Skip to content

By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.

Silicon Valley GirlSilicon Valley Girl

Head of Instagram Reveals How the Top 1% Actually Work With AI

I asked Adam Mosseri to show me the AI setup he uses at work and in his personal life.

Adam MosseriguestMarina Mogilkohost
Nov 17, 202644mWatch on YouTube

Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.

Great conversations run long. Their best ideas don't have to.

uListen is free on iPhone, iPad, Android, Chrome, and Claude.

See how

Transcript

  1. 0:00 – 1:13

    Inside Adam Mosseri’s AI Setup

    1. AM

      These models are smart, but they're kind of the average of the internet out of the box. But if you can build loops to help train them, they get much more interesting.

    2. MM

      This is Adam Mosseri, head of Instagram, the app three billion people use every month. He built his own AI command center to help him run it. Every night, it goes through his meetings, emails, and messages. But the first version-

    3. AM

      It was garbage. It was, like, more stuff to read, and it helped me zero. So that loop turned these briefings from a time waster into something that's very helpful. If you want to make AI work for you, you have to embrace a learner's mindset. I'm learning every day now in a way that I haven't had to in years.

    4. MM

      I'm just thinking, like, if I'm not a technical person, to build out something like that-

    5. AM

      I think you kind of need to-

    6. MM

      This video is sponsored by HubSpot. You-- I think you're the top, not even one percent, top zero point zero one percent, uh, of people using AI to manage a huge organization. I heard you say you talk to your four AI agents almost every ten minutes.

    7. AM

      The setup itself evolves a lot. I think one of the key things right now, if you want to make technology, and in this

  2. 1:13 – 3:26

    Roger: Keeping Track of Promises and Follow-Ups

    1. AM

      case AI, work for you, is you have to keep a very-- you have to embrace a learner's mindset because w-what the models and the tools can do today is gonna be very different than what they can do in six months. And so the things that I use to try and do a better job change, like, month over month because the technology changes. And so it's not, like, a thing where you can be like, "All right, cool. I'm gonna learn how to do this new thing, and then I'm gonna do this new thing for m- for forevermore." No, you have to kind of embrace a, a approach of constantly evolving, which is, you know, a little-- can be exhausting at times but also very exciting.

    2. MM

      So do you still have your four agents?

    3. AM

      Yeah. I, I mean, so I have stuff at home, and I have stuff at work. But at work, I've kind of put everything into one big sort of experience, um, but there's different pieces. Last time we talked, I told you about a couple of the pieces. I think I told you about Roger-

    4. MM

      Yeah

    5. AM

      ... uh, and Caddy.

    6. MM

      We talked about multiple apps-

    7. AM

      Yeah

    8. MM

      ... and I wanna dig into them.

    9. AM

      Yeah, yeah.

    10. MM

      Um, so w- from what I remember, you have this app that blew my mind. I'm like, "I need to have this," is the app that has access to your Slack, has access to your email, and tracks things that people promised to you and did not deliver.

    11. AM

      Yeah, Roger. I call that Roger. [laughs]

    12. MM

      Okay, that's Roger. Okay, talk to me about Roger.

    13. AM

      When you run a large team, one thing that happens a lot is you ask for something, and people say, "Yeah," and then, like, they never do it, or they don't do it for a very long time. And so much to my team's, um, chagrin, I have built a thing that sort of keeps track of things I ask for, uh, either, you know, in meetings or in email or in Slack or anywhere else. And then I can go back and take a look and say, like, "Hey, you know, I asked, I don't know, Max for this thing two weeks ago, and he never heard back. Do you wanna nudge him about it?" Um, and so I am often, uh, [laughs] sending out a bunch of follow-up notes being like, "Hey, you know, friendly reminder, like you said this thing would come back in a week, and it's been, you know, a week and a half. Like, what's going on?" Uh, so Roger is, uh, one way I try to stay on top of things I ask for and keep accountability high. I mean, the way it works mostly is, like, now it all just kind of gets fed into one big sort of to-do list for me. The way I will always picture things, forget about AI for a second, is like I feel like I have a lot of work to do, and there's the thing I'm trying to do right now, and there's a list of other things I need to do, and that list is so long it kind of goes out of my view.

  3. 3:26 – 5:33

    How Adam Built His AI Command Center

    1. AM

      Uh, and so I'm like, "All right, cool. I can kind of see, like, two or three in my peripheral vision of what's next, and then-- but there's the one thing I'm focused on." And I actually built something that kind of looks like that, where it's here's the thing that I think is the most important thing to do, and then here are all the other things, and then y- I can just, you know, blast through them. Um, and I can snooze them, or I can edit them, or I can, you know, whatever. Or, you know-- But it's a lot of read this and send back feedback on this roadmap or respond to this, you know, question or unblock this team or nudge this person about this thing you asked for two weeks ago. Um, but I-- The thing that-- The desk-- I will call it like a desk. The desk I built is basically how I've been visualizing my work forever.

    2. MM

      Hmm.

    3. AM

      But now I can actually make it a thing that... It's not perfect, um, but it, it certainly helps me be more productive.

    4. MM

      So desk is like your command center, right?

    5. AM

      Yeah.

    6. MM

      Where you see all the tasks.

    7. AM

      Yeah, everything's in there. So Caddy's in there. Roger's in there. All my briefings from all my meetings are in there. All the history of all my meetings are in there. Like, everything is in that one sort of command center.

    8. MM

      How'd you build it?

    9. AM

      We have an internal harness we use for building things, and it-- you know, I can use multiple different models. It has basically w- what we call the Meta CLI. So it has a set of internal APIs for things that are inside the company only. Um, but I just built a web app with a simple SQLite database. It's not a complicated thing. Um, but then there are these automations that run at night that try and find all the information that it can about how, how meetings went, what the action items were, what I asked for, what people asked me for, what people messaged me about, what people emailed me about, and then constantly update and populate it.

    10. MM

      I want to understand this command center. I'm th- just thinking, like, if I'm not a technical person-

    11. AM

      Yeah

    12. MM

      ... to build out something like that. So as far as I understand it, I use whatever AI I'm using. I'm connecting my Slack, my email, everything, and ask it to build a command center. Is there anything else in terms of prompting that I need to think about?

    13. AM

      I mean, I think we're, we're not at a place yet where I think if you have no experience building software, you can ask for something like that and get something good. I think you will eventually get there, but right

  4. 5:33 – 7:18

    How He Gets AI to Write Emails That Sound Like Him

    1. AM

      now I think you kind of need to give feedback, like, "Hey, this is how I think it should look. Here's what my goals are. Here's..." It will take a pass. You ask-- You give it feedback. It iterates, et cetera. But I think for most people who have a lot less, hopefully, to-dos than I do, uh, I think something more simple would work great. I think working with any of these proactive agents, obviously ours that we're really excited about is called Muse, to just, uh, engage with you just via text might just be a simpler u- i-interface, and just be like, "Hey, I wanna stay on top of everything. Make sure I don't Go more than a day or two without replying to any emails you think are important. And if you give it access to your email, your agent will then, or your sort of little assistant, will then just tell you, "Hey, like this person is waiting on you. It's been two days. You might wanna follow up." You can build much lighter weight things than my sort of overly complicated-

    2. MM

      What about your emails? I think the last time we talked, 70% of your emails were drafted by AI.

    3. AM

      They get... So they get drafted, but then I edit them. So here's the thing that I think a lot of people, even whether or not you're really technical or you're completely untechnical or a Luddite, um, I think it's important to understand is n- these models are smart, but they're kind of the average of the internet out of the box, and they might not do what you want how you want right away. But if you can build loops to help train them, they get much more interesting. And so-

    4. MM

      How do you build a loop?

    5. AM

      So like an ex- an, uh, a simple one would be like this email example you brought up. So I have one that will just draft responses to certain emails for me, but I don't have it send those. I take a look at those responses. They're in my drafts folder. Then I edit them, and then I send them. But the loop comes into place. I've also told, in this case Muse, to look at the draft that, to save the draft it wrote,

  5. 7:18 – 9:52

    Turning AI Meeting Briefings Into Something Useful

    1. AM

      look at the actual email I sent-

    2. MM

      Mm-hmm

    3. AM

      ... and compare them, and to learn from the delta. And so at first, it was writing very generic emails that didn't really sound like me, and I was rewriting them, and then it slowly learned my tone, my values, things like my schedule. Uh, and it got a lot smarter over time because it automatically looks at the end of the day every day at all the emails it drafted, all the ones I actually sent, and tries to learn from the differences. That's when it gets interesting because if you just ask for it out of the box, you're gonna get the sort of most vanilla version. And you, you want, in this case, to save time, and it's not really saving me time if you're writing email like somebody else that I would never send. But you don't wanna have to tell it feedback manually every time. It's nice to create a loop and then have it learn on its own, so you sort of set it and forget it.

    4. MM

      What other loops have you created?

    5. AM

      There's a lot. Um, for at work, for meetings. So I have a tool that writes me a briefing. So in a job like mine, you're in a lot of meetings.

    6. MM

      Mm-hmm.

    7. AM

      Uh, you get a lot of these really long documents. These are slideshows or just, you know, Google Docs or whatever, and, you know, they can be 10 to 100 pages long. And so I have an agent that basically reads it for me and then tries to summarize it and tell me what matters most, like, "Don't miss these three or four things." I still read the full document, but it helps me to anchor-

    8. MM

      Mm-hmm

    9. AM

      ... before I read on, like, knowing what I care about. What are the three or four most important things? I call that a briefing. When I started it, I just-- The way I created this is I just wrote a really long prompt to tell, I think, in... I forget what model I was using back then, this was almost a year ago, how to create a briefing for me, and it was garbage. It was like [laughs] it was like more stuff to read, and it helped me zero. But the loop I created was go find the materials after the meeting, so the PM's n- email after about what the action items are, the AI notes from the meeting, if you have those-

    10. MM

      Mm-hmm

    11. AM

      ... my notes from the meeting, if I took notes in the meeting, any Slacks about the meeting, if those happened. And then try to grade yourself on how w- good the briefing was. So if you were like, "These are the three most important things," and we talked about none of those, it's like, "Oh, Adam actually-- This is what Adam actually cares about," or, "This is what the strategy actually is," or, "This is what the big, um-

    12. MM

      Mm-hmm

    13. AM

      ... contentious decision might be." And so that loop m- turned these briefings from honestly a time waster into something that's very helpful for me over the course of a month or two.

    14. MM

      Mm-hmm.

    15. AM

      Uh, and it just gets sort of

  6. 9:52 – 12:17

    The Podcast App He Built for His Kids

    1. AM

      smarter and smarter every week.

    2. MM

      That sounds amazing. The question is, do you have to build apps for all these particular use cases, or can you just do it within the chat app? Like, why would you build an app for all the things that you're doing?

    3. AM

      You don't need to. For me, it's helpful because it's just so much information.

    4. MM

      Mm-hmm.

    5. AM

      Right? So I will, I'll pull up my desk, and I can see all my meetings today, all my meetings for the rest of the week, where I've got the materials have already been sent, where they haven't. It-- For me, that's helpful 'cause I can just be like, "Okay, tomorrow I know I have is a heavy day, but I've got some time in the morning to my, to sort my thoughts." You know, and Thursday's light or whatever it might be. So that, that helps me to be able to visualize that.

    6. MM

      Mm-hmm.

    7. AM

      But for just the brass tacks, I'm going into a meeting or I'm, I'm meeting somebody, I need context, you know, that could just be a message.

    8. MM

      Yeah.

    9. AM

      That could just be a text.

    10. MM

      [swish] Quick question. How many things did you promise someone this week? A deadline? A file you'd send later? Now, how many of those things do you actually remember? If you run a business, that's not a small problem. These days, a lot of clients find you on Instagram, and sooner or later, those conversations in the DMs turn into calls, and that's where things can slip through the cracks. A missed follow-up can cost you a client you'd already won. That's why I like HubSpot Smart CRM. It's a customer platform that updates itself, so the work after the call doesn't fall on you. Just add HubSpot's Notetaker to your Zoom, Google Meet, or Team calls. Once the call ends, the transcript is automatically added to the customer's record. HubSpot then suggests deal updates, pulls together the next steps, and drafts a personalized follow-up email based on your conversation. You review it, make any changes, and send it. That way, nothing falls through the cracks. In our team, this has become essential because with 35 people across different departments, having every agreement from every call land in one place is what keeps anything we promised from slipping through. If you're running a small business or just starting to land your first clients, you can try HubSpot Smart CRM using the link in the description. Now back to Adam. [swish] I wanted to go back to your personal setup that you mentioned.

    11. AM

      Okay.

    12. MM

      You know, the app that you created for your kids with the podcasts-

    13. AM

      Pods, yeah

    14. MM

      ... on different topics.

    15. AM

      Not a good name.

    16. MM

      Uh, so I use NotebookLM for podcasts.

    17. AM

      Yeah.

    18. MM

      I think it's very, uh- Kind of l-like the voices are not for kids. I use it to prep for my own-

    19. AM

      Yeah

    20. MM

      ... thing. Uh, what are the prompts and what are the things

  7. 12:17 – 13:53

    Switching AI Tools Without Starting From Zero

    1. MM

      that you injected in this app to make it exciting for kids?

    2. AM

      So my-- I've got three boys, and they are very into a variety of topics. But one is really into sports scores. So I just built a series of APIs to get a really good access to the most recent scores from any of the major leagues, and also really ground it in time 'cause LLM struggle sometimes knowing what day it is. So really [chuckles] kind of, you know, built some scaffolding around that idea. And so the main one for Blaze, my middle boy, is usually just like, what are the top scores for, you know, what are the major games of the Premier League over the weekend, and what are the scores and what happened? My oldest is much more into kind of history and certain to- topics, including sports. So he loves things like ten minutes on the best wingers of all time in soccer or something like that.

    3. MM

      Mm.

    4. AM

      Or who are the five best goalies today and why? I built a little app which allows you to just say how old the, how old the audience is-

    5. MM

      Mm

    6. AM

      ... uh, how long you want it to be, and then what you want it to be about.

    7. MM

      Mm.

    8. AM

      It's very-- It's a very lightweight, simple thing. But, um, I picked a voice that was very kid-friendly. I set it up in a way that it would be very, very kind of engaging for my kids specifically 'cause I know them. I was talking to my dad about it, and he's like, he's like, "I don't understand how AI works." So I just dragged the age slider to seventy-three, and I said, "Explain how AI works." [chuckles] When you create, you can say, what is it? What do you want it to cover? How long is the pod? You can drag it. You can just say, "I want it like a f-twenty-six-minute podcast." There's the listener age. My boys are six, eight, and ten right now. This is actually indexed to their birthdays, so these will change as they get older. But you could just be like seventy-five. You know, you write the prompt, and then you hit Make, and then you get a series of these podcasts.

  8. 13:53 – 15:10

    Using AI to Stress-Test Decisions

    1. AM

      Um, and there's some basic settings. You know, what model are you using? Here you can see I'm using Muse Spark, and, uh, you just set the speed. You can download it. You can send it to friends. So I could send you an email with a file of one. But this is just-- I build a lot of small apps. This is just one of those apps.

    2. MM

      Is there anything else you can show?

    3. AM

      This one is a trainer, so this is, like, my workouts every week, asking about, you know, how many things. I haven't done anything this week, so that's embarrassing. But you can see the overall plan, all this kind of stuff.

    4. MM

      Oh my goodness, that's a sophisticated app. What about-- Can you-- I don't know if you can show, but your knowledge database.

    5. AM

      Oh, yeah. So this one's called Custodians. I switch tools all the time. So, you know, we have our own tools, Muse, Meta AI. I've used Claude a lot. I've tried Duet recently. I've used ChatGPT a lot. Tried Dots. Used OpenCloud for a while. And when you switch all the time, it's annoying to have to try and reteach every new tool all of the things about you. So I wire each of them into this thing called Custodian, which just has shared activity and memory and skills, and so that they all can benefit from each other. So if I write like a, you know, this one's like an email voice, so this will say how to write an email like me and who has access to it. Um, but then they all can benefit from that.

    6. MM

      So it--

  9. 15:10 – 17:14

    The Three Qualities Adam Looks for When Hiring

    1. MM

      and it connects to all the apps that you're using and-

    2. AM

      They're all wired into Custodian, yeah

    3. MM

      ... and every app is instructed to send-

    4. AM

      Yeah

    5. MM

      ... facts, and then it checks whether it already has this information and-

    6. AM

      Yeah.

    7. MM

      Wow.

    8. AM

      So, uh, yeah, so every-everyone-- So basically, there's read and, read and write. So there's a memory system, which is just memory. It's text files. It's all mapped into an embedding space, so you can search based on concept, not just on keyword. And so they can all look up memories, and they're all instructed to, with certain guidelines, save memories into Custodian. That way, if something else comes out, I can just plug it in and see how it works for a while, and then unplug it.

    9. MM

      But you also-- You only use it for your personal. You don't have anything like that for work.

    10. AM

      I have a kinda... Yeah.

    11. MM

      Oh, you have?

    12. AM

      No, no. Work is a, a different system.

    13. MM

      How do you use AI to make decisions?

    14. AM

      I use it to stress-test them often.

    15. MM

      Mm.

    16. AM

      So if I'm-- If-- I mean, I try not to make a lot of-- This is gonna sound weird. I try not to make a ton of decisions. I probably make more than I should.

    17. MM

      You have this decision counter.

    18. AM

      Yeah, yeah.

    19. MM

      That's awesome.

    20. AM

      Yeah, I used to [chuckles] yeah, I just have the little-- I used to have a clicker. I used to click.

    21. MM

      [chuckles]

    22. AM

      When you manage a large team, you want to delegate as many decisions as you can to smart, capable people who are closer to the work, and if you're making all the decisions yourself, you become a bottleneck. I know this is contentious right now. I know this sort of, um, it's sort of in vogue right now in the Valley to sort of delegate as little as possible. But my philosophy is a little bit different. And so that doesn't mean I don't make decisions. I make decisions all the time. But whenever I can, if there's someone else who's better qualified to make that decision, I try to, what's the word? I try to delegate the decision to them, but in a very public way.

    23. MM

      Mm.

    24. AM

      So the whole team knows this is actually Max's decision, or this is Jenny's decision, or, you know, this is Nat's decision.

    25. MM

      Mm.

    26. AM

      Because if the team knows it's their call, they act very differently 'cause they know that person's on the hook, and that accountability is actually very empowering for those people. Otherwise, the team will often just keep defaulting to me, and I don't wanna become the bottleneck

  10. 17:14 – 18:25

    Why You Have to Be Willing to Sound Like an Idiot

    1. AM

      for progress.

    2. MM

      [chuckles] Love it. Uh, talk to me about how you're hiring now. Uh, is it different from five years ago?

    3. AM

      Yeah, no, I mean, it's... and it's changing really rapidly. So it depends on the function. So some of the things stay the same. So for all functions, functions are what we, we use the word function for, you know, engineering's a function, design's a function, data science is a function. So for any hire, for a long time, I've always looked for a combination of three things. Do you have some grit or hustle? And can you pick things up quickly? And can-- are you self-aware?

    4. MM

      Mm.

    5. AM

      So that you-- If you've got drive, you're a quick learner, and you can take feedback, you can kinda learn how to do anything with enough time. And if you are missing any of those things, you're often stuck. So I've always looked for those three things, but now I-- now some of that even matters even more, particularly the quick learner thing, because things are changing so quickly.

    6. MM

      How do you test for that during an interview or?

    7. AM

      I think you try to ask questions about things that they sh- wouldn't know about, and then, and then see how quickly they can pick it up if you have a conversation about it.

    8. MM

      Mm-hmm. Mm-hmm.

    9. AM

      And you can also take a look at their history or, um, you know, how quickly did they grow their responsibilities in their

  11. 18:25 – 21:00

    Why More AI Usage Doesn’t Always Mean More Productivity

    1. AM

      previous roles? Did they switch types of roles multiple times and find themselves successful? That's another indication that they're learning-

    2. MM

      Is that a good signal, changing roles?

    3. AM

      Depends on how-- You know, if you change roles every three months, it's-

    4. MM

      Yeah

    5. AM

      ... raising questions about y-your, you know, why. Um, but if I'm hiring for a very senior person and they spent a long time in data science, and then they moved into product staff or into engineering, and then they moved into another function, and they did a couple years in each, that's-- looks like a pretty well-rounded person to me, and I think that's pretty appealing. But the thing that's happening with, with all this technology shift is that the lines between the jobs, the roles themselves, are getting blurred, and so that's, that's also becoming much more complicated. You've got designers checking in code, and you've got data scientists putting together prototypes and mocks, and you've got engineers, like, pulling data and doing analyses.

    6. MM

      Yeah.

    7. AM

      It's all getting blurred.

    8. MM

      In this environment, who do you fire?

    9. AM

      Ooh. Um, [laughs] firing is hard.

    10. MM

      Who can't keep up?

    11. AM

      Like I said, what you-- it's just the inverse of what you're looking for. You're looking for people who are not necessarily senior but who are growing, who are becoming stronger and stronger, who are getting more capable month over month.

    12. MM

      Mm-hmm.

    13. AM

      Um, and so the people who make sense less to keep if you have to let go of people, I think, are people who are struggling to evolve and continue to learn in an environment where it's sort of we all need to continue to learn. I'm learning every day now, uh, in a way that I haven't had to in years. And so you just try to focus on helping everyone you can understand how the world is shifting, get comfortable trying new things, get comfortable failing. I feel like learning new ways of working is like learning a new language.

    14. MM

      Yeah.

    15. AM

      You have to be willing to sound like an idiot, um, until-- and get-- be corrected before you figure it out.

    16. MM

      Mm.

    17. AM

      And I think one barrier to entry for learning a second language for a lot of people, a lot of people can write relatively well, but they can't speak-

    18. MM

      Mm

    19. AM

      ... 'cause they're afraid of just sounding dumb. But I, I-- it feels like that to me. You have to be willing just to try, and every once in a while, it's just not gonna work, and that's okay.

    20. MM

      Is there any way you measure how AI impacts productivity? I know you're against, uh, token maxing-

    21. AM

      Token maxing is-

    22. MM

      ... whatever

    23. AM

      Well, we look at who uses the most tokens, and like-

    24. MM

      Mm

    25. AM

      ... half the time it's really, really productive, and half the time it's a complete waste of money, and you just try to decipher the difference. But I don't think it's a goal in and of itself.

    26. MM

      But how do you know-- Like, do you even care if AI is making someone more productive, or it's just the amount of work that they're able to do, and you don't care how they use AI?

    27. AM

      I care about how productive people are, and if

  12. 21:00 – 23:02

    What Instagram’s Algorithm Can and Can’t Measure

    1. AM

      AI or any other technology is an opportunity for people to be more productive, I care about them taking advantage of that opportunity.

    2. MM

      Mm.

    3. AM

      But you have to be really careful. But large companies can be very focused on metrics. Uh, they conflate metrics with goals-

    4. MM

      Mm

    5. AM

      ... all the time. You know, a goal might be we want people to feel comfortable sharing more on the platform. The metric might be how many people today share a story on Instagram.

    6. MM

      Mm-hmm.

    7. AM

      Um, but the goal is a-- the pr-- the metric is a proxy for the goal. It's not the goal in and of itself. The, the goal has to be expressed in English or in a language, in prose.

    8. MM

      Mm.

    9. AM

      Um, and then you pick the best proxy metric you can. I don't set goals explicitly for AI usage or token usage or diffs or lines of code or experiments per developer, but I look at those things as leading indicators of productivity. Um, but at the end of the day, I care about actual impact. How much did you ship?

    10. MM

      Mm-hmm.

    11. AM

      How much more useful is Instagram? How much better do people feel about Instagram? How many more people use Instagram? Um, not, not these AI productivity metrics.

    12. MM

      If you've been hearing about AI agents and wondering how do you actually set up one, I put together a guide that walks you through your first one in about twenty minutes. It helps you choose what to automate and get it working with no coding required. You can get it through my free newsletter. The link is below. It's such an interesting topic that you mentioned, trying to quantify things that are hard to quantify. How do you look at it in general? Like, I don't know, how happy are people with Instagram? Is there a metric like that?

    13. AM

      Yeah, for sure. I mean-

    14. MM

      And how do you measure it?

    15. AM

      There's-- So we, we measure sentiment by we ask thousands of people, if not more every day-

    16. MM

      Mm-hmm

    17. AM

      ... uh, questions on the platform about satisfaction. How fun is Instagram? How satisfied are you with the experience? How much does Instagram care about its users? These types of metrics. These are called sentiment metrics. Most industries have what's called net promotion score, NPS.

    18. MM

      Yeah, NPS. Yeah.

    19. AM

      And so, you know, we look at that. Um, we look at how many people use Instagram on a given day. We look

  13. 23:02 – 25:59

    Why Views Don’t Tell the Whole Story

    1. AM

      at how many people have sent a message, how many people shared a story, how many people-- how many impressions, so how many things somebody saw. All these are all, these are all metrics that indicate overall the health of the app. But if you over-focus on any one, you risk becoming myopic and doing-

    2. MM

      Mm

    3. AM

      ... something that doesn't make sense.

    4. MM

      Is there something you still can't quantify?

    5. AM

      Yeah, there's lots that we can't quantify.

    6. MM

      What do you do then? Can you give me an example? Like...

    7. AM

      Ranking is a great example. So w-we rank content, and just for saying this, there's gonna be some angry comments on whatever you post, so just brace yourself. But we rank because we wanna-- there's way more content on Instagram than anyone has time to see, even from the accounts they follow, but certainly from the accounts that they don't, that they might love, and so we want to get them the best stuff. We make mistakes, and ranking isn't perfect by any means, but if we don't rank content, we see not only does usage of Instagram go way down, but sentiment goes way down.

    8. MM

      Mm.

    9. AM

      Satisfaction goes down. But ranking is imperfect in a bunch of ways, like I said, and one of them is you rank for things you can measure, and I haven't seen anything that suggests that there's a correlation between how easy it is to measure something and how valuable it actually is.

    10. MM

      Mm.

    11. AM

      Right? So-

    12. MM

      Oh, yeah

    13. AM

      ... if you think about people's preferences and interests, right? One way to think about it is there's first-order preferences and second-order preferences. So I like how this donut tastes is like a first-order preference. It's like, you know, it's a positive experience in the moment. A second-order preference is like, "I wanna be healthy."

    14. MM

      Mm-hmm.

    15. AM

      "I wanna eat well."

    16. MM

      Mm.

    17. AM

      You know, "I wanna lose weight," or, "I wanna gain muscle," whatever it might be. Those two things can be in tension, and w- First-order preferences, because they are so quick, are often easier to measure. So did I like the reel? Did I send it to a friend? These are, you know, am I gonna click on the image? These are much shorter, uh, interactions. They're much more correlated with your first-order preferences, and they're much easier to measure-

    18. MM

      Mm-hmm

    19. AM

      ... than was this reel worth my time? Is this something I'm gonna talk about with my wife or my best friend tonight?

    20. MM

      Yeah.

    21. AM

      You know, is this something I'm gonna think about tomorrow? Uh, but those aren't any less important. They're just more difficult to measure. And so we are always trying to evolve and try to think of ways to better understand these harder-to-measure, but possibly and often more important proxies for how interested you might be in something.

    22. MM

      That's s- such a great thing because I just did a questionnaire for my audience. I posted three amazing guests, and two of them got incredible views, and another one I knew was iconic, Brian Chesky. He got less views, like, than average. But I knew from my conversation with him that I've learned so much, and, like, we're optimizing for views. But, like, there's something else I'm sure people are remembering. So we did a survey, and 44% pointed at Brian Chesky being the best episode.

    23. AM

      Yeah.

    24. MM

      And it's like, for me as a creator, I want to optimize for that. I want to optimize for impact-

    25. AM

      Yeah

    26. MM

      ... not just views. 'Cause

  14. 25:59 – 27:47

    How Adam Decides What Not to Build

    1. MM

      views, you go into this vi- virality, uh-

    2. AM

      Yeah

    3. MM

      ... strategy, and it doesn't bring you satisfaction. Like, I don't get satisfaction from just views. How would you solve the problem with, with the algorithm?

    4. AM

      On the-- your point about your interviews, I think it's super important to say that r- reach is a means to an end, not an end in and of itself.

    5. MM

      Mm-hmm.

    6. AM

      Right? You want reach so that you can evangelize a cause or sell a product or build a brand or nurture a community, r- ins- And not all views are created equal.

    7. MM

      Yeah.

    8. AM

      Right? So maybe the Brian interview reached less people, but it had more emotional impact with the people it did reach, which might mean a more loyal audience or a more excited audience for you long term in a way-

    9. MM

      Mm-hmm

    10. AM

      ... that's very difficult to measure. Um, views is a relatively shallow measure of engagement. Did you view it for three seconds?

    11. MM

      Yeah.

    12. AM

      Um, that's not a huge bar.

    13. MM

      Mm-hmm.

    14. AM

      You know, you can-- There are metrics that might be, m- might have looked different. I don't know. But if you can look at watch time, for instance-

    15. MM

      It was pretty much the same. That's the thing

    16. AM

      ... pretty the same thing? Yeah.

    17. MM

      Yeah.

    18. AM

      So then it might have just been a more narrow audience that cares more.

    19. MM

      Yeah.

    20. AM

      But that might be okay because maybe that's the brand you wanna build or maybe that's the audience that's gonna be the most valuable to your advertisers, right? You know, there's lots of examples of this. I mean, Acquired is a great example, right?

    21. MM

      Mm.

    22. AM

      So David and Ben are doing amazing things. They do not have the biggest audience. They have a big audience, but it's not the biggest podcast audience by a long shot.

    23. MM

      Mm-hmm.

    24. AM

      It is a disproportionately valuable audience.

    25. MM

      Mm.

    26. AM

      Their audience are people who are mostly or a lot of C-suite executives who are making pretty big purchasing decisions.

    27. MM

      Mm-hmm.

    28. AM

      And so for their advertisers, who are often, you know, B2B enterprise products, it doesn't really matter, you know, if they got ten

  15. 27:47 – 30:15

    How to Spot Your Next Opportunity

    1. AM

      thousand or a hundred thousand or a million views on a thing. You know as one of their advertisers that you're buying access to a disproportionately important, powerful decision-making audience, even if the number isn't that big-

    2. MM

      Oh, yeah

    3. AM

      ... uh, where, you know, one incremental sale might, like, make the entire investment worthwhile. So, you know, it's important to know what kind of business are you.

    4. MM

      Mm-hmm.

    5. AM

      What kind of creator are you? And then not to just lose context in the averages or in the totals.

    6. MM

      It is so hard, though.

    7. AM

      It's very hard.

    8. MM

      'Cause you are distracted by all the signals around you, by all the new people emerging. Um, how do you make those decisions about what not to build at Instagram when there is so much noise? I know you're building a lot, but so-

    9. AM

      I know. I wish that we're building too much.

    10. MM

      [laughs]

    11. AM

      [laughs]

    12. MM

      Well, how do you make a decision about not doing something?

    13. AM

      I've always wanted to, and I have been more and less successful at different times in my career at having not just me, but my team focused on doing less things and doing them at a higher quality bar than doing more things and spreading ourselves too thin. In a world where it's becoming so much easier to build more things and there is so much competition and people are becoming more efficient and I wanna make sure I'm employing as many people as possible, there's really strong incentives just to do more and more things.

    14. MM

      Mm.

    15. AM

      And that is just a very tough tension to wrestle with. And our ability to build more things isn't just based on the number of people we have. It's often based on talent density, for lack of a better term. So how many people do you have that you can really trust to make sure that something goes really well?

    16. MM

      Mm.

    17. AM

      If you have, you know, 20 really difficult, gnarly problems and you have a thousand people working for you, but you only have 15 that you can really trust to be fully focused on a really gnarly problem and have it make sure it go well, you're in trouble-

    18. MM

      Mm-hmm

    19. AM

      ... even though you have a thousand people. So we try to look at what is the opportunity, not just today, but in a few years' time. So we try to invest in things that are gonna be more important in a couple years, not less. We try to invest in things that are directly related to our identity. We're about connecting people over creativity, right? So we're not building, you know, a shoe manufacturing plant or something. It has to be, like, related to what we do, and it has to be relevant with our most important audiences, which are, you know, young people and creators.

    20. MM

      Hmm. You mentioned something of you, how you t-think about what's gonna be important in two years. For people who are listening, what would be your advice to them

  16. 30:15 – 32:39

    Finding Useful Feedback in Comments and DMs

    1. MM

      based on what you see, uh, in the future?

    2. AM

      As a creator or as a engineer or anything?

    3. MM

      Anyone who's looking for a new opportunity.

    4. AM

      I think that... I mean, the generic advice, maybe this isn't that interesting, so you should call me out if this is a cop-out of an answer, is that trajectories matter more than points. So it doesn't matter that much where something is today, so how many people use a specific thing or how strong an engineer is, you know, today. It matters how, you know, is it, are they getting better or worse-

    5. MM

      Mm-hmm

    6. AM

      ... or is the market growing or shrinking? And if so, how quickly? I would much rather build a product for a market that is growing rapidly than build a product for a bigger market that is declining.

    7. MM

      Yeah.

    8. AM

      I'd much rather hire someone who is relatively inexperienced and not as capable as another candidate, but is getting much better really, really quickly than hire someone who is completely capped out and is no longer really evolving or learning. So whatever your industry is or whatever your job is, I think it's important to have a point of view if you wanna think long-term about not just where everything is today, but what the trajectories of all those relevant pieces of information might be.

    9. MM

      Something that I've talked about, uh, to my community as well, 'cause I was building this study abroad booking platform.

    10. AM

      Right.

    11. MM

      And then study abroad was completely ruined by COVID, and, like, I was swimming against the, the stream, right?

    12. AM

      Yeah.

    13. MM

      And then I pivoted to being a creator.

    14. AM

      Yeah.

    15. MM

      And creator economy is just growing and booming. And, like, how with less effort I'm able to achieve more. How do you think about the future? What signals are you looking at?

    16. AM

      We try not to look at just where is the attention, but what's culturally relevant. Like, what is resonating with the audiences that we care about most? And so, you know, you can take a look... There's all sorts of ways to get signals, right? You can, you can do surveys, you can do ethnography studies, you can run research. You can just read your DMs-

    17. MM

      Mm-hmm

    18. AM

      ... or your comments, which can be a little crazy, but there's some real gems in there. You gotta travel. A big thing for me is to get out. It's hard 'cause I got little kids, but I try to get out there in the world and try and meet people who are using the platform in other countries, 'cause the vast majority of our users are not here in the US.

    19. MM

      Yeah.

    20. AM

      Um, and ground yourself in what people's actual experiences are. It's not something you can complete. It's a process that is ongoing.

    21. MM

      Before we keep going, I'd love to get to know you a little better.

  17. 32:39 – 33:56

    What Creators Can Ask an AI Assistant

    1. MM

      I share a lot about my work here, and I'm curious about yours. What do you do, and what you'd like more help with? We've put together a little survey. Your answers will help me decide what to cover in future conversations and what resources to put together for you. I'll leave the link in the description. Please, please, please take a couple minutes and tell me about you, about your business, about your ambition, so I can make better content for you. I really, really appreciate you. Thank you so much for doing this. You mentioned something about comments and feedback. How do you work when you're getting so much different feedback from really negative to really positive?

    2. AM

      So I think that, okay, i- if you work on a platform like ours, a lot of people are gonna use it, and they're gonna have... Even if it's a small percentage of people, particularly if it's not, there's gonna be a lot of strong opinions 'cause it's just a lot of people. And so you kind of have to embrace that and understand that even though you might get brigaded by a bunch of people who are very frustrated with you or the platform or the team for some reason or the other, it's actually a privilege in a lot of ways to work on something that reaches so many people. So I try to embrace that. Uh, and then I try to find the signal, and I try to ignore the noise.

    3. MM

      Mm-hmm.

    4. AM

      And so just because I'm in the

  18. 33:56 – 36:02

    The Metrics That Matter for Reaching New Audiences

    1. AM

      job I'm in, every time I post anything, there's gonna be a lot of negative comments, uh, because... you know, about small things. So you're just gonna get a lot of that, and you just need to accept that.

    2. MM

      So what's a signal, and what's noise?

    3. AM

      I try to just find the more interesting bits. So I try to go into my DMs and just find interesting p- points of view and questions.

    4. MM

      Mm.

    5. AM

      Or I, I mean, I do an ask me anything every Friday on stories, and I go through, as you know, there's usually a couple thousand questions, and I try to find some good questions and answer them. This is actually one area where AI's been pretty useful. Actually, Muse, I think, is gonna be, if we do, if we're clever, really useful for creators 'cause I have it now read all my requests-

    6. MM

      Mm

    7. AM

      ... in my inbox and let me know if there are any interesting questions in there or if any notable people have reached out. 'Cause I've missed, like, big names-

    8. MM

      Right

    9. AM

      ... have DM'd me before, and I just have not known for a year-

    10. MM

      Mm-hmm

    11. AM

      ... and been really embarrassed. And so I try to do both. So, you know, every day it sends me a couple every morning, you know, "This person messaged you, and you probably should write back," and then, "Here are five, like, legitimately interesting questions."

    12. MM

      Mm-hmm.

    13. AM

      And then I'll just go answer them. And, um, so I'm, I'm, [laughs] I'm outsourcing that a little bit right now. But for years I've gone into my requests folder and to my comments and just try to find a couple people to respond to.

    14. MM

      Mm-hmm. I really like using Muse as a creator because it spots signals that I can't spot with just exporting a CSV from my Meta Business Suite. I think one of the signals it gave me is that my family lifestyle content doesn't reach new audience, but it reengages my old audience.

    15. AM

      Yeah, that was a really interesting one.

    16. MM

      And all the other tools would tell me to just drop that content 'cause I'm not growing with it.

    17. AM

      Yeah, yeah.

    18. MM

      And I had this internal feeling like I need to deliver for my audience, and then Muse confirmed that.

    19. AM

      Yeah.

    20. MM

      What are the top things creators should be asking Muse?

    21. AM

      So it's not actually as good as it should be at a bunch of different use cases. I'm interested in a couple areas. I'm interested in leaning on Muse to help me understand and engage with my audience. I'm, uh, interested in leaning on Muse to help me understand what's working and what's not about my content strategy. I'm interested in leaning on Muse to understand

  19. 36:02 – 36:59

    Do “Comment for a Link” Tactics Work?

    1. AM

      what trends are happening out there in the world, not just on Instagram, but on other platforms. There's, like, a bunch of different things that I think it could do really well, and right now it's just not as good as it could be at a bunch of those, but it's getting better and quickly. So one interesting question you could ask as a creator are who are the most influential people who follow me? I actually think it's surprisingly good with very little work at understanding trends right now.

    2. MM

      Mm-hmm.

    3. AM

      When I asked about what trends are happening on Threads and on Instagram that are relevant to my job, it gives me a really good answer, um, which I was pleasantly surprised by 'cause we haven't done any custom work for that. Whereas for basic audience management, there's a lot of stuff that's not wired up.

    4. MM

      Mm-hmm.

    5. AM

      So, um, so there's a bunch of stuff to do there, so... But I do think it's gonna get much better quickly. I'm pretty excited about this.

    6. MM

      What's the metric you're optimizing your content for?

    7. AM

      I don't think I have one metric. I feel like it's my job to explain w- what we're doing and why, and that

  20. 36:59 – 37:59

    Adam’s Favorite Underused Instagram Features

    1. AM

      might not Convince people, right? There's m- people will disagree, and that's reasonable. My job isn't to convince people necessarily, but it's to at least explain that there's some rationale-

    2. MM

      Mm

    3. AM

      ... and most of these decisions-

    4. MM

      Mm

    5. AM

      ... are complicated, and there almost always are trade-offs. I try to focus on that. In terms of what correlates most with reach, it's what you'd expect, you know. So I think a huge thing that matters for unconnected reach, so reaching people who don't follow your account, it's definitely sends per impression. You always have to look at engagement rates, not just totals. So of every 1,000 people who saw my Reel, how many sent it to a friend?

    6. MM

      Mm-hmm.

    7. AM

      That matters a lot for how much it's gonna grow, uh, reach people who don't follow you. Whereas for people who follow you, watch time and like rate are much more important. Um, but I try to just focus on that core thing. And so I'll, I'll do things that I don't think actually... That I know don't perform as well, but they're more aligned with the mission or the goal. So I do these creator spotlights where I just ask a creator... It's changed how I do them 'cause what used to work doesn't, you know, has changed

  21. 37:59 – 39:03

    What Trial Reels Were Actually Built For

    1. AM

      over the years. But if I do one, I ask them for a selfie, and I just put a carousel of a bunch of their content, and I say, "Go follow this person, they're cool."

    2. MM

      Mm-hmm.

    3. AM

      And I love that because what makes Instagram great is all these creative voices on the platform. Those carousels don't do as well as a product announcement or a photo dump of, like, my personal life or a explana- explainer about how ranking works, but I don't care 'cause they're m- they're actually 100% aligned with what we want. We wanna be highlighting creative voices. We wanna make sure that we com- uh, communicate how much we care about creativity. We wanna spread the love and give people their flowers. And so it does well enough that I don't need to evolve the format very much, but it doesn't do as well as my other series, so to speak, for lack of a better word.

    4. MM

      Mm-hmm.

    5. AM

      Um, and so again, that's a good example of y- you have to remember, why are you there? Why are you posting? It's not just to maximize the number of views you get.

    6. MM

      Yeah, totally. Have a couple quick questions about some other features. DM me to get something.

    7. AM

      Mm.

    8. MM

      This is the feature that really helps me get

  22. 39:03 – 40:20

    What Taste Has to Do With Choosing Music

    1. MM

      a lot of views, but at the same time, people are complaining that the comment section was something interesting, and now it's a repetitive keyword and a generic answer.

    2. AM

      Yeah.

    3. MM

      What do you think about those? [laughs]

    4. AM

      It's a little bit of an engagement hack. I don't love it-

    5. MM

      Yeah

    6. AM

      ... but it definitely does work right now. Uh-

    7. MM

      It does. And the, the reach is like, we tested a couple Reels, 100% difference.

    8. AM

      I know, but is that, is that a good thing though? Like, from a consumer point of view. So it works as a hack for you to get more reach. Even if they don't care about the comments, is that Reel that gets way more reach actually more interesting to more people-

    9. MM

      Mm

    10. AM

      ... or is it just getting more reach?

    11. MM

      It's getting more reach. I'm getting more subscribers to my newsletter. I know, it's working for me.

    12. AM

      Well, if people are messaging you and having conversations with you, that's pretty cool.

    13. MM

      Peop- don't kill it. People are gonna kill me if [laughs] I brought it up anyway. [laughs]

    14. AM

      It's not the first time I've heard about this one.

    15. MM

      Please don't. [laughs]

    16. AM

      They also do comment for... I mean, it's, you know, it's a thing.

    17. MM

      Okay. What's the most underused feature on Instagram that will help people grow?

    18. AM

      Oh, I was gonna answer one that is not gonna help you grow. I mean, I- my favorite features on Instagram are not used by a lot of people. My favorite features on Instagram, and they're not gonna help you grow, but there's, um... So I'll, I'll come back to that. Blend. Do you know what Blend is?

    19. MM

      No.

    20. AM

      See? I know, no one knows about Blend. So in a DM thread with someone, you can start a Blend, which is a feed of Reels for both of you, and you can see which is for you and which one's for me, and it's a wildly interesting

  23. 40:20 – 44:38

    How to Get Better at Learning New Things

    1. AM

      way to get a little insight-

    2. MM

      Mm

    3. AM

      ... into what someone's interests are.

    4. MM

      Okay.

    5. AM

      It's rad. Uh, or your algorithm. So if you go to the Reels tab or Explore, there's a little couple hearts with some sliders. You can tap, and you can actually see what we, what topics we think you're interested in.

    6. MM

      Mm.

    7. AM

      So it's the first version of being able to see your algorithm and then edit it.

    8. MM

      Mm.

    9. AM

      Again, doesn't help you grow, but these are some of my favorite features. In terms of helping you grow, I mean, th- there are no major secrets here. I think, uh, for those who create videos, Trial Reels are really helpful. And-

    10. MM

      Can I drop a feature request here?

    11. AM

      Yeah.

    12. MM

      Trial Reels. So the, right now what I'm noticing on Instagram, you post one video that is maybe you staring at the camera, but then you test different, um, overlays with different text.

    13. AM

      Yeah.

    14. MM

      Instagram reads this as the same-

    15. AM

      Yeah

    16. MM

      ... video, so I can't really test-

    17. AM

      Yeah

    18. MM

      ... different-

    19. AM

      I think we need to build it in a way where you could just say, "This is a test"-

    20. MM

      Mm-hmm

    21. AM

      ... and we could treat it as a test. Right now, the point of Trial Reels is not to be a testing platform. It's to be a place where you can post a Reel and see how you feel about it and how it does before putting it on your profile forever and bothering-

    22. MM

      Mm

    23. AM

      ... your followers with it. So yes, it's like a test, but it's not like an A/B test-

    24. MM

      Yeah

    25. AM

      ... I wanna run five variants.

    26. MM

      What we're doing is basically A/B testing different-

    27. AM

      Yeah. You're hacking it to do A/B tests.

    28. MM

      Yeah.

    29. AM

      So I think we should figure out a way to support that first class-

    30. MM

      Mm-hmm

Episode duration: 44:38

In the uListen app, the transcript lights up word by word as you listen. Get uListen

Transcript of episode CN8c1edTCew

Get more out of YouTube videos.

Read along on your phone, search it beside the video, or ask your favorite AI app what was said.