one the CEO of Instagram, Adam Seri. This is a really
00:00:20
interesting interview. I'm coming at it from the perspective of a creator. I upload to a bunch of different platforms, which I'm totally transparent
00:00:26
about. You guys know me as a YouTuber, but I also make shorts and I make Instagram content and then also Tik Tok
00:00:33
stuff here and there. And so I do have a lot of thoughts on each individual platform. And I thought it would be
00:00:38
interesting to ask him some of my questions about how Instagram has come to be what it is today, where it came
00:00:44
from, where they're going into the future, and pay careful attention to the way he answers the same question both at
00:00:51
the beginning of the interview about what Instagram is and at the end of the interview because it's a different
00:00:57
answer. And I think it's really interesting to see the framing and the way we talk about the platform. And yes, we talk about paying creators, too. So
00:01:03
stay tuned for all of that. Without any further ado, here's your bonus episode interviewing the CEO of Instagram, Adam.
00:01:11
All right, Adam, thanks for joining me. Thanks for having me. Yeah, I appreciate the time. Okay, so you're the CEO of Instagram and and I'm
00:01:18
a creator. I I work on a bunch of different platforms. I know what I think Instagram is, but it'll be cool hearing your own words. What is Instagram today?
00:01:26
So, it's changed a lot. Um I for for better or for worse, I got a lot of
00:01:31
feedback and I think that's fundamentally a good thing. Um, but at the heart of what we do has always been to try to bring people together over
00:01:37
creativity. You know, we started out by making it so that anybody with an idea could make a photo that they were having
00:01:43
that they were confident sharing with those pretty intense filters and borders. Remember them?
00:01:49
Yeah. No, it's pretty pretty pretty opinionated in retrospect. Um, but how people connect with their
00:01:55
friends has changed. How they communicate has changed. How they entertain themselves has changed. And so we've had to evolve. Um, but at the
00:02:01
heart has always been these two ideas of connecting with the people you care about over your interests. Connecting with the people you care
00:02:07
about. I mean, it's a social network obviously, but it it has both chronological and algorithmic ways of
00:02:14
sharing stuff with you, sharing stuff between people. Um, and there's a lot of features that have been added on top
00:02:19
over the years, which is why I wonder how you think about how to add a new feature to Instagram. You probably get
00:02:25
tons of suggestions and ideas about things to add, things not to add. Yes. If things are removed, exactly. What's
00:02:31
your filter like? How do you decide like, okay, we were just photos this many years ago. Now it's all these things. How do you decide what to add to
00:02:37
Instagram? I think if it's it has to be connected to, you know, creativity or connecting
00:02:44
people. Um, but then it also has to be something that is going to be cared
00:02:50
about or we think used by enough people that it'll matter because obviously the app is getting more and more complicated. I think we need to do more
00:02:56
to actually simplify things and collapse some of the patterns into less patterns. Um, but we also have to evolve. The
00:03:03
biggest risk of any platform of our scale is that the world changes and we failed to adapt and we slowly become
00:03:09
irrelevant. People think of Instagram as a feed of square photos with those borders.
00:03:14
Yeah. You know, but if you look at the last 10 years, you know, if we didn't have stories, if we didn't have DMs, if we
00:03:20
didn't have reals, we wouldn't be talking today. uh we wouldn't be an important enough platform
00:03:26
to really be a platform that you focused on. And so we've had to figure out how to evolve um and try to stay true to
00:03:32
those principles and then try to also keep the app simple. Sometimes we do better or worse at one or the other.
00:03:38
Sounds like a balancing act because I mean keeping it simple is on one end of the spectrum and keeping up with all the
00:03:43
competition and all the evolving things happening in the world is another. Yes. Are you leaning towards one end or the
00:03:49
other? I mean you have a lot of competition now. Yeah. I make videos on five platforms that accept video uploads. Like how do
00:03:55
you think about more and more creat? Yeah. And I used to never do that but you know that's a slightly different language that I speak
00:04:01
to each of those audiences. Different audiences. Yeah. So depending on the year we'll focus
00:04:06
more on building new things or simplifying. But in general my bias is going to be to make sure that we are
00:04:13
taking risks. I think that what happens a lot with large companies is they slowly go down hitting their goals the
00:04:20
whole way down. And I would much rather look back at 5 10 years of working on
00:04:25
the app and have a couple examples of where we pushed too hard or went too fast and it kind of um we got some
00:04:32
blowback for it then the alternative which I see is primarily like you never
00:04:37
upset anybody but you also didn't evolve fast enough and you just matter a lot less. Yeah. So, I'm going to bias and I think
00:04:44
we as a culture at Instagram bias towards trying to take some bigger bets and trying to take on some risk and
00:04:50
sometimes they will pan out and sometimes they will be painful. Fair enough. I I definitely feel like
00:04:55
that applies in a lot of different genres. It applies as a creator. If I just do the same thing over and over,
00:05:02
I'll be happy because I'm achieving the goal of doing the same thing, but I'll probably get left behind. And in other
00:05:07
product categories, that's the same thing. Like you can make the same product over and over but then your competition will pass you.
00:05:12
Yeah. U but when you think about those risks you know there are some that you decide not to take obviously and there are some
00:05:18
that you decide are worth the challenge of like p pushing yourself forward. Yeah. Is there some sort of filter you run
00:05:24
through to decide like okay this would be cool but I don't think it's the right direction for Instagram. Yeah. There's a couple different things.
00:05:32
One thing I talk about a lot internally with my team is like it it it needs to matter. And for for something to matter,
00:05:38
it either has to be like novel and interesting or it has to actually like meaningfully move the app forward or the
00:05:45
business forward. If it's a small change, but it's not interesting, it's not novel and it's not going to really
00:05:50
change how people use the app, then it's probably just noise. Uh it also has to be connected to what we do. Like I like
00:05:57
we're going to for instance, you know, I'm sure we'll talk about AI today. We're going to try to figure out in what
00:06:02
ways AI can help empower creators, empower creatives, connect people as opposed to all the other things that AI
00:06:07
can also do. So it has to be connected to our mission and our reason to exist. Um, but you also have to decide how many
00:06:15
like so like I'll give you an example that's maybe a little bit less obvious from the outside. Usually we're talking
00:06:20
about how many you know we're talking more about how many new features to build. But when you're one of the most important things about Instagram is how
00:06:26
ranking works. Mhm. And there's a couple, you could call them bets that we are making that don't
00:06:33
maximize the amount of time people spend on Instagram or revenue we make or engagement that we have. So we focus a
00:06:39
lot for instance on originality. Making sure that we try and drive as much traffic or distribution to the original
00:06:45
content creator versus an aggregator. Yeah. That actually leaves engagement and
00:06:50
revenue on the table. But the bet is over time you will build more of a rapport with the creative community and
00:06:57
that that'll benefit in some indirect way over years. Yeah. Timeliness. Similarly, if you want to be
00:07:03
culturally relevant, things need to break quickly. But sometimes you're going to drive more engagement by
00:07:08
showing something that was really funny that's 5 days old. Yep. So, how many of those bets can you take on at once is another thing that we have
00:07:14
to juggle. Interesting. And I think what a lot of people also wonder is how much of that comes from meta. Obviously, you're a
00:07:20
part of Meta, but to some degree independent. How much of this comes from Meta's goals and Meta's decision-making
00:07:28
in budgets versus like what Instagram independently thinks that they should be doing? It's a balance uh depending on the year,
00:07:35
but in general, there are a few ways in which we try to make sure we're supporting the broader company. We also
00:07:41
try to make sure that we're leveraging all of the good work and technology from the people across the company. So we
00:07:48
build on their infra build on meta's infrastructure. It's meta's ad system. We use meta's safety technologies. Um
00:07:54
but then try to also carve out our own space right we're more focused on creative expression particularly visual
00:08:01
creative you know threads which we also work on is more focused on ideas and perspectives. Um Facebook is more of a
00:08:10
social marketplace for all things. So we try to make sure that we are cognizant
00:08:15
of the rest of the company when they when there are ways for us to contribute back that we do but that we don't lose
00:08:21
sight of um what our own identity is. So in 2022,
00:08:27
okay, I tweeted about how there should be an iPad app for Instagram. And you may remember this because you
00:08:33
replied to this. I do remember and I I remember reading that reply. I don't know if I have it exactly in front of me, but it was clearly not a priority
00:08:39
at the time. There was not um there is one now. There is what changed?
00:08:45
So I mean a couple different things. One is in general we always wanted to build
00:08:51
one but it wasn't the next most valuable thing we thought we could do with our time. Two is I think that we saw an
00:09:01
opportunity to try and lean into where we think Instagram is going over time. So, you know, more and more of Instagram
00:09:09
time spent is watching short videos, watching realels. Um, people it's they're actually some
00:09:15
they're actually pretty social. People send reels to their friends all the time and start conversations. That's where the iPad app starts now.
00:09:21
Yeah. So, the idea was in a world where messaging and reals are driving our
00:09:26
growth, how could we reorient the app more around those concepts? And then if you look at how people consume video,
00:09:33
it's usually they consume video less times per day than other things, but for
00:09:38
longer each time. So less sessions a day, longer duration per session. Okay? And that's also the shape of tablet
00:09:45
usage. People use tablets less times per day, but longer per time than their phone, for instance. So, we thought it
00:09:52
would also be an interesting opportunity to try a real's first version of the app
00:09:58
um without all of the challenges that come along with moving people from one
00:10:03
version of the experience to another. I I kind of wonder if that was it because you they've tested a real first version of the app in some other places,
00:10:11
but this was a new version of the app. So, it's not like people on the iPad are switching from feed to reals.
00:10:16
Yeah. Um, they're switching from if you had it like the tiny little mobile phone
00:10:22
version. Yeah. Yeah. Yeah. Um, and it's interesting also that, you know, I was asking for an app
00:10:27
for an iPad and I I wonder how much you listen to user feedback about what they
00:10:32
hope to see in Instagram or what they wish would come back from old Instagram or whatever it is. As a creator, I deal with this all the time. I have to have a
00:10:38
filter for like, okay, I know people will always want the old thing, but I have to have a balance of moving forward as well, like you talked about. Do you
00:10:46
listen to or factor in a lot of user like like verbal feedback like Adam
00:10:51
bring back this Adam stop doing this? Are you listening to that at all? For sure. I mean I get into my comments.
00:10:57
I get into my DMs. I get into that requests folder on a weekly basis and I try to find the signal uh and try
00:11:05
to not get overly focused on the noise. But I also try to understand what's the root thing someone's asking for. Um, so
00:11:11
sometimes you're asking for a specific feature, but there's like a thing that Instagram isn't as good at that it used
00:11:17
to be, you know, at a lower level and then where I can where we can meet that
00:11:23
need in some other way without inventing yet a new thing or going backwards, we try and do that. But I do think it's
00:11:29
also just true that what works changes. I mean, actually curious for you um cuz I've even experienced this on cuz
00:11:35
I actually post on Instagram, you know, so I get a little bit of um empathy for what it's like to try to be successful
00:11:42
on the platform. Okay. Uh I actually had my my followers start to decline like a year and a half ago and I was like, "Okay, I got to rethink
00:11:48
the investigation." Yeah. Yeah. But so for you like have you found that certain things sure people
00:11:53
are asking for you know the version from a couple years ago but have you found that certain constructs or series you do
00:12:00
stopped working and you needed to reinvent them? I think in the past a little bit more. I
00:12:06
think now the priority for me has been to abandon something before it trickles
00:12:12
out and dies. Yeah. So we have series that we do. We have like almost recurring videos every year.
00:12:17
You know, I'm going to unbox the new iPhone and review the new iPhone. Yeah. At some point, eventually, we're going
00:12:24
to move on from that type of video, and I want to be ahead of that rather than behind it. Yeah. Like a TV show getting out before
00:12:30
it gets before you get cancelled for not having, you know, the things people looking forward to. But it is interesting. And I
00:12:36
also think yeah there's some amount of noise where you know I I hear people in
00:12:41
the comments asking for certain things but I have to sort of interpret read between the lines and understand what they really want is something more
00:12:48
fundamental. Um what's the new Instagram iPad app now? What's the new most commonly asked
00:12:54
for feature that is simply not a priority right now? Oo that's a really good question. So the
00:12:59
most requests I get aren't for new features. They're like verify me. Uh it's my birthday is my favorite version of that. It's my birthday. All I
00:13:05
want for my birthday is a blue check mark. Interesting. I get that a lot. Uh a lot of people who disagree with content decisions. If some content was
00:13:12
taken down and they feel like they're censored, we want to make sure that we u provide feed um recourse there. What's a
00:13:19
feature that's asked for for that that's not a priority right now? I know a lot of I see a lot of people saying prioritize photos again.
00:13:27
Yeah. And I just feel like obviously that's where Instagram came from. Yeah. But it may feel feel like it's going
00:13:33
backwards. There's a lot of photos first creators who feel like they're being left behind by focusing on reals a lot.
00:13:39
But we've tried, particularly with carousels and being able to add music to photos and music to carousels to give
00:13:45
people who don't want to make reals a way to still create content that's going to be engaging. We're not trying to push
00:13:52
video because we have some interest in video. Actually, video in a lot of ways is not great for our business. If you watch one video, um, you probably
00:13:59
could have seen a few different photos. So the number of ads you're going to see per hour is going to go down. So we call
00:14:04
that monetization efficiency. But the so the reason why we're leaning to video is not because it's, you know, we have some
00:14:11
like thing for video. It's because that's what's driving engagement. That's what's driving pulling people away from
00:14:16
Instagram to other video platforms. And if we want to stay relevant, we need to do that. So the question with photos is
00:14:22
not everyone's going to be able to make videos. Making videos isn't easy. You probably know that better than anyone. How can we make well how can we make
00:14:29
making videos easier but also how can we make photos more engaging so that they can compete in a video first world. Um
00:14:36
but most of the features are like much smaller. They're like just weird broken things that drive me nuts. Like if I add
00:14:41
you to a collection, we can have shared collections now and you know it goes into the DMs every time I save
00:14:47
something. I can't retroactively add more people you know to that collection.
00:14:52
There's lots of little things like that that drive me nuts. Sure. Um, and so I'm constantly trying
00:14:58
to get those things. So that was the big whale. Now it's just that was the big whale. I mean, there's other ones like, you know, people people
00:15:04
ask to go back to chrono feed. We have a chrono feed. You can get there. We're not going to make it the default.
00:15:09
Uh, so maybe that's maybe the most requested, but that's actually a very loud but relatively small percentage of
00:15:16
people. Mhm. Um, but that one is I do get a lot too. Um, but that one I think we've sort of
00:15:22
found the balance there. But people ask for more power. They ask for a lot of power features. Like I want to be able
00:15:28
to not see any photos from my high school friends above their kids. And it's like, okay.
00:15:34
Well, it's interesting you say that because there's this new feature that you've started talking about that I just got access to 25 minutes ago.
00:15:40
Oh, really? I haven't really fully dived in yet, but I would love to hear you explain it because giving users control is a very
00:15:47
risky thing to do. It and many don't even pretend to want to do it. Yeah. Uh, but there's a new feature
00:15:54
where you'll be able there's no name yet. You tune your algorithm essentially. Explain what it is first of all and then
00:16:00
why you figured this was something you wanted to try to do. Yeah. So this was a meme on threads about two years ago almost now called
00:16:06
dear algorithm and people just wrote these letters asking for things that were very sort of
00:16:12
bespoke like that example from before and it was but it's clear that you know
00:16:17
people want more control over the experience and I think that makes sense and a world where more and more of not I
00:16:24
mean look what the the primary people actually share on Instagram is not even through feed or stories it's through DMs
00:16:30
so the connecting side is really healthy on the on and direct messages. But in terms of what they consume, more and
00:16:36
more of it is recommendations, content from accounts that you do not yet follow, particularly reals. And so in
00:16:41
that world where following still matters, but matters less than it used to. How can we give you more control over the experience? And so we took a
00:16:48
look at that that meme and then we took a look at some of the new um
00:16:53
opportunities that some of um some new technologies and AI allow and we created
00:16:59
a way for you to actually see what we think that you're interested in. I saw that I got that far.
00:17:05
Yeah, you got that far. So I got I hit the button and I saw the eight or nine top topics that it thinks
00:17:10
I'm interested in. Yeah, it's not comprehensive. It's the top ones. Okay, that's funny cuz the number one one is like the model name of
00:17:16
a car that I've been looking at for years, which makes perfect sense. Uh, so I shouldn't be shocked that it knows
00:17:21
that about me, but it is funny that it's gotten that just from how essentially how I've engaged with content that's
00:17:27
been recommended to me. Yes. Okay. And then you can add topics cuz maybe you're into something that we don't know
00:17:32
you're into. And then you can um add things that you would not like to see. Um, so you could
00:17:38
just be like, I'm not interested in American football or I'm not interested in politics or whatever it happens to be.
00:17:43
Um, the thing that it doesn't support yet that I'm really excited about is the ability to see something that we think
00:17:48
you're into and for you to override us. So, for instance, you could be like, you
00:17:54
know, I don't know, maybe maybe you were all into um March Madness and then your team lost
00:18:03
in the first round and you're just like, I really don't want to be pummeled with March Madness for the next four weeks.
00:18:08
So, you can pop it out. Um there's lots of examples like that where we might
00:18:14
misinterpret um a signal as that you're interested in and it might be more that it's like
00:18:19
upsetting you or that it was interesting to you a week ago but isn't now or I was hate watching it and I really just wish it would stop showing up so I
00:18:26
could stop hate watching. Hate watching is a thing. Uh this isn't hate watching but I have a friend who um
00:18:31
got stuck in this place where she saw was constantly getting like videos of like animals in like tough situations.
00:18:38
It's like kind of sad and like rescues and like sometimes you don't know if they get rescued or not. It's kind of sad.
00:18:43
Yeah. And she but she couldn't help. Yeah. And so like doom scrolling. Yeah. Yeah. So
00:18:49
um I actually for her um this because this was a couple years ago. I was like look we can you can just do a full reset. I I don't recommend it but
00:18:57
for everybody because it's a big deal but you can just be like forget everything hard reset. That's a button like an algorithmic
00:19:02
forget I said anything type button. Yeah that exists. Okay. Um, but this might be a nicer way
00:19:07
to do that because maybe she doesn't want to have an Instagram that is much less interesting. She just wants to get rid of pets.
00:19:12
Interesting. What What is the riskiest thing about this? Because I I I imagine I mean there's Tik Tok. They've never
00:19:18
even pretended to want I mean I guess at the beginning when I sign up they're like give us your basic interest and that's kind of the end of that.
00:19:24
Yeah. YouTube there's categories I can click on, but my algorithm is just doing its thing and I can feel it shifting when I
00:19:30
watch a little bit more podcast stuff or when I watch a little bit more car review stuff. What is the risky thing in
00:19:36
your head about giving users these controls over their recommended stuff? Two big things. One, we might mess up,
00:19:43
right? So, like we won't deliver. You'll not get as much of the thing you ask for or you'll say you don't want something and you'll still get it at which point
00:19:50
you're going to rightfully point your finger at us and that will happen, right? We're not we do make mistakes um
00:19:56
particularly at scale. And the other is that it'll be um less efficient uh at
00:20:04
you know driving engagement and driving the business, right? Like showing you what you ask for might mean you use Instagram less than showing you what you
00:20:10
tend to interact with more. Um or it might be there's a bunch of ways in which it can get more more or
00:20:17
just spending time on making sure that's working is an opportunity cost and we could have we could spend that time elsewhere making Instagram better in
00:20:23
some other way. Uh so those are very real but I think we're going to a world where this is possible at a level of
00:20:30
execution that it wasn't before particularly because on the back end if you look at how ranking has evolved over
00:20:36
the last five or plus years a lot of the most advanced techniques have become
00:20:41
less and less human interpretable. So we're like we're looking at these things where we take videos and we map them
00:20:47
into this space that we call an embedding and all these things are these giant vectors and we can't even describe
00:20:52
why these seven videos are near each other just happen to be because of behavior. Yeah. But then it turns out with some of
00:20:59
these new models we can be like oh that's you know in your case you know a
00:21:04
specific type of car. Uh and make and so we can so that allows us to do things like not be like hey are you into sports
00:21:11
but you can be like no I'm into you know I'm an Arsenal fan I want to I want to see like the invincibles on re
00:21:17
highlights you know vintage you know you know goals and we can go out and try and
00:21:23
get that for you. Interesting. And also I wanted to ask if you'd ever heard of this extremely
00:21:30
discreet theory about crustaceian evolution. No. But have you heard about how uh how it seems
00:21:37
like there's a bunch of crustaceations are all evolving into lobsters or crabs
00:21:42
specifically because it seems to be the ideal form. It's the ideal form for living under, you know, super high
00:21:48
pressure cold water. No, but I'm into this. So, it's interesting because I like analogies and I like to take that and
00:21:54
look at it in other lenses. Yeah. And sometimes it feels like products do that. Yes. where they're all competing
00:21:59
with each other and trying to get better and they're eventually just becoming the same thing
00:22:05
whether or not it's the ideal thing is there a possibility that social networks are doing this where you are
00:22:11
running Instagram and in order to compete as I've heard you say uh you're adopting you know people are
00:22:17
obviously engaging with videos reals things like that so we're going to do more of that maybe potentially losing the core of the
00:22:23
photos and the things that came beforehand that brought people to Instagram is Is it possible that the competition
00:22:30
is all sort of getting at the same thing in a way? Yeah, I think you can make that argument
00:22:35
for sure. You we're all sort of regressing to the mean. I mean, I think so in some ways like we're trying to do
00:22:40
certain things differently. So, we're trying to do reels differently by making them more social, more friendly. You know, things like that friends feed
00:22:46
inside of the reals tab where you can see what your friends have interacted with, leaning more in re-shares, etc. Yeah. But it but it's also true that if
00:22:52
you take a big step back and you look at all of social media, not only have the apps become more and more similar, the
00:22:58
rate of convergence has increased. So you look at you know uh like when did
00:23:04
each app introduce ranking and there was like many years between like the first app and the last app.
00:23:10
When did each app introduce stories? There was a few years. I mean it was popularized by Snapchat. It was invented
00:23:16
by cacao stories in Korea. Um but then like every you know Pinterest even
00:23:21
YouTube x everywhere. Yeah. Uh and that was faster. When did we all
00:23:27
start leaning into mobile first video? You know Tik Tok style, real style,
00:23:33
YouTube short style that happened faster um recommendations faster. So I think
00:23:39
that the competition is fierce and as a result there is a very real pressure to
00:23:45
borrow and steal ideas from the competition that works. But you also
00:23:51
have to figure out how do you also stay true to your own identity and how do you differentiate because if we're all the same
00:23:56
then you can just switch apps and it doesn't cost you anything. And I guess is there a pressure to well there is a
00:24:03
pressure to evolve the feature into your own thing but do you feel like leaving
00:24:08
behind some of the core things like chronological feed or just a ton of
00:24:14
photos everywhere. Do you feel like there's a risk to that or is it just sort of smaller numbers to the point
00:24:19
where it doesn't have to matter as much? So there's definitely a risk. My hope is that
00:24:25
the core ideas are the same if the how is different. So, you're still connecting with friends, you're still
00:24:31
exploring your interests, you're still expressing your creativity if you want to, but how is changing. Uh, but yeah,
00:24:38
there's definitely risk and there will always be people who are going to be upset and who going to want us to move
00:24:43
back. And look, I feel that way too sometimes. But I think the alternative
00:24:50
is to just to become irrelevant because I don't think you can just not evolve
00:24:55
and hope to continue to grow or even maintain your scale. I think you're going to see people we we do we see
00:25:00
people when we don't evolve fast enough leave us for the competition. I'll give you an example. We look all the time.
00:25:08
Pick a team, the ranking team, the video creation team, whatever. We look at two things. We look at the
00:25:13
top line. How is the trend doing? Are people, you know, sharing more on Instagram? And then we look at a back
00:25:19
test. So all the changes we make in a in the six-month window, we will hold out or not launch to 1% of people. And so
00:25:27
you can see what is the value that the team is creating and then you can also see what the overlying app trend is. So
00:25:34
I'll make up numbers. Let's say we increased how much people are sharing photos and videos by 5%. But the overall
00:25:40
amount of photos according to the back test, but the overall photos and videos shared on Instagram per person per day
00:25:45
was totally flat. That means there's a 5% headwind. That means if you don't do anything, you're losing 5% every six
00:25:52
months. um we see that across um a lot of different key metrics which is just a
00:25:58
signal of either people moving on or more likely just the competition getting stronger. Uh and so if we don't evolve,
00:26:05
you're just going to see us shrink and then that spirals. If there's less people using Instagram, there's less reach to go around. Creators start
00:26:11
leaving, less people then use Instagram, there's less reach to go around still and you're in a negative feedback cycle.
00:26:17
Interesting. The bottom line for the company matters, of course. Obviously, you want you want to keep the core values of the of the product, but also
00:26:22
you need to still be a business for sure. How do creators make money on Instagram
00:26:29
today? Primarily through um branded content
00:26:34
deals and also increasingly through what we call partner ads. But there's at a high level three ways to make money
00:26:40
uh on all platforms. You can either get paid by a brand or a company, you can
00:26:46
get paid by the platform directly. YouTube is best at that. Or you can get
00:26:52
paid by your fans, users. And we have things like subscriptions and tipping, but they're smaller scale.
00:26:58
One thing that's interesting in terms of getting paid by companies, historically, it's been primarily brands doing branded
00:27:04
advertising, and that's great, but there's a lot of money in advertising that isn't brand
00:27:10
dollars. It's what we call direct response. So trying to sell a specific chair or watch or whatever.
00:27:15
Yep. What partner ads do is allow creators to create content and then opt in to letting brands use that creative
00:27:22
as ads, which then benefits from the entire measurement system of our ads
00:27:27
system, which is very good for direct response advertisers, which then increases the pool of dollars for
00:27:34
potential dollars for creators to make by allowing them not only to do off-platform brand deals with big brand awareness
00:27:40
campaigns, but also deals with direct response, which is a which is a massive
00:27:46
actually the majority of our business. Yeah. I'm interested in the breakdown of that cuz I think when I first started as a creator actually when I first started
00:27:53
none of them were really a thing. Yeah. But uh especially on YouTube for example
00:27:58
uh made lots of videos made no money and then the first version of this was revenue sharing with the platform.
00:28:04
Yeah. And the first checks were extremely small but it was just the easiest bottom
00:28:09
way to get like a foot in the door to start to see some revenue from the work that you do. Yeah. And then it felt like you
00:28:16
graduated to the point where you were able to work with a brand on something. Uh what is the breakdown like on
00:28:22
Instagram? Is there is it mostly uh people working with outside companies and then putting their stuff on
00:28:28
Instagram? Yeah, so Facebook has a meaningful revshare business. Instagram doesn't we've been working on it for years now
00:28:35
and we haven't been able to make it sustainable and I can go into the details there. Sure. On what I think sustainable looks like.
00:28:40
It just has to break even. We can't just be burning money like you said. we are a business. Um, but the vast majority le, you know,
00:28:48
it's easily north of $10 billion a year, probably a lot more. This is a couple years old at that data point. Um, is
00:28:55
through offplatform deals that creators make directly between them and brands and then they use in they use their reach on Instagram um to monetize that
00:29:04
way. And now what we've seen is we've been able to scale up partner ads, these direct response ads to a really
00:29:11
meaningful scale. Um, which we hope is just increasing the overall pie. We still have subscriptions. We still have tipping. We're still working on what we
00:29:18
call um bonuses or performance-based payouts, the sort of rev share. Um, but
00:29:24
they are much smaller in scale. Yeah. So, how do you how do you grow that revshare piece? cuz I think a lot
00:29:29
of creators think about that and would love it to be more but as as you say it's not as sustainable. What are the
00:29:35
what are the major things keeping it from being sustainable? So there's three things I think that the pl that the program needs to be to make
00:29:41
sense um to scale up. One is the it just needs to not burn money. Um
00:29:48
so what the theory here is that you pay creators they create more content then you have more advertising and then you
00:29:55
can make back what you pay creators. um I don't need to make money off this if it's even close to break even or break
00:30:01
even. That seems like a great thing to do. Yeah. Two, the criteria for being eligible
00:30:08
needs to be transparent. It can't be some algorithm and you're in and I'm not and we have no idea why. It has to be like, all right, you need
00:30:13
to hit might be might not be like automatic. It might be like you need to average a million views a week or whatever it is, but there
00:30:19
needs to be some transparent eligibility. There should be a page that explains it. Exactly. But we need to get the program to be break even without using
00:30:28
algorithmic ways of identifying creators so that you can just actually explain why you are are not eligible. And then
00:30:35
three, the checks just can't be embarrassing. Like if I sent you a $4 check, you'd be like, "Thanks, but now I'm kind of offended." Um, that doesn't
00:30:42
have to be that high. It just has to be something that's like, "I don't want to be sending a bunch of creators pennies." Yeah.
00:30:47
Um, I just think that'll that'll just be a bad look. It's interesting you say that. I I heard you say that on Colin and Samir and I
00:30:54
actually disagree. I think you'd rather get pennies. I would rather get pennies and I it's not really me specifically, but I think
00:31:00
as budding creators because again it's it's hard to go off platform and get good at brand deal and manage the inbox
00:31:06
and negotiate and do all that. Yeah. And so to just have like a base way of of having like splitting revenue
00:31:12
with the platform, even if my first YouTube check was $1.74, like that was
00:31:17
motivating in a way for me to see, okay, if I can scale my reach, if I can focus
00:31:23
my efforts on certain things, then I know that I can increase what I'm making. Yeah. And obviously, it's not going to be a
00:31:29
full-time massive chunk of the business for every creator. But I think I'd rather get $4 than $0 if I know that
00:31:36
there's a reason I can move it to $10. I might be overreacting to angry
00:31:43
DMs I get from people who are in the program who who are like, "You make X
00:31:48
billion dollars a year. You sent me a check for $9." Yeah. It's maybe because it's hard to
00:31:54
explain exactly how you arrive at the number cuz I know on YouTube Yeah. you can very easily determine what ad came
00:32:00
from what video and how it's associated with the uploader. Maybe that's harder on Instagram, so it's harder to explain
00:32:06
to attribution is tougher. Yeah. Yeah. But I look if I'm over I would love to drop one of the question to make it more
00:32:12
possible. Um that's not the gate right now. Um the first two are um but um you know we
00:32:20
might get there. Maybe we'll get there in some countries and not other countries. But um I could very well be
00:32:27
over rotated on that piece of feedback. I'm always trying, like we said, how do you step through the signal and the noise?
00:32:32
There's a lot of strong opinions about what we do, particularly from people who rely on us. And I get that. And so I'm
00:32:38
always trying to figure out um when am I just being
00:32:43
when am I really focused on what's best overall and when am I might be overreacting to not being not maybe
00:32:48
liking getting yelled at? And this might be an example of the latter. Yeah. I just I when again the arc on
00:32:54
other social networks especially with YouTube that I came up on so that's why I reference that a lot is like when I
00:32:59
was first starting sure the $1 check didn't meaningfully make a difference to me but it was motivating and then
00:33:04
eventually it was the biggest part of the business and then when I was able to get the skills to go out and sell to
00:33:10
others who wanted to like work with me on ads and manage the inbox and all that. Yeah, that became the next biggest chunk and
00:33:18
then the revenue share with the platform became secondary. Yeah. And it's also it's always going to
00:33:24
fluctuate. Like I don't think any smart business depends entirely on the revenue share of the platform. Yeah.
00:33:29
But it does represent motivation and a nice like bar to continue to see if it
00:33:34
goes up or down. You can kind of associate that with the performance of the entire channel. So I feel like it is
00:33:39
uh something that most I think you're getting a vocal minority being very angry, but I think most people would be
00:33:45
happier to see something than nothing. Yeah. I mean, it it it's a compelling
00:33:51
argument. So, you're giving me something legitimate to think about. Uh, which I I think it would not be
00:33:58
surprising to me if I was overreacting to a vocal minority. I'm always trying not to.
00:34:03
Um, but it's hard. Uh, and so just pattern matching in general, the kinds of mistakes we make, that's a common
00:34:09
pattern. So, you're probably right. Fair enough. [Music]
00:34:21
Support for this episode comes from Masa Chips. You know, sometimes the best snacks are the ones that have felt like they've been around forever. Simple,
00:34:28
straightforward, they're done right. That's exactly what Masa Chips are aiming to do with their delicious tortilla chips made the traditional way.
00:34:35
Masa makes their chips with just three ingredients. Organic Nyximalized scorn, Redmond sea salt, and 100% grass-fed
00:34:41
beef tallow. And the result is a chip that's sturdier, crunchier, and more flavorful than the big brand bags you're
00:34:47
used to. So, one superpower these chips have is since they're so sturdy and crunchy, the
00:34:52
amount of guac you can pick up in one singular scoop is kind of like out of
00:34:58
this world. That's like four times the amount of guac versus chip ratio I have in my hand right now. And they come in a
00:35:03
bunch of great flavors. You know, there's the obvious white, blue, traditional. They also have an habanero
00:35:09
flavor that's fantastic. And what surprised me the most was this churro flavor, kind of like a dessert tortilla
00:35:17
chip. Can't get enough of it. But Masa's mission goes even further. They only source their beef tallad domestically
00:35:22
and are focused on creating their products with sustainable practices. So if you're ready to try the most delicious tortilla chip you've ever had,
00:35:28
which are sustainably produced, go to masachips.com/wwayform and use code waveform for 25% off your
00:35:35
first order. Link in the description. How often do Instagram ads get you to buy stuff?
00:35:40
Um, more often now than they used to. They do. They get you. They get me. I always thought like I I never get got
00:35:47
by my own ads. Obviously. They still get you. They get me sometimes. Well, cuz sometimes it's something I didn't know
00:35:53
about. It's like very tough. I mean, it's perfect. It's like I'm a dad. I'm in my 40s. I live in San Francisco. I
00:36:00
like gadgets. So, they're like, you know, they're like growing like they go after me with the things that it they're
00:36:05
pretty good. I mean, what do I like? Uh, what have they got me with? Uh, coffee gear. I really like coffee gear. That's a popular one.
00:36:11
I really like coffee. I make I'm got deep into pourover for a couple years. I got deep into espresso for a couple years. So, you know, a little tiny
00:36:20
coffee grind vacuum that's perfectly designed to like fit into your setup. They say just the right thing with just
00:36:26
the right pictures. Yeah, I know. I was like, "Ah." Um, my wife always jokes. Um, she I we we
00:36:33
once a year usually have an argument about whether or not Instagram is listening to your microphone. Um, she's
00:36:39
sometimes doubts my position on that. Yeah. Um, it seems like the ads get really good sometimes and have to defend that.
00:36:45
They do. They do. Just for the record, they are not. We're not Oh, for a couple reasons. Okay, hold on. Yeah, I should
00:36:50
be careful. Yeah. Um, so, uh, you might have been
00:36:56
looking for it online, uh, because we do work with advertisers who send us data about what people were looking on their
00:37:02
website, uh, who they want to target those people. So, you might be looking for, like we were talking about, a camera lens and then get that lens.
00:37:09
Yep. Um, you might have been talking about it with a friend and they might have been looking at it online.
00:37:14
That's the one that seems to be the creepiest but is actually real. Yeah. where that person's getting ads for it, that person bought it, I talked
00:37:20
to them about it. And is it proximity or is it just like my relationship with it's more it's it'll most likely happen
00:37:27
if you two are similar on paper. So it's like two guys, you know, in tech roughly the
00:37:32
same age. It's like that kind of thing. Yeah. Um, and then sometimes it's just people
00:37:39
don't realize they actually saw the ad and cuz you scroll by them and sometimes that registers subconsciously and then
00:37:45
you're talking about that area. That's kind of how advertising works. Um, those are the three main. But here's
00:37:51
the other thing. No one believes me, but the other thing is if we were if we were listening to your mic, your battery would drain real fast.
00:37:57
Sure. The if the phone was remotely competent, you'd be seeing the little light saying
00:38:02
that the thing was on. Yeah. Um and it would just be a gross violation of privacy most importantly.
00:38:08
Yeah. So like we don't um but I do have this conversation once in a while. Funny.
00:38:14
Uh okay. So what is the future of Instagram look like? Cuz uh there is this evolution happening in social networks and obviously you've seen ev
00:38:21
the Instagram of old to what it is today. What is Instagram in and I love
00:38:26
asking you this because CEOs hate talking about future products but I'm going to ask you anyway. What is Instagram in 5 years, 10 years? Is it
00:38:34
very different? Yeah. Is it still keeping the core of what it is now? So, I think there's a period of time
00:38:39
that's more clear and there's a period of time where it gets a lot less clear. That's always true. The further on the future, the less
00:38:44
granularity you have, but there's a little bit of like a um like a a threshold. So, over the next couple
00:38:51
years, I think the biggest thing that's going to really change our business is going to be AI. I think you're going to see a lot of
00:38:57
great things where you're going to see more creators can create more content. Uh so you can create more content or
00:39:02
better content if you're already a creator or people who didn't have the ability to make content can make compelling content who couldn't used to
00:39:09
before. Um but there's also lots of concerns, right? There's also, you know, deep fakes. How do you know what to trust?
00:39:15
How do you know that something was actually captured with a camera? You won't be able to. So there's opportunities and challenges that come
00:39:21
along with more and more content that is created with AI. We too often talk about
00:39:26
AI as binary. It's either synthetic or real content. And the truth is AI assisted in the middle.
00:39:32
It's going to be mostly that is my guess. Um uh so so I think that's going
00:39:38
to change what content you see on Instagram, who you follow on Instagram, who you even talk to on Instagram. These
00:39:44
things are going to change and there's going to be a lot in the blurry middle. We focus mostly on the extremes. Go back
00:39:49
to the way of old or don't talk to like the unsupervised AI bot. But I think most of it's going to be in between
00:39:56
those two. Where things get much more fuzzy is when the form factor changes. Um we obviously
00:40:02
have been very public about the fact that we believe that eventually phones will still be part of our lives but not
00:40:09
nearly as dominant. Um, one example of how it could turn out is how, you know, desktop and laptops have turned out,
00:40:15
which is, you know, you have more than one probably. Um, uh, you know, it's
00:40:21
still an important part of your life. My computer's important part of my life, but it's not the most important in a world where, you know, wearables, but
00:40:28
particularly glasses, I think, get better and better more in, you know, the five to 10 years. And obviously we we've
00:40:34
talked about this publicly with um Met Ben Displays and with um Orion.
00:40:40
Um I think that that makes it much less clear what Instagram is. Yeah. Cuz there's certain things that
00:40:46
naturally flow like most of what you do in terms of connecting with Franc's message. You can imagine that.
00:40:52
Yeah. Um you can capture photos, but then what does a consumption experience look like?
00:40:57
Yeah. in in a glasses first world is I mean I've got a couple
00:41:04
educated guesses but I think it's going to evolve a lot. Does any of it worry you? Cuz I made a reference at the end I got to see the
00:41:10
metaray band displays and I made a reference at the end of my video to like you know there is an Instagram app on
00:41:15
those glasses and there is a actual real possibility that I could be talking to
00:41:20
someone wearing the glasses and they could just be scrolling through Instagram. Yeah. And not paying attention to me. Looking through glass at me.
00:41:26
Yeah. That's pretty close to that Wall-E meme. Like that's extremely with the slurpies,
00:41:31
the people. Yeah. So like is that I mean it's it's possible now that could actually happen. What do you think about that?
00:41:36
I'm a worrier by nature. So I worry about all of the things. Um there's like how can it get misused? How it could be
00:41:43
bad for people. There's also how it could be bad for the business, how it could be bad for us. I mean the thing that we've maybe talked about a little
00:41:49
bit less publicly uh in the context of AI and new technology is like yes it's changing what our apps are and what
00:41:56
people do on their app on our apps. It's also massively changing how we build our apps. We we're trying to wrestle with it
00:42:02
even internally as well. So I think it's important for us to be honest about the
00:42:08
risks and the downsides and then try to do what we can to maximize the upsides
00:42:13
and minimize those downsides. So yeah, I worry about it. Um, I've got kids. Um,
00:42:18
what what is their life going to be? I mean, so when I was 13, I'm 42. I was born 83. When I was 13, it's mid90s.
00:42:25
This is like AOL dial up 2400 BPS early internet. Yeah. Like
00:42:30
on the phone. Yeah. So, look at my parents had no idea what I was doing online. Um, and I think
00:42:38
one of the risks for me as a dad, my kids are too young to use social media, but they're nine, seven, and five.
00:42:43
they're going to come up as a lot of this tech comes up and if it's going to be hard for me to
00:42:49
even understand possibly um what the risks are if I'm not proactively leaning
00:42:56
in. So yeah, I I I think about I'm excited about all the positive opportunities and I am concerned about
00:43:03
all of the risks and I I try to think about both. It's an interesting position because as a social network with now it's three
00:43:10
billion monthly active users. There's three billion. It's like a hard number to wrap my head around.
00:43:15
You become kind of like fabric of society type big like decisions that you make to get people to use Instagram more
00:43:23
get humanity to be on their phones more. Like that's the type of repercussions that come with what's happening at
00:43:30
Instagram, you know, which is interesting. So I I actually wonder what you think of you've seen these AI creators like
00:43:36
creators on Instagram that are entirely AI. Well, yeah. Well, there's two kinds. Well, right. And I'm I'm talking about
00:43:42
the ones that are are completely independently AI, but not just independent AI. It's a persona. It's a person who right there's
00:43:48
also like if AI who's like fully synthetic content, but they're creating these dreams, random all types of stuff. Yeah. I mean,
00:43:54
there's like a creator with a name and a face and it's AI and they make tons of stuff for no effort.
00:44:00
I figure that's good for Instagram in theory because it's very efficient. It's a creator that doesn't demand a ton of
00:44:06
revenue, but like creates tons of engaging content. Yeah. But on the other side, as a as a human creator, yeah,
00:44:12
it's horrible. I hate seeing that. Yeah. What do you think when you see AI creator doing what human creators do?
00:44:20
Um, both. So, I mean, so on one hand,
00:44:26
I think that in a world where anybody can like make any type of video with
00:44:34
like, I don't know, $150 worth of AI credits, whether it's personifying a person that doesn't exist or anything
00:44:39
else. Um, like what do we have left as creatives? And I think it's really our
00:44:45
taste and our understanding of culture and our own creativity. And so I I'm in
00:44:53
some ways um bullish on that those ideas that
00:44:59
creatives will always have taste, have creativity and have a position on culture. And therefore whether or not
00:45:04
the content was created with AI, without AI, or entirely with AI, they will be
00:45:10
telling stories, pushing boundaries, driving culture, and that'll be good.
00:45:16
Um, but I also worry about, and it's also great if there's people who, you
00:45:21
know, maybe you just were never comfortable in front of a camera and now you can do it. But then I also worry about deep fakes. I worry about just an
00:45:29
explosion of competition. If there's 10 100 times as much content,
00:45:34
there's no way there's going to be 10 or 100 times more engagement. So, the average reach is going to go down. We'll probably get blamed for that. Um,
00:45:42
what does that mean for people who, yes, it's great who people who couldn't create content before can now create
00:45:48
content, but for people who aren't don't have access to or aren't technical enough to leverage this technology, how
00:45:53
to make sure that they still have a voice. Yeah. So, uh, my mind pingpongs between the
00:45:59
exciting opportunities and the scary risks. Yeah. Does it make sense to create rules around it? like you want to enable
00:46:05
obviously you want to raise the floor and make it more accessible for people to create but I don't I don't think this
00:46:12
is me being biased as a human creator I don't think we want Instagram to be full of AI only creators and there are some
00:46:18
that are getting tons of engagement very successful profiles but I I don't love the world where Instagram is full of that
00:46:24
is it up to the rules of Instagram to prevent that from happening or if they just make good enough stuff is that's just the way it's going to be
00:46:30
it could be I mean I have a really hard time imagining it getting to a place where like that's
00:46:38
all that there is because I just don't see how that content is going to cover enough culturally relevant content.
00:46:45
You're very you believe in humanity a lot. I think people are interesting and like and look AI will be interesting
00:46:51
um over time but right now like there's a reason I mean on one hand like it's unbelievable what
00:46:58
some of this tech can do. Um, like you know, for instance, I was living in the UK for a while. One common thing for
00:47:03
banks to do is voice signatures to like sign in. Like you should definitely not do that anymore. Way too easy to fake.
00:47:09
As someone who has thousands of hours of my voice in high definition on the internet, don't do that. Don't do that. Yeah.
00:47:14
Um, on the other hand, like I haven't found a funny AI yet. I haven't found an AI that has um
00:47:23
said something that I thought was like on the bleeding edge of what's interesting in culture right now. Um and
00:47:30
that will happen. Um but I think that's further away. I think right now people particularly creatives
00:47:37
um bring something to the table that AI does not. Even though AI can completely
00:47:43
with particularly with direction fabricate something that didn't exist. Yeah, I think to an extent some of it is
00:47:50
interesting now because it's AI, which may be why some of it is blowing up, but I think the other reason so many people
00:47:56
call it slop is because it kind of just appeals to the lowest common denominator. We're we're sitting here thinking, okay, we've never seen
00:48:02
anything really interesting or creative or funny. I've never really thought that hard about something AI,
00:48:08
but at the very lowest level of just like basic humor, basic interesting
00:48:13
imagery, they're making that stuff now. Yeah. So that's but the thing about that from an
00:48:18
Instagram business perspective is that it's not it's it's not that valuable. So
00:48:27
like for it's too commoditized. You'll be able to find that anywhere.
00:48:32
So for us, sure it might be nice for some of the content on Instagram to be content that you could find elsewhere
00:48:37
and drive some engagement, but it's fundamentally not differentiated. It's not defensible. It leaves us exposed. If
00:48:45
we become all content that you can find anywhere else, then it's very easy for you to leave us.
00:48:50
Um whereas if there's content that is more unique or differentiated, then I
00:48:56
think that it's much more valuable beyond the amount of engagement it might drive or how amount of revenue it might
00:49:02
drive. That's one of the reasons why we focus on creators. The creators do post across multiple platforms more and more
00:49:08
so, but it is fundamentally more differentiated than like sports highlights or AI slot for instance.
00:49:14
Yeah. So, I mean, yeah, sounds like there's at least some incentive for Instagram to keep humans on the site
00:49:20
happy and lots of great stuff. I mean, look, there's different there are different forces in different directions at play. Um, it's just like
00:49:28
most big questions, not black and white. Um, not simple. Yeah. Well, I guess then I have one more
00:49:36
question for you. Yeah. Which is the same question from the beginning. M you know Instagram 3 billion monthly
00:49:41
active users reels upcoming AI you know maybe leaving some features behind
00:49:47
evolving forward what what is Instagram right now 2025 Ooh right now um it depends a lot on the
00:49:57
country but I'm not going to use that as a copout but it is wildly different by country I believe that
00:50:02
um in so maybe you know we're here in the US so maybe in in the US I think
00:50:08
that we are there's some strengths and there's some weaknesses. I think on the strength side, you know, I think we are
00:50:17
still the de facto home for most creators. They use multiple platforms. YouTube is better at long form video.
00:50:23
They're better at revshare. Tik Tok is better at breaking new talent and helping people get discovered.
00:50:29
But we are a more complete suite of tools for you to not just tell your stories, but engage with your audience,
00:50:35
etc. So that's why if you look at how many creators are on Instagram, it seems like they're more than any other platform. Um, we are trying to continue
00:50:45
to stay culturally relevant by investing in originality and timeliness. Um, but
00:50:50
there's real risk that the competition will execute better than us. Um, and
00:50:57
I don't want to be in charge of Instagram when we drop from being like a
00:51:02
tier one app in terms of cultural relevance to a tier two. And that to me that feels kind of
00:51:07
existential. Um, I want us to be the best platform for creatives, the best
00:51:12
form for people to talk to their friends about their interests. In some ways I can point to numbers that suggest that
00:51:18
we're the first and other ways I can point to numbers suggest that we're the second or the third. Um, and so we're, I
00:51:24
think, at this really intense, possibly pivotal moment where the world
00:51:30
is changing faster than ever. The competition's fiercer than ever, and we're trying to navigate that and
00:51:37
somehow also stay simple and stay true to our identity. Yeah. Well, sounds like you're a busy
00:51:44
man, so I'll let you go. Adam, thanks for the time. Such a pleasure. Appreciate it. Hopefully us again. Yeah, it'll be a blast.
