Unorthodox frameworks for growing your product, career, and impact | Bangaly Kaba (YT, IG, FB)

[00:00] You were early growth PM at Facebook. You were head of growth at Instagram. You're VP of product at Instacart. You're now director of product management at YouTube. And I've heard that you've had a lot of impact on a lot of different cultures. I found this framework travels with me. It's got like these five components to it. Vision, skills, incentives, resources, action plan. And you need all of those to have change. And then within those buckets, you've got to figure out what are the right levers that you need to pull? What are the things that are missing? You're really big on something you call understand work. [00:30] the anti-pattern of what we want to do. Someone says, hey, you know what, this would be great to build. Then you go pull data to go justify why that would be great to build. Call that identify, justify, execute. First, you have to really understand from first principles, what is actually going on? So understand, identify, execute. You wrote this legendary blog post called how to choose where to work and what to work on. There's impact that you're really trying to draw. And the impact is only achievable by looking at a set of variables related to the environment, a set of variables related to your skills. [00:59] Today's guest is Bengali Kaba. Bengali was an early growth PM at Facebook where he was responsible for how people make friends on Facebook. [01:09] He was head of growth at Instagram, where he helped scale the platform to over 1 billion users. [01:13] He was also VP of product at Instacart. [01:16] He's also worked with tons of amazing startups as a growth advisor, including Twitter. He's now Director of Product Management at YouTube, where from what I hear, he's already made a huge dent. This conversation went long because there was so much gold to be extracted from Bengali's head, and I could not stop myself from learning everything I could in our time together. This episode is for anyone looking to level up their product and growth chops, or also just do better in your career. We dig into his framework for how to choose where to work and what to work on,

[01:46] the importance of spending time on something he calls understand work, his adjacent user theory and how it can help you drive growth, a bunch of advice for coaching, product managers, and managers of managers, tons of lessons and stories from his time at Instagram, Facebook, and YouTube, and so much more. If you enjoy this podcast, don't forget to subscribe and follow it in your favorite podcasting app or YouTube. It's the best way to avoid missing future episodes, and it helps the podcast tremendously. With that, I bring you Bengali Kaaba. [02:18] Bengali, thank you so much for being here and welcome to the podcast. [02:22] Thanks for having me. Excited to be here. So many previous guests have recommended that I get you on this podcast, which I already knew. [02:29] Funny story, when I first launched this podcast, I asked you to be on it, you're like, sure. And I included you on my launch poster of all the guests that are going to be on the podcast. [02:38] And then you decided to take on very hard work and jobs that kept you from having time. And so I'm really excited that we're finally doing this. [02:47] I'm glad we're finally making a reality. Sorry about that, Lenny. No sweat. You actually mentioned to me that somebody came up to you in Zurich and was like, "I'm excited for you on Lenny's podcast." [02:57] Yeah, it was crazy. I was like, you know, visiting a team that I managed there, about to get back on a plane to go back to SFO. [03:03] And just standing there, you know, doing some work, minding my business. And I get on the plane. I'm talking to a colleague. And someone comes up to me. I don't think I've ever seen them before and said, [03:12] Hey, sorry to interrupt you. I'm so excited for your podcast with Lenny. I can't wait for it.

[03:17] and then just walks away. And I was like, what is going on right now? Lenny is a big deal. I don't even know how this person knows me. And that's how I knew, Lenny, that I had to reschedule with you. [03:29] People are coming up to me and telling me that they're excited. I was like, there is a lot of anticipation. And Lenny, like the power of your reach now is like legit. [03:37] That is hilarious. That's like a new strategy for me to get people on the podcast, just say they're going to be on the podcast and then the pressure will start. [03:44] Oh, yeah. I mean, totally. Okay. So there's two broad topics that I want to spend our time on today. [03:49] I want to talk about career advice and growth advice, and they're both essentially growth-oriented. One's career growth. One is product growth. How does that sound? [03:57] Sounds perfect. This episode is brought to you by Wizard, empowering product leaders to ideate and iterate faster than ever before with the power of AI. As a product manager, I often spend hours taking screenshots and then annotating them with feedback for my team. With Wizard, I can simply upload my screenshot and Wizard's AI will turn them into a fully editable UI design that I can then take, make tweaks to, and then share with my teams in minutes. [04:27] totally new ways to improve our product experience, I can use Wizards AI to generate new design concepts from simple text prompts and turn them into interactive prototypes effortlessly. There's a reason that over 2.6 million people have trusted Wizards to accelerate every phase of their product lifecycle and speed up time to market. Developers can even export UI components to React and CSS to speed up their development. Wizards drag and drop editor is super easy to use

[04:57] your entire team, even your CEO and customer service teams can contribute. Unlock all of Wizard's game-changing AI-powered features and more with 25% off Wizard's Pro Annual Plan. Visit uizard.io slash Lenny and use code Lenny to check out today. That's uizard.io slash Lenny. This episode is brought to you by Mercury. Mercury knows your financial operations are complex. [05:27] this way. I've been a super happy customer at Mercury for over a year now, and it's honestly hard to imagine a better online banking experience can exist. Most founders and finance teams have to couple together a patchwork of tools to reconcile transactions from different sources, work extra hard just to get a holistic view of cash flow and have maps to company priorities, and struggle to get answers from platforms that all speak different languages. Mercury knows that [05:57] pay and accounting capabilities, you can pay bills faster, stay in control of companies' spend, and speed up reconciliation. The end result is the precision control and focus that startups need to transform how they operate. Apply in minutes at mercury.com and join over 200,000 ambitious startups like mine that trust Mercury to get them to perform at their best. Mercury, the art of simplified finances. Mercury is a financial technology company, not a bank. Banking services

[06:27] Financial Group and Evolve Bank & Trust, members FDIC. [06:31] So let's maybe start with the career. [06:33] You wrote this legendary blog post called "How to Choose Where to Work and What to Work On" that a few people have mentioned to me was really impactful in their career. [06:42] And just to remind people of your career path, which they'll hear in the intro, but just to give people a reason to listen to your advice on career. [06:49] You were early growth PM at Facebook. You were head of growth at Instagram. You were VP of product at Instacart. [06:55] You worked with a ton of amazing startups as an advisor, including Twitter. [06:59] You're now director of product management at YouTube. [07:02] This is a career that many people would dream of having. [07:05] So let's just spend a little time on this topic of [07:07] how to choose where to work and what to work on. And I know you kind of have this framework in this post, so maybe that's a good way to start of just kind of how you broadly think about [07:15] where to work and what to work on. [07:17] Yeah, that blog post that you're referring to actually came out of [07:21] a personal struggle that I had when I was at Facebook and trying to decide [07:25] what my next move should be. I felt like I was kind of stuck. [07:30] I felt like I was working harder but not getting... [07:32] seeing incremental kind of benefit to the work that I was doing. And I knew that I needed to change, but emotionally I understood that, but I couldn't really [07:43] have an objective way of thinking about it. [07:45] And so I really push myself to figure out what is actually going on with my situation and how do I create a way that I can rely on objectively to understand what's actually going on. [07:56] And so I looked at that situation and I wrote that post and it's, you know, the framework is really that there's impact that you're really trying to drive. And that is the thing that is the most important. And the impact is only achievable by looking at two sets of kind of variables, a set of variables related to the environment.

[08:14] a set of variables related to your skills. [08:16] and really breaking down each. [08:19] in understanding what's happening in the environment bit by bit, and what's happening with your skills, [08:24] Where are you hindered? [08:26] structurally within the environment, [08:29] Where are your skills kind of lacking? And what do you have control over? And using that whole kind of output, that framework, to decide what makes the most sense. [08:39] Why is impact the key [08:41] output of this equation. I think for a lot of people that isn't necessarily [08:45] the intuitive [08:46] variable that they think is important to focus on. Why is that so important in your experience? [08:51] Yeah, I mean, I didn't really know how to think about what the right [08:57] thing to optimize for was initially. [08:59] I realize that it's not compensation. Compensation is a reflection of the input. [09:05] the impact that you're having. [09:06] Thank you. [09:06] And so and you're leveling like how much senior you are, how much scope you have. [09:13] is a derivative of how much impact you're driving. The more impact you're driving at your company, more people feel like you, [09:20] can operate independently and drive real results, the more that scope they'll give you. [09:24] So really impact became the thing to optimize for. It is the import and compensation becomes an output based on it. [09:32] I think this is a really important point that is easy to miss, and this is what I always [09:36] Tell people when they're looking for ways to get promoted and do well in companies just find ways to have more impact and [09:41] Can you maybe make it even more concrete? What does impact mean to you when you talk about impact?

[09:46] Yeah, I mean, so impact can be a lot of things. But I think for a product manager, for example, it's really one helping to get [09:54] drive extreme clarity about where there are problems with the product, where there's opportunities [10:01] and what is the right focus and prioritization. [10:03] That is actually a form of impact, just creating the clarity that people need to understand and believe in the investment. [10:10] The reason why I named this and it feels a little counterintuitive is that the more senior you get, [10:16] the more there are questions of are we even investing in the right place? [10:20] Is this area, is this team, is this org the right investment? And so being able to even create the clarity that there is opportunity, it is the right thing to do, it is strategically and structurally important, is a form of impact. [10:35] delivering on that impact, showing that you can make progress quickly. [10:39] that you can deliver fast lane wins, as Casey Wenders would say, or medium and slow lane wins, and then actually showing that you can do this again and again, [10:47] is how you actually validate the impact that you can see where the opportunity is and what's going on. Awesome. Okay. So this equation is impact equals environment type skills. [10:57] Can you talk a bit about how to work on these two elements? So the environment was the one to me that I think is most people overlooked and I overlooked when I was first thinking about this. [11:08] environment in this case, I think I kind of discreetly named a few things. One is your manager, [11:14] Then there's the resources. So what kind of team do you have?

[11:18] Is your team staffed appropriately? Do you have the right P&L, whatever budget, to get the things done that you need? Then there's a scope, like what is in your remit versus not in your remit. Because if you don't have enough scope, then you can't actually focus on the things that are most [11:34] The team itself, [11:36] the skills of the team, your compensation in some ways is part of the environment, because if you're not compensated failure, you don't believe you are, then it's hard to feel like [11:45] the work that you're doing is meaningful. [11:46] And then there is the last part is the company culture. [11:49] So to what extent? [11:51] Is the culture a place where you feel supported, included? You feel like you can do your best work. And so you're really looking at each one of these variables. And I look at this every year and I say, how is my manager doing? Like, how do I think about my manager? How do I think about the resources I have, the scope, the team, the conversation, the company culture? And to what extent? And I score them. [12:12] I score them as a one means it's kind of neutral. [12:16] A two means that I'm greatly, greatly benefiting from this situation. And like something even closer to zero is not in a good place. So I assign a score in quarter point increments, 0.25, 0.5, 0.751, up to two. [12:32] Every year, I really ask myself, what is the state of each one of these? And to what extent do I believe that they can and will change? [12:40] Wow, I love this. Okay, so... [12:42] There's this formula, impact equals environment times skills within environment. [12:47] There's these five variables, and they add up to 10 if they each give you two points. That's so cool. Okay, so the five, just to be clear.

[12:53] your manager, the resources you have, [12:56] is team's number three. [12:58] compensation and the culture of the team. And then your scope as well. And then scope. Okay, got it. Oh, there's six. Okay, got it. So it's up to 12. [13:08] And then, OK, so the idea here is [13:11] you score each of these of how you're feeling, [13:14] how the environment is contributing to the impact that you're delivering. [13:18] And if one of these is not [13:20] a great score that's an opportunity to improve your impact which will improve improve your career. [13:25] That's right. And you have to really be honest. I think part of what makes this framework so powerful for me, at least, is that it helps you to be honest around. [13:35] What are the things that are limiting your ability to have impact and for your skills to really land? [13:41] And to what extent [13:43] do you believe that you can help to change or try to influence the change in the environment [13:49] And because, you know, no one wants to be in a place where there's a bad culture and culture is a bad fit for you. [13:53] But if you're not really thinking about it objectively and naming that, [13:57] or maybe it's the culture of the team that is not the right place for you. [14:01] It really forces you to evaluate what's going on around you that's limiting your impact. [14:05] Is there an example from your career you could share where one of these was not [14:09] where it needed to be and either help change it or realize you had to get out of there. [14:14] When I was actually at Facebook, I was running the team that does all the people recommendations. [14:20] And it was a great team. I actually had a massive, massive

[14:24] I had 30 engineers I was working with, 15, [14:28] machine learning engineers and 15 front-end back-end engineers. [14:33] Great team, incredible team, a lot of resources. [14:36] ton of scope, in some ways it was too much scope. [14:40] And to me, that was problematic because I really needed to build out [14:45] I felt like multiple teams [14:47] to support the work that we were doing or to break it up because it was [14:52] I mean, the pace, the velocity you can imagine at Facebook was incredible. And I felt like... [14:58] between all of the work that we needed to do, the amount of engineering capacity that we had, [15:03] and the amount that was on the table, I felt like I wasn't resourced. [15:06] in a way where I felt like I could actually [15:09] deliver on all of the things that were necessary without burning myself out. [15:13] And I felt like I was burning myself out, and I couldn't really see the forest from the trees because there was so much to do. [15:19] And so that was a situation that I didn't really understand how [15:24] to navigate at that point in my career. And there were like two or three manager changes concurrently. [15:29] And so I didn't have a manager to lean on that I felt like I had a relationship with to help me to navigate that space. [15:35] And I felt like the scope was actually too much for what needed to be done. And so I needed to find a better kind of fit for me. [15:42] There was nothing wrong with the team. It was like a great... [15:44] learning environment but it's like the confluence of scope and manager [15:49] and all happening at once just wasn't a good fit for me. [15:53] A lot of people are in these situations where they...

[15:55] They'd say they go through this score... [15:57] they identify I have way too little scope, way too much scope, kind of like you described. [16:01] And there's always this question of, can I actually make a change, or is this just not [16:04] Am I not in a position, especially IC, product managers, [16:07] There's just always a lot of like, I can't actually change anything. [16:10] What do you often tell people around this that just feel like there's nothing I can do here and my manager sucks? What am I going to do? [16:16] Well, I do think one of the things that I recognize is your manager, not all of these variables is created equal. [16:23] And manager is the most important variable in the environment. [16:27] because a great manager who [16:30] is empathetic, who is aware of what's going on, who is a great communicator, [16:36] has the ability to move [16:38] the chess pieces around and to like fix some of these for you either immediately or in time. [16:43] There's no one other than the manager who can really help to increase your scope or to help make sure that the team has the right pieces in place or dial in some of the issues that you might see in culture. And so that is why people say they don't leave a job, they leave a manager. [16:59] because the manager is the one that has a lot of the power to fix a lot of these variables. [17:03] And so really the question becomes like, to what extent have you been able to [17:08] clearly articulate what of the, you know, and dispassionately articulate what are the challenges that you're having, you're seeing across some of these variables with the manager, help them to [17:18] tie it back to how it's impacting your work. [17:21] and like see if they can help you create a plan to like

[17:25] alleviate some of these things. [17:27] That's super interesting. Then your experience of the manager is kind of the core of a lot of these variables. [17:32] Is there anything-- [17:33] You recommend to people if their manager is not someone they like, is it just like try to find a new manager? [17:38] or potentially leave [17:39] You're never going to always like your manager, really. That's not the goal isn't to like your manager. Ideally, you [17:44] Respect them and they respect you. You feel like there's things that you can learn from them. [17:49] Finding a new manager is always an option, but I guess sometimes the real question is actually spending time to understand what is your manager optimizing? [17:57] A lot of times I think there is [17:59] a big disconnect between an icy [18:02] focusing on their discrete area and try to optimize for local maxima versus understanding, OK, my manager is thinking about these things and this is how I fit in. [18:12] and understanding [18:14] Maybe they have a gap to understanding why your area is important. [18:17] Or maybe there's stuff that's on your manager's plate that is actually adjacent to your remit that if you understood they were optimizing for, you can take that on and you would find more synergies with what they're trying to do. [18:27] I love that advice. [18:29] Let's talk about the other side of this equation. We've talked about environment. Let's talk about skills. [18:33] What do you advise there for folks that want to improve their skills? [18:37] The skills part is really, really big, and it's something that requires, I think, consistent evolution of your own. [18:44] abilities and so you know I broke it out in that blog post or kind of communication and [18:49] your ability to influence and your leadership. [18:52] strategic thinking and then execution, right? Like actually getting things done.

[18:55] The communication of these, again, not all of these are created equal. I think communication is the one that [19:01] tends to be [19:03] the most impactful. [19:05] And you see this in a lot of ways. You see this with people who are poor executors, but incredible communicators. And they seem to continue to rise and rise because, like, you know, they can tell a great story. But when you look under the covers, there's nothing there. There's no substance there. Right. So communication for better or for worse is one of the most important things. [19:21] But building out your skills, I think, is really just kind of a it's an interesting time to be in product and to be in tech right now because you have so many more ways to build out your skills than what previously existed. Just like so many incredible blog posts, like your podcast and blogs, so many incredible people who come on here. [19:40] who tell you things like how to go to market, how to think about B2B SaaS and metrics. And so like, you know, like listening to you, listening, reading Ben Thompson, understanding his mental models. You know, if you go to look at Elena Verner or whoever, right, like there's so many thought leaders who are out there. So I think being a voracious reader, [19:58] is really, really critical because it helps to build your toolkit. And you need arrows in your quiver to really understand how to think about the right framework and the right mental model at the right time. [20:09] One thing I also tell people is people think about mentorship is like, I have a mentor. Lenny is my mentor or John is my mentor. [20:17] I tell them it's actually better to have a stable of mentors. [20:20] You want to have three or four. [20:22] And ideally, what you do is you meet with each one of them once a month.

[20:27] more on a different Friday of the month. [20:29] Right. And so you might have three or four people on every Friday, you're meeting with someone different. The reason why this is so important is because of you, Lenny, or a mentor to me. You're busy one Friday and we only once a month. Now I don't have anyone to talk to for two months. [20:41] And you're going through a lot of things. You're not really able to bounce ideas or get built skills or learn how to influence others. But if you have three or four mentors, [20:50] If you have someone different to talk to every week, and if one or two of them cancel, you still have two people to talk to that month. [20:57] And you still can continue to grow and to build your thinking. So that's something I always recommend. And then the last thing I actually talk about is, [21:05] When it comes to execution, I think people [21:08] Don't especially product managers don't do enough of watching and learning from others. [21:14] There's an art and a science to product management. [21:17] And in a lot of ways, I come from a background of education and a lot of ways in product management. [21:22] It's very similar. You want to watch how other people [21:25] hone their craft, how they deliver, how they lead teams, and kind of steal things from them, like figure out, well, wow, Lenny did this really, really well when he was at Airbnb. It's really incredible. I'm going to take that. I'm going to take how he runs that team or how he landed that framework, [21:39] or how he communicated this message, [21:42] and note it and use it in my own toolkit later on. [21:45] And so if you're really a student of product, you can't just be a student of theory. You've got to be a student of practice, too. [21:51] Meaning you're going and you're looking, you think, hey, I'd love to sit in your team meetings. Can I come watch? Right. Or your PM meetings or your leadership meetings.

[21:58] And you really like doing that for the sake of learning is really, really incredible. And that's some of the ways I talk about building skills. [22:04] Is there an example of that latter lesson that you [22:07] learn from someone that comes to mind. [22:09] of watching someone and like, "Oh, I'm gonna do that." [22:13] Earlier in my career when I first got to [22:17] instagram um i wasn't the first pure in growth beyond but i was one of the first and um the guy who was there before me his name was georgia lee he actually stood up the growth team and [22:28] ended up leaving six months right after I joined. But like I watched George was like a really good [22:33] listener. [22:34] He had just the really incredible ability to be in a room [22:38] and hear what was going on to recast it back and make sure everyone felt good about where we were and what was being said in the path forward and really crystallizing [22:48] the [22:50] the action items and also like how people felt about what the next steps were and walking out of that meeting it felt like he always had [22:58] buy-in and clarity. [23:00] And it was helpful to build trust. You could see him almost like winning trust. [23:05] in every moment of that meeting. [23:08] And that was something that I really kind of admired and learned and [23:12] kind of sought to emulate later on in my career. That's an awesome example. It comes back to the skill of communication that you talked about, just the power [23:19] becoming really good at that, and this is such a simple [23:22] skill you just described is just recapping. Here's what we talked about. Here's what we're going to do. Here's the decisions we made. [23:27] There's action items. It's not actually difficult. It's just... It's not action. Yeah, but what was magical about the way he did it, it's like he would name people and their contributions and show how all of this came together into what is the path forward. That's amazing. It's communication, but it's also the other side. It's the other coin, the side of communication, which is listening to figure out how you communicate back to others. So good.

[23:50] Following up on something else he talked about, which is mentors. Anytime someone talks about finding mentors, everyone's always, "How do I find a mentor?" [23:57] Well, how do you what advice do you share with folks of like how to find a mentor to help you say once a week? [24:02] or once a month. [24:03] What I see people find the most success is they ask, they tell people what they're working on or what challenges they have. And they say, do you know someone who... [24:14] I might be able to learn from who has done this or has good thinking about this recently. [24:19] Right. So instead of just going to someone saying, hey, can you be a mentor? So coming to me and say, hey, Bengali, I'm actually trying to figure out how do I change the way this team operates because we need to go from this model to that model. But here are some of the challenges. So, you know, someone who's actually really good at changing the way teams work or really good at communicating a new vision that I can talk to. [24:40] And so that I think is really important because [24:44] What you do is you're creating a seed. [24:47] And the seed is, you know, there's like a triad. You have the person who's looking for the mentor, you have the recommender. [24:53] the recommender and then you have the potential mentor. And the recommender is basically saying, here, there's common purpose between the two of you. And I can see it. And I think this is important. So I want you all to come together. [25:04] And that triad in that moment [25:07] creates a higher affinity, a higher likelihood that the person who needs the help and the person who you're connecting with [25:13] are going to see mutual benefit from one. [25:16] Right. And I think I find that like just doing that and seeing that and being really,

[25:21] focused on what the opportunity or the challenge is, [25:25] tends to lead to better connections as opposed to just reaching out and saying, hey, I like your style. We'll be better. [25:30] So essentially share with folks you trust. Here's what I'm working on. [25:34] See if they recommend someone that could help you with that specific skill versus assuming that there's this person that can help you with this skill. [25:40] That's right. I love it. [25:42] Going back to your advice on career, something that you wrote about in your post, plus the [25:47] folks have told me you're really big on. Is this something you call understand work? Does that ring a bell? [25:52] Yes, it does. Okay, talk about that and why that's important. [25:55] I certainly cannot take credit for Understand Work. This is a part of an old Facebook framework, understand, identify, execute. What I probably will say is I guess I've been the shepherd for Understand Work and other companies. I've taken it from my time at Facebook and really institute it, or it rigorously at Instagram, then at Instacart, and when I worked, helped with Twitter, now at YouTube. The way I like to talk about it with my teams is, [26:21] that, you know, first I tell a story actually, and the story is, [26:24] We've all had a moment where we have worked on something with a team [26:28] Super excited finally launches and we go out to dinner and we celebrate and [26:33] And, you know, we're celebrating everyone's juiced. You know, it's great night. You go back to the office the next day and you look at the metrics and metrics are flat. [26:42] And everyone's a little bummed, like, why did this happen? [26:45] Why are the metrics flat? What happened? We worked so hard. And ultimately, when you unpack it, you realize that you built something that you thought was going to be a good.

[26:54] idea. [26:55] but you really didn't understand a lot of key components of what people really needed, what pain points really existed. [27:02] or what were the alternatives and what the real value of those alternatives were in the market [27:08] or exactly how the product needed to work with the flywheel or what the experience needed to be. [27:13] And so that is what I call the anti-pattern of what we want to do, and I call that identify, justify, execute. [27:20] When you identify something, [27:22] Someone says, hey, you know what? This would be great to build. You identify that. [27:25] And then you go pull data to go justify why that would be great to build. [27:29] and then you sink an ungodly amount of time working on it, [27:33] in order to make it work, but it ultimately doesn't succeed. And that is the anti-pattern to what is kind of like the Facebook kind of way of thinking, which is understand, identify, execute. [27:42] So first you have to really understand from first principles, what is actually going on? So when we talk about understand work, [27:49] There's a few ways to think about it. One is, [27:52] It is an intentional affordance [27:56] in your execution. [27:58] to do the work that helps you to de-risk a project and to learn what's going on. [28:03] When I say intentional affordance, meaning [28:05] You put it down like Lenny is going to work on this thing. Lenny as a product manager is going to write a strategy. Janice as a designer is going to design a prototype, like whatever the thing is. And you put it down as actual an item as opposed to assuming it's going to happen in the background. [28:20] we don't make an affordance for understand work, then the work doesn't get done, and everyone's just sprinting on execution.

[28:26] And so it's a planned intentional time. [28:30] to the team's bandwidth to figure out what is it that we need to do to understand what's happening. And so, you know, for example, with Instagram, we get a lot of understand work of [28:40] What makes for a good connection? [28:42] Like, how do we want to think about that? How do we want to make sure that we're actually growing a graph [28:47] a social graph that makes sense. And so I might work with data science to understand work, to pull a funnel and look at the different types of connections [28:56] for different types of users and make it make sense. And engineering might do understand work to instrument logging to make sure that we have the data that we need in order to tell the story. And so you're doing this understand work to basically better understand the gaps in knowledge. [29:14] What's also really interesting is that UnderstandLock helps you to clarify what is a root cause? [29:19] Like, what is a job to be done or like, what is the right use case? [29:24] And because the team adopts this mentality, it becomes this forcing function for execution. [29:32] So when someone says, you know what we should build, we should build this. [29:35] you have a team that's enabled and empowered to say, [29:39] That's a good idea, but we don't actually understand [29:42] These three things. [29:44] before we start working on it. So maybe we should do understand work [29:48] to make sure that that is actually the right idea. [29:51] Right. And so what you end up getting by embracing this concept is two things you get.

[29:56] parallel paths to work. [29:58] So every sprint, let's say you do a three week sprint or three month roadmap, [30:03] You're executing on the things that you have a lot of conviction around. [30:06] And you're also doing understand work in parallel. [30:09] And so at the end of the sprint, you have learnings from what you executed. Wow, that test or that launch worked really well. We should double down. Oh, wow, that test or launch didn't work. Here's how we should pair it. And you have insights from the understand work. [30:23] And you use both of those to plan the next sprint or roadmap. [30:28] Right. And so because you have [30:30] Pillow path. [30:31] you end up getting this velocity multiplier over time. [30:35] Right. So the next sprint and the third sprint and the fourth sprint, every subsequent sprint [30:40] You've de-risked new ideas. You've gotten more clarity. And so you do more execution, you do better execution, and you move faster. [30:48] The things you ship, you have a higher win rate on the things you ship. [30:52] right and you know your shipping when i remember when i left instagram this is many years ago but we had like i had 15 teens and [30:59] We might have... [31:00] been running 12 to 20 experiments, a quarter a team. [31:04] And I would say probably 60 to 70 percent of them were like, [31:07] positive. [31:08] and shippable. [31:09] which is incredible. I mean, you think about the most magnitude of that. And it was because we were so effective at de-risking and understanding what was going on. [31:17] This is really interesting. [31:19] I imagine many people listening are like, [31:22] put time into understanding. We run experiments. We write strategy docs. We do user research.

[31:27] What is it that you think people are missing? Is it that you dedicate people's [31:31] like actual time, like you're just doing understand work instead of building a new thing. [31:36] for this next sprint, you're just going to be telling us [31:39] what the problem here or opportunity is. How do you actually operationalize this versus just what people probably already do, which is [31:45] These are research data dives. [31:47] What I found is a couple of things. [31:50] when your team does not fully understand a problem space, [31:56] the balance of work [31:58] tends to be [31:59] higher. Like I think when I started at YouTube, we were doing 60% execution, 40% understand work. [32:06] And over time, as we understand more and more, [32:09] The mix. [32:10] shifts. Now my teams are probably doing 80 percent execution, 20 to or 80, 85 percent, 20 to 15 percent on the same. And so it's not just about writing a strategy. It's about saying, [32:22] okay if we have these themes of stuff that we want to work on [32:26] What do we know with confidence because the data tells us [32:30] And what do we need to understand because we don't have the data, we don't have the research, we don't have the insights. [32:35] and really being honest with yourselves around like what are the things that are [32:41] low to medium effort, but high likelihood of being impactful, because you have the data. And what are the things that you-- [32:48] or interested in perhaps assuming, but because you're missing something, you're missing [32:53] research or feedback from users or data or some insights or strategy, you should actually say, I'm going to go do this before I do something else.

[33:02] Right. You know, an example of this is one of my teams does ships paid virtual goods in the live experience on YouTube. [33:11] That's a really, really important and hard problem space. [33:15] And when I joined, we didn't really fully understand the whole live ecosystem, like how we lived in that live ecosystem. [33:21] how our products worked. [33:23] where the biggest opportunity was. So the first and most important piece of understand work was instead of shipping iterations to the current product, [33:31] we needed to actually get the funnel of what was happening. [33:34] what's happening with [33:36] How many people are watching live every day? How many people are clicking through? How many people are seeing our experience? How many people actually buying it? We didn't really... [33:43] fully have the experience mapped out. [33:46] And to understand where were the gaps and where the problem places and where should we be focusing? And so we had to do that. And that was understand work from multiple people, for multiple teams. We had to do on the engineering side. We needed data science as analysis. We need the PM to go and dog food. The experience to figure out what was broken. Like all of that was intentional and an affordance on our roadmaps. [34:07] What's interesting about this is that this is very counterintuitive to how people would probably approach [34:12] hey, we need to speed up execution. We need to speed up growth. We need to ship more... [34:17] More execute. Go faster. Do more. [34:19] And what you're saying is you find [34:21] the impact [34:23] comes from actually slow down [34:26] how much we're doing and spend more time understanding to [34:29] execute more intelligently. [34:31] That's right. Slow down and speed up.

[34:33] Fascinating. [34:34] Because it's interesting, I have a bunch of questions that emerged out of folks that you've worked with, and many of them [34:39] around how you turn a culture around. [34:42] speed things up and drive growth in a really [34:45] meaningful way. [34:46] And it sounds like this is one of your... [34:48] key strategies is [34:50] get people to spend more time understanding before [34:53] diving into a bunch of stuff. It does sound counterinsortive, but if you actually think about it, [34:57] What is a better outcome? Is it a better outcome to just ship more faster now, but most of the things are unimpactful? [35:05] Right. Or is it? [35:06] a better outcome to ship fewer things [35:09] but really work on making sure that you're shipping them in the best way [35:13] and de-risking a lot of other things so that a year later your win rate's higher and your velocity's higher. [35:19] Yeah, I think counterintuitive is the wrong word. I think it makes sense. I think it's just no one does this. [35:24] Usually everyone's like, "Move faster. We need more ship, more experiments." Well, I mean, this is the irony of growth is people think growth is overnight success. [35:33] And it's not. It is a lot of short wins and short term [35:39] execution for a longer term gain and really understanding you have a lot of short turns towards the longer term outcomes. [35:48] so that people can take away this lesson [35:52] Can you help people understand just when you say understand work, what does that look like generally? [35:57] just dedicate a time to dive into data and answer a bunch of questions you've [36:01] symptom is it

[36:03] uh, [36:04] running experiments, still test hypotheses? What does that usually look like when you're here, let's do 40% understand work? The understand work sometimes comes from me, but most often comes from the teams themselves. [36:16] and every function can do and should be doing understand work at some point. [36:22] And so it really depends on what the function is. So for an engineer, it could be looking at the code and saying, OK, we want to improve this. I need to do understand what to understand. Do we need to refactor this code and how scalable is it? And what do we need to do to make sure that we can execute fast and make sure that we're not going to have a lot of starting stuff? [36:43] That could be understand work. It could be actually instrumenting the data and making sure that we actually have full visibility into what's going on. [36:50] data science, like we work a lot with data science doing activation metrics and like understanding proxy metrics, right? Like that's understand work, but because it helps us [36:59] to figure out what we need to build and where to focus. [37:03] you know, for product management, sometimes understand work is [37:08] figuring out the partnership strategy. [37:10] ahead of like actually launching the product because you need to go figure out how the pieces are going to come together. [37:16] And so it really just depends on what the function is. But when I say it should be coming from the bottoms up, what I mean is I encourage the team when we plan a sprint or plan a roadmap, [37:27] to ask [37:28] or questions to identify the key themes that they need to work on [37:32] And when they ask a question on a key theme, also ask,

[37:35] What else do we need to understand to make this happen? [37:38] And to in that planning session to make sure that you're including cross functional partners, not just product, not just product design and data science and you also include go to market. You also include marketing. [37:51] Because if you're not inclusive, then you don't really understand what the issues are. [37:55] My PM brain is... [37:56] afraid of creating too much understand work and nothing getting done. How do you find that balance? So we're just sitting we're going to be understanding for hours and days and weeks and months and [38:08] Not shipping much. [38:09] How do you do someone? I tell them you have to ship. I mean, you always have to ship, right? Like, it's-- sometimes I give them-- [38:17] Sometimes it's helpful, especially early on, because it is a it takes a while for people to get their head around why this is so critical. [38:24] I say, you know, [38:26] we should choose sometimes I would give them guidance. [38:29] We should choose three to four understand projects that are going to really help us. [38:34] this roadmap. [38:36] and figure out what they are. [38:37] So choose your top three or four, and then let's talk about it, and I'll give them guidance. [38:42] or I would say figure out [38:44] Um, [38:46] there should always be some execution. So initially, [38:49] you're looking for low effort, high impact things to execute against. [38:54] build a portfolio of work to do every sprint or every roadmap, [38:58] some of which should be low effort, high impact, some of which should be medium effort, high impact. [39:03] And sometimes understand what actually looks like doing a cheap test.

[39:07] Doing a task that's going to help us to learn as fast as possible, that we think is a good enough experience, [39:13] that can inform us, right? So identifying ways to do that. [39:17] And so it's really about just managing expectations and helping people to be clear that [39:21] The goal is to ship the product, but you want to ship the things that you have more confidence and understanding. [39:28] Versus that. [39:29] So I think a big takeaway here is if you want to [39:32] have more impact, move faster, [39:35] Try to spend a little more time understanding [39:37] the problems you're going after in the opportunity space. That's right. I'll give you a very tactical example. When I joined Instagram in January 20, [39:45] In 2016, believe it or not, the onboarding, the signup flow at Instagram had literally no logging to it. [39:53] It had logging of how many people started and how many people ended. [39:57] And I joined in January and was like, we had to write a roadmap. [40:00] And so the roadmap looked like this. [40:03] Okay, we know this amount at the top of the funnel and this amount at the bottom of the funnel and there is eight steps in the funnel and we don't know what is going on. [40:10] Right. So the first bit of understand work was we needed to do [40:14] like the instrumentation, [40:16] of that funnel as fast as possible to get the data to figure out where the drops were happening and what to fix. [40:23] Because we knew of what was happening at the top of the funnel and the bottom of the funnel, we can go and play around with the experience. [40:31] and see what was broken. So we ran a bunch of tests of stuff that was obviously broken to see what would [40:36] what would improve, right? And so it was like a mix. And so what we did, we set up time where we did,

[40:42] the beginning of life. [40:43] First couple of weeks, add logging, ship it to code, [40:47] reevaluate in the middle of the quarter, look at the full funnel, and then add more things that we can do later on once we got that. But like in order to get that done, that involved understand work with growth marketing to figure out what was the schemas with instrumentation. [40:59] The engineering to actually do that log in, right? Data science, like pull together funnels and dashboard. [41:04] All of that had to come together all at one time. [41:07] I really like just this concept of calling it understand work. I think that alone is a powerful [41:11] tool. We're going to spend more time on understand work. [41:14] It feels like it gives meaning to stuff that otherwise people would brush aside. Right. Where it's just like, no, no, let's just ship stuff. Let's just try stuff. Let's just... [41:22] test this step. We'll see what happens. [41:25] That's right. [41:25] Is there anything else you have seen and often do to help a team you join [41:31] move faster and grow. [41:34] bigger. And I've heard that you've had a lot of impact on a lot of different cultures. [41:38] So I'm curious if there's anything else that is really effective. [41:42] I found myself in a bunch of interesting situations where I had to come in and help improve cultures or change cultures around teams. [41:49] I found this framework. I don't even actually remember where I got it from, but it travels from computer to computer with me. [41:55] We're seeing the team. [41:56] It's called Managing Complex Change. I actually think I got it from a business school or something. [42:01] And it's really interesting. It's got like these five components to it. [42:05] of this vision [42:07] skills, [42:08] incentives, [42:10] resources, action plan, and you need all of those

[42:14] to have change, right? Like you need teams that have vision, they need to have the skills, [42:19] They need to have the right incentives. [42:21] Sometimes some teams are incentivized to do some things versus others. You have the right resources in the right places. You need to have a clear action plan. [42:28] And what I love about this framework is if you can visualize it, maybe you can share it with your podcast. Yeah, we'll put it up on the screen on YouTube so folks can see what you're talking about. [42:37] Basically, what it does is it shows where if you're missing any one component, you get different [42:43] outcomes right so if you're missing the vision component [42:46] you end up in a state of confusion. [42:48] Or if you're mentioning the incentives, you end up in a state of resistance because people aren't incentivized to do the right work. [42:54] Or if you're missing an action plan, you end up in a state of false starts. [42:59] So I use this and I think about this a lot actually because [43:03] I you know, when you come in as a leader, as someone who's supposed to influence change, you have to really observe what's happening. [43:10] and figure out what are the challenges, like anecdotally, and what are you observing? And then I figure out where I can plug in and what I can do to make the teams better. And what I tend to find is-- [43:22] you know, moving from the right side to the left side of this, like action plans are easier to institute. [43:27] So if I see a team that's struggling to execute, I wonder, do they have the right type of PRD framework? [43:33] Are they communicating well? Do they have the right type of team meeting structures? But those are like kind of lower hanging fruit. It's a lot harder to change vision and skills [43:42] it will come in time. [43:43] But what I also have done over the course of my career is I've

[43:46] built this deck [43:50] that comes with me of like different skills that matter. [43:54] Like what are different skills? What are different frameworks? How do we think about it? [43:59] to help up-level teams fast. [44:01] Because I find that sometimes you walk in and not everyone is grounded in the same mental models of concepts. [44:07] You know, one example of this I talk about a lot is [44:10] you know, Instagram had just this fantastic culture of [44:14] thinking about how do we ship high, high quality products. [44:17] What does product craft mean? [44:20] And so that is something that, you know, I came, I came to YouTube and my team was particular. They didn't have a mental model for product brand. [44:27] So I found myself talking into an echo chamber in some way. [44:31] A deck. [44:32] that showed, OK, here's how I think about product craft. Here's a framework for it. Here's how I think about all of these different things. And so now we have a shared language, shared communication, [44:42] for an repository of skills that we're going to build. [44:46] Thank you. [44:47] So I'm looking at this image that [44:48] will have up. So in this case, [44:50] Their skills weren't necessarily there. [44:52] which in this framework leads to anxiety. And so what I'm hearing is essentially you come into a team, [44:58] You're like, what am I feeling? Is it confusion? [45:00] frustration, [45:01] the wrong thing and that kind of tells you which of these buckets to spend time on. [45:05] That's right. [45:06] And then within those buckets, you've got to figure out what are the right [45:09] levers that you need to pull, what are the things that are missing, how do you really focus and where do you spend your time? So interesting.

[45:17] I love that this image is just like this very grainy... So old. ...screenshot from some old McKinsey deck or something. Yeah, it's like the 2006 PC or something. I love it. So you have this thing, you have this deck. [45:28] You just come in with all these tools in your tool belt. [45:31] Is there anything else in that? [45:32] list of things that you bring with you to help change culture and [45:36] Health teams. [45:38] - I think one thing I think about a bunch is, I come from, my background's a little bit different than a traditional tech executive. [45:47] There was actually three phases to my career. I was in education for six years. [45:51] Tartan, inner city, D.C. [45:53] and then was a dean of a boarding school in Switzerland, which is like a little bit of a plot twist. [45:57] I went to business school and I worked on Wall Street for a bit. And then I left my job on Wall Street, quit and started a startup. [46:03] Startup was a glorious failure, as many startups are. But it was like a very nontraditional path towards tech. [46:10] And I think a lot of my time in these other industries actually shapes the way I think about products and product management. [46:17] and actually changing teams and building teams. [46:21] And what I mean by that is like, you know, I'm... [46:24] you know, [46:25] I think there's a lot of [46:27] similarities between education and product management, believe it or not. [46:30] You know, when I was interviewing for my first set of jobs in tech, recruiters would say to me, how does your background relate? Like, I don't really see it. [46:39] And I would tell them, you know, when I was in education, [46:43] I would walk into a classroom of 24 seven year olds [46:46] And these kids owed me nothing.

[46:49] And I needed the only way I could be successful or impactful is I needed to be able to [46:55] be a strong communicator. [46:56] I needed to be able to have a clear vision of what was going on. I needed to be able to influence them. [47:02] in a believable way such that they would get on board with what needed to be done for 270 days of the year. [47:08] It's like, you know, when you walk into a room as a product manager, engineers, designers, researchers go to market. Nobody owes you anything. [47:15] And the only thing that you're going to do in order to be successful is you need to be a strong communicator. You have a clear vision of what's going on. You need to be able to influence them to do the things together of what matters, right? It's very similar skill set, just different domain expertise. [47:29] And so because of that, right, like, [47:31] I really adopted a mindset of like, how do I coach my teams? How do I enable them? Because it's really about the sum of the parts. [47:39] versus me being a top-down leader saying, you have to do this and you have to do that. [47:44] And so I think a lot about that in both approach [47:47] and the processes that we create. And to give you a couple of examples of this, [47:55] There's two things I want to call out one is [47:58] You know, there's actually this education [48:00] framework. It's called Bloom's Taxonomy. [48:03] I think it's changed over time, but when I learned it, it was basically a pyramid. [48:08] And the pyramid was Bloom text only describes what's the different levels or order of critical thinking you need in order to be a master of something. [48:17] and at the bottom of the period was knowledge, and then it's comprehension.

[48:21] and an application analysis [48:24] synthesis and evaluation. [48:26] And so going up the pyramid means it's higher order thinking. [48:30] And I think about that a lot in trying to understand, where are my teams struggling? Do they have the core knowledge that they need? [48:36] Do they understand it, but they can't apply it? Are they applying it, but they're not able to analyze it across different business segments? And you can use this framework. [48:46] not only with the ICPMs and the teams themselves, but also for your managers. [48:50] What is breaking down? [48:52] And then using that and attaching that to like the skills that we want to build to figure out how do you fill in those gaps? [48:58] And it really is a grounding for me, a grounding mental model. [49:02] for how do you build teams that are actually effective, how do you meet people where they are, as opposed to just saying, hey, you need to figure this out. [49:09] I find that too often in tech and also in product, [49:12] People ask. [49:14] to figure things out, but not given the support to get there. And there's no way to really connect the dots for them. Amazing. I'm pulling up the Bloom's taxonomy. [49:24] So essentially, if you see a PM, [49:27] struggling. [49:28] What you try to do is figure out which of these things do they not have. How are you not supporting them? How can you support them better? So it could be they don't have the skills, they don't have the understanding. [49:37] What are some of the other things that often you find that maybe hinder a product manager's success? [49:42] A lot of times it's they might have the understanding, but they haven't had a chance to apply it to a variety of different scenarios or haven't seen it applied to multiple scenarios.

[49:52] So oftentimes you might understand a concept like machine learning. [49:55] but you haven't actually worked on it [49:58] in multiple against multiple scenarios. So you maybe have one way of doing it that doesn't make sense. And you need to like have two or three. You don't even know that two or three ways of doing this. [50:08] Right. Like that is like often a common failure point. Right. Or it's maybe you know how to apply it, but you can't synthesize [50:14] why this thing that you're doing actually matters to the business context. Right. Like oftentimes that becomes a challenge with managers. It's like they know what to do, but they don't understand how to tie it back to the business context. [50:27] and overall strategy needs. [50:29] And so where to prioritize. Right. And so what I find is that like [50:33] You know, we're trying to manage managers. You're really trying to lift the top of the pyramid. [50:38] You are responsible, managers responsible for basically owning that pyramid. [50:42] for all of the areas that they operate, but they need to be able to live at the top of the pyramid [50:47] across all. They need to be able to synthesize and evaluate what's happening [50:51] for each product team that they own in order to kind of make the bigger picture connections. [50:56] By the way, I love your metaphor of product managers are like this group of seven-year-olds where you have to learn how to manage. [51:01] uh influence communicate [51:04] That's great. Yeah. I mean, it is not like a group of seven year olds, but I think it's actually, it's the same skills, same skills. I mean, I think it's true. Like I found it actually significantly harder for me to get 24, seven year olds to believe in what I was doing than to walk into a room with Kevin system and my creator, [51:20] and explain to them what's the strategy for growing the next hundred million users on

[51:24] or on Instagram. These are very logical adults who can reason with you. You've got a classroom of kids that's not quite the same. So those skills are really critical. I love that. I definitely want to ask about your Instagram days. [51:39] Is there anything else in this bucket of wisdom of... [51:41] You kind of talked about [51:43] Things you've learned about how to manage product managers and managers and managers. Is there anything else there that might be helpful to folks that you've learned? [51:50] one thing [51:51] I also think about a lot, and I don't know if this is just a me thing, but I think about PM as a [51:57] Team Sport. [51:59] Right? It's leading product teams is really about [52:04] Being the coach and helping other people to like kind of see what their role is on the team and maximize them and [52:10] People talk a lot about product as a, you know, you're the CEO and I don't actually fully believe that analogy. [52:16] you know, [52:17] If you think about it as a team sport, [52:19] there's a few things that kind of shake out. One is not everyone's going to be a star player. [52:24] But not everyone needs to be a LeBron James or Kobe Bryant. You need role players. You need really strong role players. You need people who feel valued in their role. [52:33] And you need to understand how to groom those role players, and how to make sure that they have the right seat at the table in the right place. [52:40] And so I think it's, yes, you're the conductor of the orchestra, but you're really more than that. You're really like the coach of the team. [52:46] Another thing I think about a lot is, you know, I have a good friend who's a college basketball coach, and he taught me about kind of this idea of your coaching tree.

[52:57] And this is a really important concept, especially in college basketball. [53:01] You, as a head coach, you take a lot of pride in who were your assistants, who was your first assistant, second assistant, third assistant, and where do they go on to be head coaches? [53:12] And what legacy do they have because you were able to install a bit in them? [53:15] And so the coaching tree of Mike Krzyzewski, the coaching tree of John Calipari, these like esteemed coaches, not only because of what they've done, but because of the tree that they've built. [53:26] And I like to think [53:28] about this as well, because I think it's really important for product leaders to think about what is their leadership tree. [53:33] Like, who have you... [53:35] help to build up and help to grow and help to get to their next role. Right. And so I think about this a lot. You know, I have people who I've worked with who are running growth astray for the CPO chief or now running stories at Instagram that were on my team in the earlier days. [53:52] And [53:53] And their success is my success, and I'm proud for them, right? I'm happy for them. And I think it kind of like-- [53:59] reinforces this mentality. It's your responsibility to coach people up to greatness. [54:04] So what I'm hearing is you've put a lot of value [54:07] in your team on them coaching folks. [54:10] whether they're managers or even ICs. [54:13] is helping them understand that it's important to coach folks on their team and help them develop, that it's part of your job, essentially. [54:21] That's part of your job. I mean, you're trying to build a repository of skills and repository of knowledge and of team velocity. And the only way you can do that is everyone is like,

[54:31] So rising tide has to lift all boats. [54:33] Amazing. And interestingly, that probably teaches you how to do your job better because in teaching, you actually learn things a lot better. That's right. [54:40] It forces you to figure out how can you get things off of your plate. [54:44] so that you can go work on bigger things. Yeah. And how good does it feel when folks that you used to manage go on to do? [54:49] Bigger and better things. [54:51] It's great. [54:52] I want to talk a little bit about growth within YouTube. I heard that you haven't been there for that long, and apparently you've already... [54:59] 2x or 3x or more, something important within YouTube. I don't know exactly the details. [55:05] And people are very impressed with the impact you've already had. [55:08] And apparently a lot of the success there and other places is how you think about growth. [55:13] through flywheels. [55:14] You always look for the flywheel that helps [55:16] can you just talk about either at YouTube [55:19] Ideally, YouTube, whatever you can share, because that's pretty impressive, especially for a company at that scale that you're making so much impact. [55:25] or other places, just how you think about [55:27] White wheels and growth. [55:28] Well, it's very generous of you. I wouldn't say, look, we've done some really [55:32] good work, some good work so far on YouTube. And it's been a journey. I think [55:38] I do I do think a lot about fly roles. I think it's actually [55:43] A lot of growth is [55:46] really understanding [55:47] what is the value profit, all of the different points of the experience, especially if you have a multi-sided marketplace, like a multi-sided marketplace. [55:56] For YouTube, it's like, you know, the creator. [55:58] what the creator is trying to achieve, both from an engagement perspective, but also for me from a monetization perspective.

[56:05] And then for the viewer or the purchaser, like where is the heck? Like what are they trying to do? [56:10] and where are they missing the opportunity? [56:13] I can't really go into specifics a bunch with YouTube, but what I would say is this is that like, [56:18] One thing that I always do when I come in is I [56:21] trying to push my team to really dog food products. [56:24] in their [56:26] adjacent user state, if you will. [56:28] right um and what i mean by that is [56:31] Often a product that you and I use, that we've been using for years, [56:36] isn't actually the product that we're building for other people. [56:42] like a power user. [56:43] who's using a product [56:45] has so much history and so much informed knowledge on how that product [56:50] on for that product to actually create a great experience for you that if you were to go and create a new gmail account and look at youtube today [56:58] I guarantee you it's a completely different and significantly worse experience. [57:04] There's a lot of [57:06] obvious opportunities missed, especially with what is the flywheel and why things are working or not working, if you don't actually go and use things in a new state. [57:14] Right. And an example of this, I want to talk about YouTube is [57:18] A lot of the YouTube graph for searching stuff is based on what have you watched in the past? [57:22] So if you go and search and you search history, [57:24] it's going to be about like, well, what have you watched in the past or what have you searched for? [57:29] And what can we show you that's going to be what we can better predict? But if you don't have a lot of searches [57:34] then they're not going to do as good of a job. So that's not exact direct translation for what we were doing. But for me, I work on freedom monetization, and there are really important flywheels around like,

[57:45] What does it take for a creator to make content that can help them to monetize? [57:50] How do we get that content to people and where and to what extent are we getting that content to people? And how do we make sure that people feel good about what they're receiving, the people who are paying? [57:59] And all of those flywheels have to work. And so part of what I've been able to do is really think about how do we connect the dots in a story [58:07] that the teams can uniquely understand, [58:10] can help them to lean in even more. [58:13] and have clarity and purpose of work. Sometimes what's super important about the fly was actually is enabling for your teams to know what to work on, what not to work on. [58:22] Right. And then it also helps us to understand [58:24] What do we know and what do we not know about creators and viewers and monetization operations so that we can do the understand work? [58:31] to improve the velocity and to prove the impact. [58:34] and that's really where we've been focused and so we did i did a lot of this in youtube but you know also did this in instacart [58:40] really thinking about when I joined Instacart. [58:43] One of the questions I had was like, [58:45] How are people in their daily lives? How is it daily life actually reflected in the purchase experience? Are we making it easy for them because? [58:52] When you buy groceries, you're not going grocery shopping because you want beautiful ingredients in your fridge. [58:59] You're going grocery shopping because you actually have a meal to put together. [59:02] So like, are we actually reflecting the real job to be done, which is I want to buy tacos. [59:08] I want to make Taco Tuesday. Can we make it easier for people to find ingredients in tacos? [59:12] as opposed to having them search for tortillas and tomatoes and avocado.

[59:18] So it's really thinking about what is the job and then what is the fly worth to make that happen and how we make this come to life. [59:25] There's so many things that I want to follow up on here. [59:27] First of all, I realized that all of the things you worked on [59:31] I'm a daily active user of or weekly active user. Twitter, YouTube, Instagram, Instacart about weekly active user. [59:39] I think that might be... Yeah. Wow. Nice job. You got me in the flywheels of all your flywheels. I was not intentional, but I'm glad my efforts had improved a draw life just a little bit. Nice work. Ideally. [59:49] OK, so a couple of takeaways here. One is you think about-- [59:53] the value prop at every interaction [59:56] of both sides if it's a multi-sided marketplace. [59:59] Or if it's just like, why would I be doing this? So it's like, why would I send an invite to my friend? Why would I share this photo? Why would I... [1:00:05] open up Instacart. [1:00:07] And then you think about the jobs to be done during the day of a potential user and how can we [1:00:12] flow into that versus not connect to their actual day-to-day experience. And is the product actually working? As I'm doing this, I'm looking and saying, "Is the product actually set up to deliver these things?" [1:00:23] Like, do we do we actually see it work? And I think there's a lot of assumptions that the product works. [1:00:28] But a lot of times teams will surprisingly build [1:00:31] the pieces of a flow, but not actually build the experience, the output. Like they don't design it in a way where you're getting the real output that you need. [1:00:39] That's such an easy to miss point you're making here, which is you just think about, hey, I have Taco Tuesday. I'm actually as a product manager on Instagram. I'm going to open up the app. [1:00:48] and use it in this use case and see how it goes.

[1:00:50] That's right. [1:00:51] And most people don't do that, is what you realize. - People don't do that. Another example of this, right? Like this was years ago, but at Instacart, [1:00:59] This got made it really hard to reorder stuff. [1:01:01] Super hard to reorder. [1:01:03] And it was shocking to me because [1:01:06] When I thought about it, when I go to the grocery store, 90% of the time I'm getting the same stuff. [1:01:11] Right. It's like, you know, maybe not any trip, but over the course of a month or two, [1:01:15] you've got a list of things that are part of your stable and then occasionally you know like [1:01:20] Holiday time, I want some peppermint bark. [1:01:23] chocolate or something, you know, but like it's not like it's like there's like random snacks you're gonna throw in. But like there's and so when we looked at the data, it turns out, [1:01:31] Like after five times when you go to Instacart, 90% of your order is the same. [1:01:36] Right. But like when you wanted to reorder, at least back then, [1:01:40] you couldn't go and reorder [1:01:43] easily you had to dig and find it like seven or eight clicks and when you did reorder you had to reorder the whole thing [1:01:48] You couldn't take pieces. You couldn't take, I want to take the milk and the blueberries. [1:01:52] Right. And so [1:01:54] You know, you think about growth, you think about what does it mean to grow an Instacart? What does it mean to actually drive better retention? [1:02:00] Well, it's actually really important to make it easy for people to make the next order. [1:02:05] And so the product wasn't really built for that, though people had the best intentions in mind. The adjacent order. [1:02:12] Exactly. I imagine that was maybe one of the biggest growth wins in Instacart history is just the reorder the same thing because I do that all the time. [1:02:20] Yeah, you would be surprised. This episode is brought to you by Vanta. When it comes to ensuring your company has top-notch security practices, things get complicated fast. Now you can assess risk, secure the trust of your customers, and automate compliance for SOC 2, ISO 27001, HIPAA, and more with a single platform, Vanta.

[1:02:50] tracking risk. Plus, you can save hours by completing security questionnaires with Vanta AI. Join thousands of global companies that use Vanta to automate evidence collection, unify risk management, and streamline security reviews. Get $1,000 off Vanta when you go to vanta.com slash Lenny. That's V-A-N-T-A dot com slash Lenny. [1:03:14] You mentioned this concept of adjacent user. [1:03:16] kind of adjacently, but I think it's worth spending a little more time here. It's like this term you [1:03:21] popularized and you explained it somewhat, but maybe explain a little bit more because I think it's really powerful for people and how you think about growth. [1:03:28] This was something that came out of my time at Instagram. [1:03:32] It was our cover framework that we came up with because Instagram really was at that time. You know, we grew so fast and [1:03:41] that the people who were using Instagram in February were completely different than people who were using it the following October and then the next January. [1:03:49] I think when I joined, we were at like 445 million monthly actives [1:03:54] January 2016, at the end of that year, we were at 636 million. [1:03:58] Like 47% that year. [1:04:00] And so when we talk to users, [1:04:03] first half of 2016, [1:04:06] You know, when we talk to women in the US in their 30s, they're like, Robert, I have an Instagram. I have a Facebook account. Literally people said that. Right. That was that long ago. [1:04:14] And then like a year later, it's like, of course I use Instagram. Like Instagram is my everything. [1:04:19] And the world changed so fast. And so when you're in a hyper growth product,

[1:04:23] It's really, really important to understand who your users are today [1:04:28] and the persona of the user, what motivates them, why they're using it. But then also to understand who is the next user, who is the user who [1:04:37] could be using this product, but for some reason it doesn't work for them. [1:04:40] Right. And understanding who that adjacent user is and when you're actually starting to see that adjacent user adoptive product. [1:04:48] And one of the ways you start to see the adjacent users starting to [1:04:52] Adopt the product especially with the data as you start seeing cohort cohort curves decline. I [1:04:58] Right. You start seeing the people who sign up today. [1:05:02] three or six months from now, they're signing up and they're doing a worse job. Nothing's changing the product. [1:05:08] but just the understanding of how the product should work. [1:05:11] is different. [1:05:12] They might be less tech savvy right on the scale of an earlier doctor versus like late majority. They might be closer to the late majority. And so we saw this at Instagram, right? Like we would always be working on registration flow. [1:05:24] And at one point we were converting at some insanely high percent. [1:05:28] And then three months later, it would go down by 15%. [1:05:31] Not because anything was broken, but just because we've broken into new markets. You're bringing on people in India or you're bringing on people in the Philippines. [1:05:38] and their understanding of how it works, the phone they use, etc. are all different. So really the core of their adjacent users is a few things. [1:05:46] You have to understand who is using your product today and why. [1:05:49] And [1:05:50] When you're growing at some really strong pace, 30%, 40%, 50% or more per year, you've got to be on top of,

[1:05:58] who is your next, who you believe the next user is and why. [1:06:02] And then you also have to be the adjacent. You have to use the product life-derminancy how it's working. [1:06:07] what's broken with it. [1:06:08] Right. And so at an Instacart, the adjacent user, the original user might have been like, [1:06:14] uh you know an office admin who is going to buy this thing every week because of during happy hours and team staff but then the next adjacent user might have been the mom of three or four or the dad of three or four who's home with the kids and they are you know they need to depend on instacart right versus you know like later on it might be they're a single person in new york who like does this out of convenience right but like what they're optimizing for how they use the product all changes [1:06:42] And the functionality and the abilities [1:06:44] fundamentally different. So you have to be them. You have to watch how they use the product. You have to talk to them. [1:06:50] And you have to visit them and literally see what they're doing in real time. [1:06:54] in order to make sure that you're enabling the right jobs. [1:06:57] I love just the visualization of the adjacent user. Basically, your growth is going to come from non-existing users. It's the users right outside that circle [1:07:05] where your users are today, and you need to think about what do they need. [1:07:08] that existing users maybe don't need. [1:07:10] You said that this is most powerful for hyper-growth companies. Is this something you think all companies should be thinking about, or is it a lot less... [1:07:17] important if you're not a hyper-growth [1:07:20] business. [1:07:21] So it's a great question. I think it's essential. It's mandatory for hypergrowth. I think it's very helpful. [1:07:27] if you have a product or company that

[1:07:31] is not growing what you want it to be. And so you're focused on [1:07:36] capturing more share of wallet. [1:07:38] instead of expanding your audience. [1:07:40] right and so sometimes you can imagine like a cosmetics company for example right that's like digitally enabled [1:07:47] you can imagine that they've hit a ceiling in terms of the growth of their users. [1:07:51] And they're really just trying to get people to buy more product. [1:07:55] but maybe, [1:07:57] they're missing like the people who want to use their product are missing [1:08:00] something from your current product, right? Maybe you're missing different skin tone shades, right? Or maybe you're missing certain types of tools. So really talking to who is just outside of your current user base, who's coming to your product, and looking around and not buying, understanding what are their needs, and figuring out how do you enable them? How do you build the right experience for them? [1:08:18] in order to become adopters of your product. [1:08:20] So what I'm hearing is spend some understand work to figure out who your potential users are, [1:08:25] and then use the product as them and see... [1:08:29] what is missing. I imagine user research goes into this. You're not actually going to understand necessarily all the things they need. [1:08:35] So it's probably fine folks in that cohort and see how they use the product. [1:08:40] Awesome. [1:08:41] The question I had on my mind as you were talking, you've come into so many companies [1:08:45] and [1:08:46] help them with growth. [1:08:47] Where do you find most of the opportunities often lie? Is it like onboarding? [1:08:51] activation, is there a trend like that? You're like, "Here's probably there's going to be opportunity here," or is it super mixed bag? [1:08:58] Usually it's somewhere in the

[1:09:03] onboarding to like habit building experience. [1:09:07] Great, great. [1:09:08] What does it take for someone to actually understand the value that first moment, that first moment in the product? [1:09:16] And a lot of teams, it's shocking how many teams don't really understand what that moment is for them. [1:09:22] And then also, how do you get them to build a habit around the product? [1:09:25] Thank you. [1:09:25] Oftentimes a lot of people equate growth. [1:09:28] to top of funnel, and that is also critical. [1:09:31] I think-- [1:09:32] having the right top of funnel motion, [1:09:35] is really critical in building on that. So I think there's [1:09:39] One part of it's like once you have the right type of funnel motion to get people to come in, [1:09:45] How do you help them make sure they're finding value and they're building habit and they're retaining? [1:09:49] Because that's the thing that helps you compound over time. [1:09:52] Like if you're bringing a lot of people up, but they're not staying around. [1:09:55] Then you just have a leaky bucket, and it doesn't matter how big a top funnel is. [1:09:59] So making sure that that first month and two month, three month experience is great. [1:10:03] And then the other part is really figuring out, how do you build compounding growth loops [1:10:08] where it's not just [1:10:10] one way of acquiring people, but you're building two and three and four ways [1:10:15] that layer on to each other, that help you to really supercharge your [1:10:20] your engine of, like, acquiring people. [1:10:21] So, on Instagram, [1:10:24] if you kind of look at where Instagram is and what [1:10:27] how I think Aaron Groom, there's a lot that goes into it, but if you [1:10:31] actually unpack the top of funnel for what worked at Instagram,

[1:10:35] There's certainly a component of it, which was our core component, which was invitations. [1:10:39] where people inviting you and making sure that those invitations work and they work well, and that people's friends are coming on and get notified. [1:10:48] But, you know, another part that goes unspoken, still critical to this day was like the celebrity partnerships. [1:10:54] was critical because [1:10:56] Basically, you know, they had this wonderful partnerships team. [1:11:01] that basically took Instagram and taught celebrities how to use it, [1:11:06] how to make it work for them, how to tell their own story and be their own brand. And that was a critical growth funnel because, [1:11:12] With that, [1:11:13] you had the ability for them to create these celebrities and celebrity creators, [1:11:19] to set the norm for how the platform gets used. [1:11:22] They also were getting picked up by the news and the media. [1:11:26] Right. For, you know, all the stuff that's happened in the celebrity world, which then added on to this other growth level, which had which was SEO. [1:11:35] And so every time a news article came out, they would link to the creators [1:11:41] like or the celebrity's Instagram account or that particular post [1:11:45] Right. And so you have this whole SEO engine that worked. And the SEO engine was because you have both Web, which we launched at Instagram, [1:11:53] which created the canonical SEO tables, and then you had [1:11:57] all of these inbound links from these celebrity sites and news media. But then what you also had is you had embeds for Instagram in all of these different sites, right? Like a news article or whatever, the posts, you know, Lenny's podcast, Instagram account, those embeds help with SEO juice. And so you have not just the invites, but now you have the celebrities. Now you have the, you know, the SEO component.

[1:12:21] And then we would do a bunch of paid media on top of that. [1:12:24] I'm using a lot of those signals. [1:12:26] And so we have and then we have our own content. [1:12:28] And so you have all of these different kind of growth engines compounding each other. [1:12:33] So every time the invitations got better, every time I got more celebrity, every time SEO gets better. [1:12:38] It's like magnifying the top funnel. [1:12:40] And at the bottom of the funnel, like or mid funnel, we're making sure that people are retaining and getting value and staying around over the long term. [1:12:47] That is so interesting. I had no idea this was such a core part of the early... [1:12:51] growth strategy. Everyone's always talking about morality and [1:12:53] or to mouth all these things and you're saying partnerships. [1:12:56] was a key part of the early growth strategy. [1:12:58] That's so funny because it's never something people talk about in growth is partnerships. So he's like, oh, there's a couple of companies that really had success from partnerships and [1:13:07] Yeah. [1:13:08] How early was this a big part of the strategy? How early in the history? I mean, the partnerships team was there before I joined. It was a very savvy and astute thing that Kevin and Mikey set up. [1:13:18] They really kind of [1:13:20] they made like the partnership team drove a lot of the word of mouth around instagram [1:13:25] But it was the partnerships work in combination with the product work that actually helped to [1:13:30] allow a lot of that to lane, right? Like it was, you know, [1:13:34] our ability to think outside the box and understand that we needed to have a web presence [1:13:38] because it was critical for international growth. [1:13:40] and it also helped with SEO. [1:13:42] And so the idea of launching run up really drove and actually increased [1:13:47] Instagram growth by 10% the minute we launched it. [1:13:50] And it was something that, you know, George Wang, who was there before me, had the idea around

[1:13:55] It wasn't really something that Kevin and Mikey believed in initially, but we had to prove to them. [1:14:00] that it was really impactful. And once it launched, then they understood why it was impactful and all of the effect of it. [1:14:07] The web was really critical for driving so much SEO. [1:14:11] which helped support a lot of the celebrity work in the partnerships work. [1:14:15] Because now every time a creator or a celebrity [1:14:19] did something on Instagram. [1:14:21] Every news media article picked it up and it just helped to drive searches and Google searches, which helps to make Instagram part of the cultural zeitgeist. This is so interesting because everyone imagines Instagram reality word of mouth. [1:14:33] SEO and partnerships sounds like a core part of Instagram's early growth, which I don't think anyone ever talks about. [1:14:39] Wow. [1:14:40] I mean, it was part of Facebook's growth too, right? But I think a little bit differently, it's not as much partnerships. But when you Google someone now, you Google someone's name, oftentimes Instagram is one of the first five things that comes up for the average person. - Yeah. - Right now, right inside. - LinkedIn, Instagram, Twitter. [1:14:54] It's like, [1:14:55] Do you ever see the video of Alex Zhu talking about TikTok's growth strategy? He had this metaphor where he was trying to grow off of Instagram. His metaphor was, "Instagram is like, you're in Europe. [1:15:08] and you're killing it. If you're in Europe and you're king, [1:15:10] You don't ever want to go to America. There's no reason for you to give it up. And America is TikTok in this case. [1:15:16] He's like, how do we convince people to come to America and try moving everything there? He's like, we need to go after the people that are not doing well in Europe who want to be the king. [1:15:24] And we're going to help you become that king in America or in the president.

[1:15:28] So he used the adjacent user theory on us, basically, is what you're saying. That's right. You did. [1:15:35] Drink up your shake. Yeah. Yeah. So I think, and I think the reason that strategy worked differently is there was already a place. So, [1:15:44] they couldn't execute what you did because there's already [1:15:47] Celebrities are already there and it's not like come here. There's no point. I already have a huge following and [1:15:51] So you make your own celebrities. Make your own celebrities, exactly. And they went after the B-list, C-list people. [1:15:57] Right. So it's interesting that people look at Instagram like, "Cool, we're going to do partnerships, [1:16:02] But I think it's important to realize things change when the market is a different dynamic. [1:16:06] You can't just do the same thing. It won't work as well. [1:16:08] Is there anything else from the Instagram early days that maybe people wouldn't know or would be interesting to share because you were there quite early and now it's maybe the most thriving social network in the world? [1:16:17] One thing that's actually, I think, interesting, is an interesting story that doesn't get talked about a lot is, [1:16:23] I think the early Instagram was built [1:16:27] where it was built in a way where [1:16:30] Every follow was created equal. [1:16:33] And what I mean by that is, [1:16:35] You know, if I followed you, Lenny, or if I followed Kim Kardashian, or if I followed [1:16:41] Selena Gomez, "All follows were treated like, [1:16:44] you know, [1:16:45] they were equally important. [1:16:47] And this is actually a really important fact. It was a really important factor [1:16:52] for a couple of reasons. [1:16:54] Early celebrities were adopted in Stavam. [1:16:57] obviously benefited from that because you know a celebrity is going to get a follow before an average person does and

[1:17:03] It doesn't mean that like, you know, their follows weren't meaningful. It's just that like, [1:17:07] you know, when you have a machine learning that's just optimized for, [1:17:11] a click through or follow, then like [1:17:13] you know, that matters. But what ended up happening is [1:17:18] We ended up looking at the data [1:17:21] and i can't take credit for this my um colleague at the time rob andrews had identified this he was the head of growth marketing is a pure mind that you know basically this is 2016-ish days [1:17:33] The average person would come on Instagram and retain, but then [1:17:37] leave after seven, eight, nine months. [1:17:40] We saw like we see a flattening and a retention curve. Then we would see it like dip again, which was very. [1:17:46] I'm like, "Why is this happening?" [1:17:48] And it turned out that [1:17:50] What was happening anecdotally is that people were repping on Instagram [1:17:55] following a bunch of people, [1:17:57] following a lot of celebrities actually, [1:17:59] because the celebrities are being shown. And then when they actually went to make their first post a few months later, [1:18:04] None of their friends were following them. [1:18:06] And so there was like posting into an echo chamber. [1:18:09] And anecdotally, people would stop using the product because they felt bad. [1:18:13] we hypothesize. No one was [1:18:16] liking or following or commenting on the post. [1:18:20] And so we had to do this thing, we called it the connections pivot. [1:18:23] right in like 2016, 2017. [1:18:26] where we [1:18:27] had to convince Kevin and Mikey [1:18:29] that it was actually not the right thing to do to prioritize celebrities to everybody.

[1:18:34] because we're basically biting on, cut was a bite your nose to spite your face, whatever that is. Like, you know, because [1:18:41] the regular person wasn't having a great experience. [1:18:44] And so it doesn't mean that we shouldn't recommend celebrities, but you have to recommend celebrities to people who are already on there. They already have the graph. [1:18:51] And the most important thing to do was actually get regular human to human connections. [1:18:56] in the first, whatever, when people first sign up. So that when you, Lenny, actually go make your first post, [1:19:02] Wake up. [1:19:03] your friends would see it and you would be validated and you would feel like okay this is a place for me [1:19:08] I have a community here. [1:19:10] And so that connection pivot was critical. It changed, literally angle changed the retention on Instagram. [1:19:16] And so if you like Google, Instagram, growth, there was like a tech crunch article about [1:19:21] Instagram's growth that, you know, I think 2017, 2018, when we're going 40, 50% year over year, [1:19:26] And obviously, there was a lot that went into that skyrocket. [1:19:30] People think that stories was the sole reason why we grew and story brought us a lot of people. [1:19:35] But I mean, we literally our retention doubled over the course of like a year and a half. [1:19:40] I mean, if you could imagine, imagine if like, you know, you're you're [1:19:43] your bank account, like the interest rate doubles every month. You know what I mean? [1:19:47] It was incredible. [1:19:48] This just the shift in making sure that people got connected with their friends early on. [1:19:53] change the way that people perceive the value. [1:19:57] And so a lot of the top of funnel work that we did, a lot of the activation work that we did really like. [1:20:03] paid off in space, ultimately. Wow.

[1:20:06] I'm curious, what's the most impactful thing you've [1:20:08] Shipped. [1:20:09] in terms of impact and experiment launch. [1:20:13] at Instagram, [1:20:15] Um, [1:20:17] You know, we saw this big problem where people were logging out and not being able to get back into their account. It's crazy. [1:20:24] like hundreds of thousands of people a day could not get back into the account. [1:20:27] And it was because, and when I say could not get back into the account, I mean, [1:20:31] They literally... [1:20:33] um, [1:20:34] would try and 28 days later, we would never see them again. [1:20:39] and we thought it was a problem because they didn't remember [1:20:43] where the email was, which email they signed up with, or what their actual handle was, or what their password was. [1:20:51] And so this was we were losing 10, 12 million people a year. [1:20:55] from what we call account churn. [1:20:57] Account Access Churn. [1:20:59] So we worked on this problem and we had spent a lot of time in South Asia and Southeast Asia trying to understand how to grow the product. [1:21:07] And from that time, we realized that [1:21:11] It's important, actually, for people to be able to log out. [1:21:14] We don't want to restrict people's ability to lock out, because there are many people in the world who [1:21:20] don't want to... [1:21:21] use background data because they own prepaid phone plans. [1:21:25] or they don't have a lot of money and they want to share, [1:21:29] They're phone with a sibling. [1:21:31] and so [1:21:32] they log out. So it's really important. So we don't want to take away that behavior. But we did have to do a better job of helping them get back in.

[1:21:39] So we did two things. One was like, [1:21:42] There used to be a time, and you can probably find it on Google if you did like an archive search, [1:21:46] If you want to log into your [1:21:48] Instagram account, there would be a tab for your email and then there'll be a tag for like your handle. [1:21:53] right and like and we didn't really make it easy for you to log in with your phone number either and you have to get to the right tab and the right place to get the right thing [1:22:01] And so we did this omnibox experience. [1:22:03] where we're just like email, phone, hang up, put it all in one place. [1:22:06] Right. And we made it super simple for you. And the second time, if you tried it twice, [1:22:12] and it didn't work. We knew that we were on a trusted device. [1:22:17] we would just send you a text message and say, hey, are you trying to log in? [1:22:20] That solved like half the problem. [1:22:23] And then the other half, what we did-- and this still exists today, when it's on Instagram. When you log out, we say, hey, Lenny, it looks like you're going to log out. Do you want us to save your credentials on your device? [1:22:32] So you don't have to worry about your password. [1:22:35] Right. And then we later added an ability to have a passcode. And that's all basically the other half. [1:22:40] Right. And so just like being really thoughtful around, like, what is actually the core job that people are trying to do? Because a bad experience would have been like, hey, let's make it really hard for people to walk out. [1:22:49] Right. And then how we like a one or two experiments really help people to get back into [1:22:55] experience. And so what was interesting was that we were able to solve this and it's actually called drive public 15, 20 million [1:23:02] extra monthly active users a year, [1:23:05] What was also super interesting was that [1:23:08] it helped us to realize that getting people back into their account more

[1:23:14] drove more account [1:23:17] product like content creation on incident. [1:23:19] in ways that we didn't expect. [1:23:22] And so because we were getting people back into the account, people getting into the second account. [1:23:26] and getting into a third account. And they were going into their Finsta account, [1:23:30] And [1:23:31] creating content more. [1:23:33] And so that actually led to the creation of a multiple accounts team. [1:23:37] which is what has made it easy for you to navigate between accounts on Instagram today. [1:23:42] And so that was understand work actually coming to life. It was, hey, we're going to do the understand work to figure out how to solve account access issues. [1:23:51] And we're going to we're going to solve it and solving it. When we looked at the data, we're like, hey, [1:23:56] Why is all this content creation happening? That was not what we were expecting. [1:24:00] And where is it happening? [1:24:02] and it was happening from second and third accounts, [1:24:05] And so that made us realize that, oh, people were not only getting worked out of the first accounts, but they were actually creating a lot of them in the second and third accounts. [1:24:11] How can we make it easy for them to get and navigate between their first account? What it's like. [1:24:15] your account and a business account your account and a fence account or like a bakery account you know what i mean [1:24:20] And so that created a multiple accounts team, which I ended up owning. And that ended up even becoming like a bigger. [1:24:26] That's like an example of like understand work to solve a real problem, real solving, solving massive solution. [1:24:32] Massive impact. [1:24:33] creating new [1:24:35] like data that you didn't expect. They drive new understand work, which creates a whole new team. So like allow you to kind of move around your Instagram account more seamlessly. You can see this now, like when you go to post, you can decide, OK, I want to post as Lenny on a post as this other thing. Or when I want to write a story, you know, you can change anywhere. That was all an effort that came originally came out of this realization.

[1:24:54] I love how many of these... [1:24:56] huge impact stories are just like such straightforward, simple [1:25:00] things like just let people log in more easily. [1:25:03] help them log out more effectively. Uh, [1:25:06] Fine, show them their friends versus celebrities. Like, so many of these stories are often just, like, really... [1:25:11] simple ideas that lead to such profound change. [1:25:14] The logout example is interesting. At Airbnb, we found the logout was actually causing a lot of-- [1:25:19] churn also. Initially, they went with the simple solution of [1:25:23] just extend the session [1:25:25] logout length instead of like a week, make it two weeks or three weeks or four weeks. You pointed out with Instagram, people actually wanted to log out. So they're like, let's leave that alone and start with making it easier to log back in. [1:25:35] Sorry. [1:25:36] Amazing. Man, you're so full of amazing stories. I'm curious with Facebook, maybe just as I wrap up here, [1:25:42] Is there anything there? Because you were an early growth PM at Facebook. You worked a lot on [1:25:47] friends and helping people discover their friends. Is there anything there that would be interesting to share? In my time, I joined Facebook in 2014 and I was responsible for people recommendations globally. And at the time, Facebook was at scale and big and dominated North America. And so really the focus area at the time was South Asia and Southeast Asia. [1:26:08] And, you know, we did and we noticed a lot of really interesting things, especially in India. [1:26:14] Facebook seemed to be broken from a [1:26:16] people graph perspective, right? Like we just think that was the data was telling us like people didn't have as many friends in common. [1:26:23] Like in the US, it might have been-- I'm just making up a number. On average, we saw 22 friends in common when you make a friendship connection. But in India, it was like seven.

[1:26:31] Right. And then there was a lot of friending and messaging, friending and unfriending. We're like, what's going on here? [1:26:37] So we had some hypotheses from the data. We ran a bunch of tests. Nothing seemed to work. [1:26:43] and be a breakthrough. [1:26:45] Actually, at that time, I had to propose, and they agreed to do this. We needed to do an understanding. [1:26:51] So we were literally on the ground in India every three months [1:26:54] took a team of engineers with me. Like I'm talking like we're in Delhi and people's homes in Mumbai. [1:27:02] And we went to go kind of investigate what was going on, like watch people make friends, watch them, like use people you may know and really understand what was happening and see what was going on, because we felt like there was context of what we were missing. [1:27:15] There's a lot of things that we learned. It was mind-blowing. [1:27:18] But one of the most interesting things learning was that [1:27:21] we [1:27:22] watched people try to make friends. And we said, hey, wait. [1:27:26] That's the profile page. Why aren't you looking at it? [1:27:29] And they would say, [1:27:31] Wow. [1:27:32] That doesn't have any relevant information for me. So I need to go look at the pictures. [1:27:36] like well why doesn't the profile page have relevant information for you like that's [1:27:40] where [1:27:41] information supposed to be. And they're like, listen, this guy's named Amit Kumar. [1:27:46] I have 10 friends named Amit Kumar. Right? Like, what is this page going to tell me? [1:27:51] Also, and they would scroll down, they'd be like, [1:27:54] all of these fields are not relevant to me. Right. And so you look at Facebook at the time was very Western centric. [1:27:59] It's probably still is. [1:28:01] It is it was like name school you went to, like job title, right? Like affiliations, that kind of stuff. That's all Western paradigms. And like,

[1:28:10] Some of these people are like, my friends sell jeans at a market. [1:28:13] Like none of this is relevant. Like we didn't, you don't like, you don't even know the name of the school sometimes, right? It might be a number or something. [1:28:18] And so all of the descriptors that we take for granted were not relevant. [1:28:22] And the names were very common. This was really illuminating. So what they would do is go look at pictures and say, is this my friend's car? [1:28:28] Or is this my friend's like, can I see my friend's animal and what's available? [1:28:32] And it was interesting because, you know, the understand where at least the moral realization. So we went back to Menloberg. [1:28:38] with the data. [1:28:39] Turned out, we said, okay, let's look at the data in a different way. [1:28:43] What are the most common names on Facebook? [1:28:45] And we looked at the top 10 common names are Indian names. [1:28:48] And like the most common name was Amit Kumar. [1:28:51] And there was like 250,000 people a month [1:28:55] Who used Amit Kumar? [1:28:57] Like who are like real people? So imagine you're in Bangalore and you're trying to find your buddy named Mick Kamara. It's probably 5,000 that we could possibly recommend to you. [1:29:05] So it's super interesting. In fact, the cultural context was so different. So we had to get creative, find creative ways to figure out how to solve it. [1:29:11] The power of understand work shines again. [1:29:15] We have a recurring segment on this podcast called Fail Corner, where I ask guests to share a time they failed in their career. [1:29:22] and what they learned from that experience. Because a lot of people look at your career and your story and just like, "Holy shit, what a journey. I will never be able to achieve this." [1:29:31] Is there a time when things didn't go? [1:29:32] Well, and... [1:29:34] something that you took away from that experience? [1:29:36] Probably for me, I think my time at Instacart was probably not my best time.

[1:29:41] I think [1:29:43] Bye. [1:29:45] Went to Instacart with the vision of what I thought the product was [1:29:49] could be or should be. [1:29:52] And... [1:29:53] It was, I think, a big story that I deeply believed. [1:29:57] But I was not, I think, aware or in tune with what the DNA of the company was at the time. [1:30:03] which was it was like a [1:30:06] company that was really great at operations [1:30:09] And [1:30:10] because that was the core of what they did, and they were really building up a lot of their product [1:30:15] experience [1:30:17] and a lot of what [1:30:19] like the new function. And so I think [1:30:21] What they needed at the time, [1:30:23] was a much more tactical [1:30:26] in the weeds. [1:30:28] Kind of like [1:30:29] get your hands dirty, do, do, do, because they hadn't really seen that before. [1:30:34] And what I wanted to deliver in that experience was [1:30:38] more [1:30:39] kind of like building the right systems, the right people, the right processes, so that we can figure out how to institutionalize the work. [1:30:46] I think for me that was like probably my biggest oversight is that like I think [1:30:51] I think they needed more technical and I was like kind of more [1:30:55] I don't want to say in the clouds, but more think about where do we want to go with the experience that we need to build? [1:31:01] versus like let me get deep in like the funnel data. [1:31:04] It felt like it became loggerheads where I didn't feel like I was probably [1:31:09] either delivering what they wanted or being supported in the way that I wanted to be supported.

[1:31:13] Is the lesson there to-- [1:31:15] Spend more time on understand work before you take a job to make sure you're aligned with what they're looking for. [1:31:20] That is actually a lesson. And I think it's part of that is, [1:31:24] is doing that, one of my takeaways is that not only with the people there, but with the people who've left it too. The people who've left the company, [1:31:32] will give you a perspective that is raw and different and much more, I think, aligned with what you want to hear. [1:31:39] And it may or may not be the same, but I think that people and this isn't this isn't specific to the car. Like whenever you talk to people who are at the company, they always try to tell you the best version of the company. The people left will tell you what the worst version there or their version. And it's on you to like triangulate that information. But you need both sides. [1:31:57] Love that lesson. [1:31:58] Bengali, is there anything else that you want to share or leave listeners with before we get to our very exciting lightning round? [1:32:05] No, I think we've got a lot. We have. With that, we've reached our very exciting lightning round. I've got six questions for you. Are you ready? I am ready to go. First question: What are two or three books that you've recommended most to other people? [1:32:19] Few I've recommended most, some of them over my shoulder actually like Range by David Epstein. [1:32:24] Think of really critical PM work, really being able to be a master of a lot of different things, or somewhat deep in a lot of different things. [1:32:31] Deep Work, Calum Report, kind of clearing space to be focused and do the things, thinking about the most important work in a deep and like undisturbed way. And then Start at the End by Matt Waller, who's a kind of behavioral scientist, a buddy of mine, really thoughtful book, helps you kind of really think about what you're building, but in a way that's like more holistic.

[1:32:49] Do you have a favorite interview question that you like to ask candidates when you're interviewing them? [1:32:54] I've got a relatively new one that I love. It is so this works. [1:32:59] exceptionally well, I believe, for more senior hires, whether they're like group product managers, directors, or even like kind of [1:33:07] senior ICs. [1:33:09] I go through and I think about what are the four, usually five skills, [1:33:14] that are really critical. [1:33:16] for someone to have in this job. [1:33:18] and I think about them myself, [1:33:20] And I think about the archetypes of like, [1:33:23] who are like what, like what is the ratio on one of these? But then I ask them, I say, hey, take out a piece of paper or take out your laptop. I'm going to give you five skills. [1:33:32] And I want you to stack length and form it. [1:33:34] One to five. [1:33:35] Right. So one being the one the strongest that. [1:33:38] and 5 being the one you're relatively weak as a [1:33:41] So like it's way better than saying give me your strengths and weaknesses. [1:33:44] It also forces them to contextualize it [1:33:47] against the skill that you're looking for. [1:33:49] but it also helps you have a meaningful conversation [1:33:52] around how they think about themselves and their self-awareness and to what extent they're ranking these skills based on the context they're in. [1:34:00] Right. Versus their own ability. Sometimes you write something a five because like, [1:34:04] something you'll just recap because you haven't had to do because there's a strong function at that company. Right. And so when they start ranking people are like, [1:34:12] I've asked this probably four or five times on a recent role. I hire people like, oh, this is really hard, but really good question. And then you can learn. I usually will dig into what is like number two.

[1:34:23] And what's number four, number five, and why? [1:34:26] And it really helps me to one, calibrate, like, am I looking for the right person? And two, are their skills actually a match for what we need at this time? [1:34:33] I love when I hear a question I've never heard before. That is genius. [1:34:36] Next question, do you have a favorite product you've recently discovered that you really like? [1:34:41] I think one that I enjoy quite a lot, actually, it's really kind of [1:34:45] them [1:34:46] Really simple but amazing is this app called Flighty. [1:34:49] A little bit. That's spelled I-G-H-T-Y. [1:34:52] It is a travel app and it is manages all of your. [1:34:57] your flight itinerators and your friends itinerators, [1:35:01] But the reason why I love it is it goes two or three clicks deeper than the average travel app or like your airline app. [1:35:08] and it tells you when your inbound plane is running weight. [1:35:11] It helps you to understand where the gates are. My wife was just in Madrid last week, and she texted me saying, hey, I'm at the airport, but no gate is announced. I looked at flighty. I'm like, your gate is E67. It's actually been assigned, but it hasn't been announced. [1:35:24] I mean, this happens a lot here, right? Like you'll be there and they won't tell you the gate until like an hour beforehand. So it gives you so much information right before it's actually publicly announced. [1:35:33] which is really helpful, especially if you really need to be somewhere, because oftentimes I know if my plane's going to be late and if I need to rebook before anyone else does. [1:35:41] which helps you, gives you a better chance of getting somewhere. Or I will be able to get to a gate. I get to SFO, and it might not be on the board at SFO, but I know exactly where gating is. So we agreed out. Can't say enough about it. I'm downloading it right now.

[1:35:54] It's got a bazillion reviews, five stars. [1:35:57] Thank you for the recommendation. Incredibly useful. Of course. [1:36:01] I'm trying out a new question. [1:36:03] You've joined a bunch of different companies. [1:36:05] What's one thing that you've done in the first 90 days at the job that has made a big impact? [1:36:10] I go and I sit in team readings. [1:36:13] and see how they operate. I just listen. [1:36:16] Hi. [1:36:17] talk to not only the PMs, but the content designer and everyone and I. [1:36:23] Try to get to know them by both name and story. [1:36:27] And what I mean by that's like, [1:36:29] Oftentimes people execs come in and they're like, what are we doing? What are we prioritizing? Why does it matter? [1:36:34] And they don't really take the time to actually learn who the people are, what their story is, what they care about, what they're passionate about. [1:36:43] both professionally and personally, [1:36:45] and really try to understand [1:36:48] how the team is working from that person's perspective. [1:36:51] And what I find is actually when I do that, people are way more willing to hear [1:36:57] when it's time for me to share, [1:36:59] my thoughts. [1:37:00] that we're more invested. [1:37:02] Because they believe I'm invested in that. [1:37:05] I love that. [1:37:06] Two more questions. [1:37:08] Do you have a favorite life motto that you really like and find useful and often share with friends or [1:37:14] Family. Family. [1:37:14] in work or in life. [1:37:16] I'm not big on mottos. There's one that I've actually been repeating. I don't remember where I got it. It might have been like Adam Grant or someone on Instagram. [1:37:23] But it was...

[1:37:25] It's one I said it to a colleague of mine. He's like, wow, that really hit differently. [1:37:29] I don't know. [1:37:30] It's that like, I think it goes something like this, like, you know, [1:37:33] people and teams don't really reach us, they don't actually reach their goals, they fall to the level of their systems. [1:37:42] And that to me was really powerful because I struggle a lot with just being balanced in life, right? Like working out the way I want to and getting the right rest and making sure that I'm spending time in all the right places. [1:37:53] And that's really like, [1:37:54] It's a gold of mine, but the problem is that I don't-- my system for that specifically-- [1:38:00] I tend to like, you know, [1:38:01] Not. [1:38:02] I don't I don't I'm not rigorous with it is what I need to be. [1:38:06] But on the other hand, at work, I'm very rigorous in systems and processes. [1:38:11] So really like that, that saying to me really like kind of hits because it, [1:38:14] is both applicable to your life and how you want to live your life but also [1:38:18] How you want to run your teams? [1:38:19] Final question. You mentioned that you were a dean at a school, a boarding school. [1:38:24] Is there something that you take away from that experience that kind of sticks with you? I know you mentioned one thing about treatment. [1:38:30] PM's skills are similar to teacher skills. Is there anything else there? Actually, so I was a dean of a boarding school in Switzerland, which was like actually one of the most fun and interesting times of my life. It was an American school in Switzerland, just outside of Lugano, Switzerland, the Italian speaking section. [1:38:46] And I'd say, you know, [1:38:49] What was [1:38:50] I learned a lot about myself during the time. I learned a lot about just like relationships and people. I actually had a call with one of my old students this morning. She's like in her 30s now. This is like 20 years ago.

[1:39:01] So I still keep in touch with them. [1:39:03] um and i think that's you know the kind of the comment i made about knowing people by name and story matters a lot like you know [1:39:09] It says a lot that 20 years later, I'm still talking to some of these students, you know, who are like their own adults, run their own businesses, but also. [1:39:16] You know, I think the human [1:39:18] like human challenges and human connection, human problems, very universal. [1:39:23] This was a really interesting place. We had kids coming from a variety of different walks of life, some who lived in Azerbaijan or Kosovo, and the families felt like, [1:39:34] you know, they couldn't be safe there. So they want to send the kids somewhere else. So some kids whose parents work at the U.S. Embassy and like just were kind of there for a little bit of time. And what I found to be true is that like, you know, just spending time with people and understanding their own story and family life. [1:39:50] really led to a lot of shared interests and passions. [1:39:55] It helped me to see that the world can be very [1:39:59] similar in so many ways, and it is very much reflected to me in my day-to-day life in tech. [1:40:05] Because as you know, in the Bay Area, so many people who come from so many walks of life end up [1:40:09] in this journey building these products and companies. [1:40:11] and just helped me to value so many different voices. Like if you want to build world class products, [1:40:18] if you want to build product that can scale around the world. [1:40:21] if you want to build product that's going to be hyper growth, you have to be inclusive of so many voices [1:40:26] And so you have to build that skill speed. [1:40:28] able to acknowledge and learn from and to live in tension with different voices. That is beautiful. I love that we're ending there.

[1:40:34] Bengali, I feel like we've done a lot of understand work. [1:40:37] And I think we've helped a lot of people with a lot of things we talked about so much. [1:40:41] Two final questions. Where can folks find you online if they want to follow what you're up to? And how can listeners be useful to you? [1:40:46] For as much as I've worked at social companies, I actually don't use my social products that much. I'm on Instagram. Like, if you look at I am Bengali, I am B-A-N-G-A-L-Y, Google that, all of my, like, socials, handles all have the same kind of, like, link, including LinkedIn. So, yeah. [1:41:04] hit me up whenever. And then what can people do useful for me? I don't know. I mean, I think [1:41:11] I don't have a short list, I would say. [1:41:15] I love to hear people's stories. I love to get advice. You know, obviously my time is a little bit limited occasionally, but I think if people [1:41:22] can share with me things that they're learning or things that they see in the market. [1:41:26] or if they have questions, I'd always love to hear it. And so I think for me, just hearing other people's stories and learning what other people are doing is probably the most rewarding for me. Bengali, thank you so much for being here. [1:41:38] Thank you. This is amazing. I really appreciate you, Lenny. Bye, everyone.