Transcript for "How Notion built a $10 billion product": Hi. I'm so, so excited to have Akshay from Notion here with us today. Akshay, thanks for taking the time to meet with us. Yeah. Thanks for having me. It's good to see you. Awesome. So I think everyone here has probably used or tried Notion at some point. We at Merage use Notion every single day. It's definitely our source of truth for all company information. So I'm really excited to have you and share and talk a little bit more about your product team, how you guys make product decisions, and just overall how product has really changed from when you first started Notion to now. Do you wanna share also maybe some background about yourself, and how you came to where you are today? Sure. Yeah. So I've had an interesting journey. So I I'm definitely a product person. I sort of worked in products for almost a decade. But then when I joined Notion back in 2018, I was responsible for everything except product. Ivan was our CEO and he was very much sort of wanting to focus on product. And so I spent about 4 years really building up our go to market teams, our foundational people finance legal teams, eventually hired leaders for all of them and nuked myself out of a job. So about a year, about yeah. My 5th year at Notion, I was essentially an IC just working on AI. And, during that year, we actually spent, a full year trying to find a, product leader, because the team had gotten sizable enough that we were looking for someone who could run product here. And, I feel like after a year of meeting candidates, Ivan asked me if I can go back to product. And so, and so so yeah. It's been about a year since I, took on the product role. And, so so it's been yeah. It's like a full circle. I sort of say that I I, I've, you know, I think, like, my 5 year detour of working in go to market and sort of company building has met has made me a much better product builder because I have so much more context for for that side of the world. Got it. And is there anything that you wish, like, since it seems like now you're re coming back in, is there anything that you wish you knew from that you have now learned from your go to market experience that you wish you guys applied in the earlier days? I think there's always this natural, like, I wouldn't say tension or friction, but there's always, like, product feeling like, why doesn't sales do it that way? Or why does marketer have to do it that way? And then I think when you run that world, I think you can quickly realize, like, it's like, oh, why can't product just deliver on the things we've delivered on? Like, why can't they give us a road map? Like, you know, so I think, like, you realize the the sort of, like, the different information you have. But also, like, you realize how hard those jobs are. Like, it's not easy to be a salesperson, especially for a product like Notion that's so horizontal. Like, it's like, oh, each person has a very different use case. And so so I don't know. I think in in many ways, like, I feel, grateful for that experience because, we as you recently, we sort of came up with, like, our strategy for 2025. And, you know, we've gone through, like, several iteration of, like, what the company strategy is and what they eat what's the product strategy and what's the good market strategy. Last month, we came over the 2025 strategy. It's a single page of doc, and that single strategy doc is our company strategy, is our EPD strategy, is our real market strategy. It's all in one doc, and we all look at the same thing. And and I think that's really powerful so that it feels like if we win, we win together. If we lose, we lose together. I totally agree. And especially since your company has grown so much over the past, few years, especially with headcount, I'm sure things got a little bit more siloed just because there are so many more people. So, it's probably been great for you to be able to reunite all these teams together again. Out of curiosity, so this product the entire team has grown and the product team has probably grown as well. What were the biggest challenges when you were scaling the product team, and how did you end up overcoming them? Well, so Notion had a very interesting history on this. We were we were essentially just quite late in terms of building the PM team particularly. So actually even up until 3 3, 4 years 3 years ago, we had, maybe, like, 2 PMs and, like, maybe, like, 3 designers. It was, like, tiny. Yeah. I think we had 3 designers, for about a 100 engineers or something. So it was, like, kind of insane. We were just very late. And I think part of the reason we were able to be late was because I think we hired engineers who were very product oriented, And then we also hired designers who were who were all coders also. So I think essentially we had these people who could take, end to end, and they were involved in meeting customers and figuring out what's wrong with the product and actually building it and shipping it. So so I think we we got lucky in that world. Of course, but as you grow, like, if you realize that, oh, there's always, like, cross functional work you need to do to just, like, even, like, work with marketing to get ready for a launch, like, do beta testing and collection of all this feedback and, you know, setting the strategy for next year, which was all time taken away from engineers to build and designers to design. Right? And so so I think we we introduced that pretty late, but then I think we've tried to keep it tight. Like, even today, there's only about 20 PMs, and and only about 15 designers at Notion for for the space we have. Probably could have had a lot more people, but we've tried to keep it, tight. And so I think the challenge has been to try to find the balance between giving people enough of a canvas to be able to do good work, but not so much that it feels like we're sort of burning them out, but also not so little that it feels like things are very siloed and you're suddenly shipping your org chart. Right? So trying to find that balance is is being key. Out of curiosity, when you only had a few product managers for the 100 of engineers, how did you decide what they would end up focusing on? I think our first product manager, worked on growth because it felt like a it's almost like a new, team we could build that's a layer on top of the core product work. I think the core product work and how it's evolved, has been constant iteration every 6 months. Because I think at the end of the day, Notion is a is one big interconnected system. And so we cannot ship, like, I think we've we've tried very hard to not ship things, like, in silos. Because I think, like, if you build something, let's say you build charts, which we did recently, but charts is a view on database. And, like, whatever you do on charts will affect all the other database, and whatever you do on the database will affect all the pages and so forth and automation and so forth. And so I think, actually, like, you know, one of the things we recently realized, like, there's only 2 big things we look for in product people here. One is systems thinking because Notion is one big interconnected system, and you need to have that understanding, to to really understand, like, how we build. It's almost more important. Like, we we actually go slower to make sure the interconnected system stays alive. And then the other thing we really care about is craft. We care a lot about, like, sort of, you know, thinking very maniacally about the details. And and some of that comes from the top with Ivan. Like, you know, I think we will not let things ship unless we feel like we're proud of it. And you mentioned this a little bit earlier also, but there are a lot of different use cases that, Notion could be used for. So when it comes to bugs and customer asks, there probably is a wide variety of things that you guys could be building. How do you end up prioritizing, especially with, a leaner product team and so many customer asks and especially with a consumer angle? I'm sure there's even, like, millions of, different thing directions that you guys could be going into. So it's definitely been a a bit of an evolution, and I think it's probably true for most companies where the early early innings of this is actually pretty straightforward. It's just like, take a look at everything that's coming in, take a look at whatever is the highest. You just go build that because I think that's easiest and perhaps the most effective. I think you get to a point where there's probably more, coming at you than you can build, and that that part you still have to prioritize. And I think, I think for us, I think as you called out, there's the other interesting aspect of just very different personas using it. It's like someone using it on their in their personal life versus someone some SMB using it versus a mid market use case versus like a large enterprise use case. The way we've we've think thought about it is is a bit like we we we take in obviously all these different asks and we we have a way to slice them across those different personas. And then we try to almost, refactor or abstract them to the underlying building block that they're asking for. So again, going back to the charts example, charts can be used by a personal use case of, like, someone just wanting to track their gym workouts. But then the same primitive can be used to track, like, your projects and your tasks and whether you're, like, you know, whether it's all leading up to your OKRs. Right? And so so we have to do this, like, it's it's almost like we have to do this work to make sure that the primitives that we build, the new systems that we build actually work for all these different use cases. And and that means that I think, like, every permit we only ship, like, 2 or 3 new primitives every year. And the the reason is that each primitive actually takes, like, 6, 9 months to get right because of these different use cases. And maybe the last thing I'd say in terms of sequencing, one of the thesis Notion has had is has been around this idea that that there's only, like, 20 or 25 key systems in in any software. And so and so as we think about the use cases we have, there's, like, a natural sequencing that we need to do to to that that makes sense. So for example, like, you know, databases have to come before automation has to come. Automation has to come before something else has to come. Right? So I think you have to sort of we almost have, like, this grid where you take a look at all the use cases and all the building blocks on on the x axis of the use cases on the y axis. And then you can sort of see, like, okay. Well, if I wanted to really make project management work, what primitives do I need to use and, like, what's the right sequencing of that? So, anyways, long story short, it's a little bit of art and science now, but it used to be, a lot more art in the very beginning. It became a lot more science just looking at feedback, and now it's a bit of a blend. I'm sure there's also, use different use cases with overlapping asks as well, especially if you are mostly looking at the primitives and the building blocks. And out of curiosity, do you weigh consumer how like, how do you weigh consumer feedback versus enterprise feedback? That's a great question. I think one of the one of the things I mean, we've obviously made a bunch of flip flops here along the way. But I think where we've landed on, which I think is pretty helpful, but at least my product team tells me it's pretty helpful, is that Notion is for knowledge workers. And and at the end of the day, these large workers will use it in their personal life or will use it in work. But in both cases, people use it for projects. So I I could use it with my wife on kindergarten applications, but I can also use it with my product team for, like, I don't know, like, the next thing we're shipping. Kindergarten applications can get really complicated too. So They do. They do. Exactly. And so so so so I guess the the the the realization is that we're not a consumer app. Like, we're not a consumer app where we're gonna make money of ads. At the end of the day, we are a work product, and and work can now happen at home or in office. Right? And so so it's a little bit less about, like, consumer enterprise. It's a little bit more about, like, how do we enable people using Notion alone with a few people and then with lots of people. Right? And and at the end of the day, people are just collaborating on things they want to do in the world, and and we have to sort of enable that. Cool. Are there any unique processes that, unique processes to Notion that help your product team move more effectively? One of the things, Fuzzy, my engineering partner and I instituted about a year ago when when we took on these roles were, I actually took about a quarter where we it was a bit of a mess. Like when we joined, we we we asked this question, which is what are the 20 projects that matter? And and people pointed us to the task database or the project database, and I don't know. There's, like, 1,000 if not tens of thousands of those things. And so I it's like, I just didn't have a sense for, like, what what are the things we track. So one of the things we did was we we we sort of called it outcomes. And and essentially we created a new database called outcomes and we said, okay. Here's the 25 things that matter. And if you look underneath, I think there's the there's a natural hierarchy of systems. So there's a task which could be like, I don't know, fix a specific bug, which is the lowest level of, you know, unit. The unit above that is project. So you could have a project which is, let's say, charts as a project, and, like, you could have a bunch of tasks for it. But then there's a there's a there's an outcome above that, which could be like there's actually, like, these 4 projects will lead us to, like, ship project management, for example. Right? And and so there's roughly about 20 outcomes that really matter, and we put it out there and people can quickly see outcomes to the projects, projects to the tasks. Right? And and so there's this hierarchy that I think everybody inside the company can see. So so we came up with that list and if that list can evolve every quarter as we ship things. And and across those outcomes, we created a very simple way for PMs to put their weekly updates. And it obviously took some convincing to everybody to, like, follow that process. But now that people follow that process, it saved Fuzzy and I so much time. Like, when we when we got on, like, we just got questions after questions, like, why is this project delayed? What happened to this thing? What happened to that thing? And now that everybody puts their weekly updates, nobody comes to us. Because essentially, you can use Notion AI to ask a question. It's like, what's the latest release date for this thing or why is this delayed? And you just get an answer for that. It's actually a funny story is that Ivan, our CEO, would message Fuzzy and I, it's like, hey. Why is this project, like, not moving as fast as possible? And then Fuzzy would spend 2 hours talking to his engineering managers, figuring out, like, what the hell is happening? And I would literally take his question, put it in Notion AI, then take a screenshot of the answer and send it to him. And Fazio was like, damn it. I should've done that. Well, then it's just that, like, year of you being, like, an AI product manager ended up saving Fazio, like, lots of time. So you're our next that's actually our next step, which is currently humans. Like, PMs actually spend, like, an hour running the weekly update. But, like, all the information is there. So so, technically, like, our hope is that 80% of that update could be written in AI. You can just come in, tweak a few things, add your own commentary, and then, you know, you'd save an hour. But it also goes to show, like again, it doesn't have to be Notion. It could be any system you use. But if you can consolidate the information into a system that can track and you have this connectivity of the systems, like, where where okay. At the highest altitude, it's outcomes, but, like, I can go from outcomes all the way down to task and actually find out what the hell is going on with that project very easily. And it's sort of truly democratized. Right? So, like, the marketing team can actually map out their road map and, like, every changes that happen for specific outcome, they can actually go do that. It's really, really powerful. Like, I I I mean, I was looking at LinkedIn and we didn't work this way. And it's almost like a like a little bit where sort of Notion on Notion is a bit like we're using Notion instead of steroids and, like, we're just breaking the system. But I think it's very powerful, way of operating and maybe it reduces the need for having a ton more people too. Right? It's like each person can do more because because everything is there. That's really cool. We might steal some of that too, actually. Do you think, I guess, was so you mentioned that at LinkedIn, you guys didn't end up doing anything like that, especially on such a granular level. Was there anything from LinkedIn that you really liked that when you started Notion, you were like, okay, we're gonna I'm gonna end up doing this here. And were there any things that you didn't like or disagreed with where you felt like because, you know, you had more control and notion that you were able to really make sure it was consistent as a business process. Well, there's a thing. There's there's an arc. It's less about product development, but there's an arc of just the the culture, the values, the mission, the strategy, of LinkedIn as a company was just extremely clear, extremely inspiring, and extremely consistent. In my 6 years there, but also, like, maybe in the last 15 years, I think the company has been very consistent. Almost to a point where the 10,000th employee within the first two weeks would just be, like, brainwashed into knowing what the vision of company is and why the work they do matters to them. And I think that is very underrated. You know, it's super underrated. Like you you sort of lose that as you scale the company. It's like your your 50th PM, do they understand what the mission of the company is? What the strategy is? Can they like can they can they say exactly the same words of the mission statement or will they, like, say some version of it. Right? I think that part is underrated and I think that's underrated because the thing if you can somehow get that, you know, get that working, then you're all rowing in the same direction. Like, you're not like, you don't have, like, different interpretation of that. And so so I I I think I think I I think a lot about that. It's less product, but it's more like, as a company, can you do that? Because then, like, each PM's understanding of what the what we're trying to do in this company is exactly the same. Yeah. And it's really hard to do especially with more people because you basically need everyone to be, like not necessarily making, like, this like, the same decisions that they would have as, like, you would have, but able to roll in the same direction, which is really hard, especially if it's too vague or if it's too specific. You don't wanna take away agency as well. So I'd note I mean, there's a yeah. There's a cool hack that we that not hack, but, like, there's something that Ivan did recently that I think I thought was brilliant. When we when we when we published our 2025 strategy doc, he obviously walked through the company with that doc and and but then he went to every leader in the company and he said he basically gave them homework. He said, here's a strategy doc. Shoot a 5 minute loom video where you talk about the strategy doc and what that means for your team. Oh, interesting. And and then he said you publish that. And basically now we have our library of everybody's videos talking about the same strategy doc and what that means. And then you you can now, like, also you know, it's basically a practice for them because you have to do that for potential cap recruiting as well. But I think it's so helpful for internal everybody to, like, recognize that and, like, everybody to, like, take it down to the next level. Right? Because I mean, you can I mean, the CEO will do that and the product leader will do that every now and then, but if everybody can actually talk about the same thing, it's it's super powerful? That's really interesting. Okay. I might seal it to you. And we're I guess, we only have a few more minutes left before, before the end. So out of curiosity, so, obviously, you've seen a lot of generations of product at LinkedIn and then also Notion as well and then back to being an IC. Where do you think the product management function is going, long term? Well, I think, one of the things I've realized, at least the difference between, like, LinkedIn and Notion is that, I think LinkedIn was a little bit more, like, much cleaner swim lanes. It's like product managers set the vision, designers design, engineers then, like, take that thing and actually build it. And I think Notion is much more, like, blurry. Our PMs have a deep understanding of tech and design. Our engineers have a deep understanding of product and design. Our designers are coding and thinking about this. So so I I I I wonder if it's like I don't know. I I I hope the future of PMs are are not just, like, sit in your castle and figure out what the vision is and hope that the team executes against it. But it's more like, you know, prototype and build and and figure figure out what is needed to be done, and and sort of enable your engineers to build and your designers to design and so forth. Right? It's a little bit sort of like more hands on. I I heard this, like, amazing definition that one time LinkedIn founder cofounder, Alan Blue gave me, which I thought still resonates with me, which was I think he basically said, like, actually, like, everybody works on product. Everybody's connection to product is different. So the engineers connection to product is technology, designers connection to product is users, and product managers connection to product is the market. And so in many ways, like, I think product manager's job, the key job that I think is still gonna be critical is to make sure that we're doing everything you can to derisk, everything to sort of enable product market fit. Right? And so I think you have to sort of have, like, this understanding of the market and understanding of this macro that that really sort of gets you to making sure, like, the product is able to succeed. And and that job is going to be as relevant as ever in the future. So sorry. One more follow-up question on that. So because of those changing requirements and now it seems like the scope is increasing and the requirements are increasing as well, how does that impact your hiring criteria for product managers? I think we we we look for those 2 things specifically. 1 is we look for builders, not just like document writers or planners. We need people who are sort of much more hands on. And I and and in today's age also, like, you just cannot have you cannot, like, you just cannot, give a reason for not being able to do that. Like, you just can't say that I don't know how to code because, like, I mean, you know, there's enough tools where you can speak English and you can code now. So, like, there isn't any reason for you to not to be able to do that. And also ask you to explain everything now too. Yeah. Ask you to yeah. It's like it's like basically it's for people who are gonna lean into this thing, like, it's it's gonna be just a beautiful new world. And people who are not gonna lean into it, I think we're we will actually struggle struggle. And then I think we're we are we're I think the other flip side, which is what I talked about is I think we're looking for strategic rigorous thinkers who are able to watch the market in a way that engineers and designers are not gonna be able to do that because they're building, you know. And so I think I think those 2 skills are gonna be very critical in terms of whoever joins Notion as a PM. Cool. Awesome. Well, thank you so much for taking the time to share all of your thoughts today on product. I really appreciate it. And if anyone has any follow-up questions or wants to get in contact with you, what's the best way to do that? Twitter or x. I'm just at a kothari, and my DMs are open. Cool. Awesome. And if you're interested in chatting with me, I'm also on Twitter. But you can also email me at chuncmerge. Dev and then you can also find us at www.merge. Dev. Thank you everyone. Really appreciate you taking the time today.