20Product: Slack CPO Noah Weiss on How to Master Product-Led-Growth, The Biggest Mistakes Founders Make When Scaling Into Enterprise & What Needs to Change with your Product, Team and Processes when Scaling From PLG to Enterprise

Noah Weiss is the Chief Product Officer of Slack, overseeing the product team’s strategy and development. Over his seven years at Slack, Noah has led various parts of the product organization, including the self-service SMB business and product-led growth; the Virtual HQ team that launched huddles and clips; and the search and machine learning teams. Prior to Slack, Noah served as SVP of Product and Analytics at Foursquare. He started his career at Google leading the structured data search team and working on display ads. In Today's Episode with Noah Weiss We Discuss: 1.) Entry into Product and Road to Slack CPO: How did Noah make his first foray into the world of product with Google? What are 1-2 of his single biggest takeaways from his time with Google and Foursquare? What model did Noah learn at Google that he applies to product today? 2.) Product 101: The Foundations: Is product more art or science? If Noah were to put a number on it what would it be? What are product principles? What makes good vs bad product principles? What are the biggest mistakes that founders make when instilling product principles? Does Noah believe with Gustav Soderstrom, "talk is cheap and so we should do more of it"? 3.) How to Master Product-Led-Growth: What are some of Noah's biggest lessons on how to master PLG? What are the biggest mistakes Noah sees early stage founders make today when going for the PLG approach? How does he advise them? When is the right time to move into enterprise? What needs to change? How do you change who you build product for? The buyer or the user? Why does Noah believe product speed will always be the most important thing in product? 4.) The Internals of Slack: How does Slack do post-mortems today? Who comes? Who sets the agenda? How has this changed in a world of remote? What does it take to do them well? How do Slack do product testing pre-launch of new products? Do they know when something is going to be a hit? What did they think would be a massive hit that turned into a flop? What does Noah believe is the biggest near death product experience for Slack? What happened? How did they get through it? Why do Slack buy other companies? How do they think through the decision of buy vs build? When do acquisitions work? When do they not work?

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Speaker 00

  The most important feature by far for any product is actually the speed above all else. That's like the oxygen for a good product experience. I think we're good at talent integration. I think we struggle at product and technical integration.

• 0:14Harry Stebbings?

  This is 20 Product with me, Harry Stebbings. Now, 20 Product is the monthly show where we sit down with the best CPOs in the world to hear how they start, scale and manage the best product teams in the world. And we have the incredible Noah Weiss, CPO at Slack in the hot seat today. Now, over his seven years at Slack, Noah has led various parts of the product org, including the self-service SMB business and the PLG Growth Engine, the virtual HQ team that launched Huddles and Clips, and the search and machine learning teams. And prior to Slack, Noah served as SVP of Product and Analytics at Foursquare, and he started his career at Google, leading the structured data search team and working on display ads. But before we dive into the show today, this episode is brought to you by Linear. Let's be honest, the issue tracker you are using today, it's not very helpful.

• 0:59Harry Stebbings?

  Well, Linear is different. It's incredibly fast, beautifully designed, and it comes with powerful workflows that streamline your entire product development process. From issue tracking all the way to managing product roadmaps, Linear is designed for the way modern software teams work. Its powerful workflows let you customize and automate processes to fit your team's unique needs. And they have integrations with your favorite tools like GitHub, Slack, and it takes just minutes to set up. Linear is the default tool of choice for high-performance teams, from startups to a wide range of large, established companies like Vassal, Retool and Cash App. So see for yourself what makes Linear magical. Visit linear.app/20vc to try for free with your team and get 25% off when you upgrade.

• 1:43Harry Stebbings?

  That's linear.app/20vc. And speaking of tools we cannot live without, I need your input on the 20VC Miro board about new guests we should feature for 20VC in 2023. Just head on over to miro.com/20vc. Miro is a tool I consider to be truly game-changing. It's a visual collaboration tool packed with the right tools, tech and templates to help you think of and create that dream product. That means you can brainstorm. Storm the perfect product with your team. Vote on the best ones and explore with your customer journey roadmap all on a Miro board. Whatever you need, Miro's infinite whiteboarding capabilities help you get there. It acts as your team's single source of truth. And now I'm using it to hear from you.

• 2:26Harry Stebbings?

  Go ahead and add your suggestions for 20 VC guests on our Miro board at miro.com/20vc. And finally, you have to try EPO, the next generation A B testing and feature management platform designed to help you run reliable impactful experiments. EPO saves you time across the entire experimentation workflow. Planning tools help run experiment scenarios beforehand. Best-in-class diagnostics mean you spend less time debugging issues. Their cutting-edge stats engine helps you reduce experiment runtimes. Plus, EPO sits directly on top of your data warehouse and your North Star metrics. So it gives you this real confidence in experiment performance and ease in conducting follow-up deep dives. No more extended analysis cycles to understand results.

• 3:11Harry Stebbings?

  And that's why companies like DraftKings, ClickUp, Momentive and Cameo all rely on EPO to power their experiments. Get access today at getepo.com. That's getepo.com.

• 3:28Speaker 02

  You have now arrived at your destination.

• 3:31Harry Stebbings?

  Noah, I am so excited for this. I've really been looking forward to this one. So thank you so much for joining me today.

• 3:37Speaker 00

  Thanks so much for having me, Hera. I've been a longtime listener, excited to finally join.

• 3:40Harry Stebbings?

  Well, that is very, very kind of you. Now, I wanted to kind of chronologically unpick your career before we dive in. And start as your career, starting as PM at Google. What is the single biggest product lesson you took from the two to three years at Google? And how did that impact your mindset?

• 3:56Speaker 00

  I think the thing that struck me the most in backroom, they used to have this model, they called it the 70-20-10 model. 70% of the product teams were focused on what are things that are known valuable things for our customers that are in some ways more incremental. 20% were new product areas that had already kind of taken off, but they were trying to further refine and incubate. So, you know, App and Gmail took off. How do you scale it to a billion users? And then 10% were the far out bets, the self-driving cars, the Android before Android existed, the building a web browser. And I think the thing that stuck out to me was the combination of that deliberate approach to what the portfolio looked like. And whatever your ideas you bring to them,

• 4:36Speaker 00

  what would 10x a bigger scale look like? So I think that caused a level of ambition for product thinking at Google that it's kind of hard to afford at most early-stage startups, but I think it's aspirational to bring to every company.

• 4:49Harry Stebbings?

  Do you do the 70, 20, 10 at Slack?

• 4:51Speaker 00

  We do talk a lot about kind of the portfolio and diversification within different product areas. So, you know, we might ask it to you, when you think about the breakdown of your roadmap for the next two quarters, where is it allocated? How much of this is maintenance, performance and quality? How much of this is refining and kind of customer delight? And how much of this is you're incubating new ideas, you're trying out new levers? We don't call it 70, 20, 10, but it's definitely part of how we'll look at the structure of product team roadmaps. And every team is different, too. You know, even at Google, I would say that was at the corporate level. But take a team like AdWords, their allocation didn't look like 70, 20, 10. They might look like 95, 5. It's a interesting framework, but I don't think it is something that is like a perfect ratio

• 5:35Speaker 00

  to apply in every context.

• 5:37Harry Stebbings?

  What's the biggest product takeaway for you from Foursquare? It's a fascinating product actually to have the experience you did on. And how did that impact your mindset?

• 5:44Speaker 00

  You wind up learning a lot more, doesn't matter what your role is, when you're actually faced with a lot of headwinds. When everything's going up and to the right, you think everything you're doing must be brilliant. You know, you throw something against the wall and the metrics keep rising. And so I think it's really hard, and I think this is a knock on, I think sometimes folks who have only worked at large companies is you have so much tailwind, you have so much momentum, you have so much distribution. You don't learn how to manufacture that from scratch, or you don't learn how to turn things around and things go sour. Within the context of Foursquare, basically the thing I took away was that product market fit. People talk about it like it's a binary thing. You unlock it and then you have it forever. And I actually think a much more accurate model is that you unlock product market fit, you get to maintain it for a while, but you have to keep renewing it as your audience changes and as it rolls around.

• 6:30Speaker 00

  New changes. So I think with Foursquare, we definitely had to go past the early adopters who were so motivated by discovering the world, seeing new places, novelty in the social side of it. And then also the world changed, understanding how competition changes the backdrop with Instagram coming out, really shifting how people viewed what it meant to share experiences in the real world.

• 6:50Harry Stebbings?

  I totally agree with you in terms of like the need to renew product market fit with the changing of company scale. What are the biggest challenges to doing that?

• 6:59Speaker 00

  I think if you look earlier stage, the single biggest challenge, because I think the first time you had product market fit, almost every company, they're designing for themselves in some way. What's the founder's right to compete in this market? What's their founder, you know, market story? It's almost always, I have this problem in my personal life, or I see this problem in my work life, and I want to solve it for me, and hopefully other people want it too. Slack, obviously, was was two stories of that. You know, the first was trying to build for this game that they wanted. That didn't take off. The second version was, hey, we built this amazing tool internally. Maybe other people might want it too. Let's hope that there's a big enough market. Once you saturate that market, other people do want the thing. I think the biggest first leap is,

• 7:38Speaker 00

  how do you have the self-awareness, the humility, and then the intellectual curiosity to learn about what is the next audiences that you should be kind of designing the product for, who look fundamentally different from you? And how do you make that leap as an organization, as a culture, as a frame of reference, where you are your next customer? That's a huge leap, and I think really hard for most companies to do.

• 8:02Harry Stebbings?

  How do you do that? And how much do you listen to your customers versus independently progress on your own research and thought processes?

• 8:11Speaker 00

  Yeah, I think you have to get even deeper into your customers' lives and better understand. So, you know, as a case of Slack, I think it was initially, if you read the initial pitch deck, which they shared when they pivoted the company, it said Slack is a product for teams of five to 50 people. The initial conception of Slack was after 50 people, this thing probably won't scale. No one's going to want to work this way, so let's just design for those small, medium-sized teams and companies. And I think the biggest change was when we realized this was the At the data that actually small teams at large companies were loving using Slack and all these different pockets as they were springing up. But to actually make Slack work at a scale of thousands or tens of thousands of people, the needs of that type of company looked very different.

• 8:53Speaker 00

  The end users wanted the exact Slack product, but the organization wanted something very different. We had to spend, it was probably end of 2016, 2017, really immersing ourselves in the world of like enterprise controls, scale, security, compliance, talking to executives at companies instead of just being the insurgent and getting closer to them. Not because everything they said we should take literally, but because we weren't designing for ourselves because we were only maybe 200, 300 people back then. So you couldn't know what a 30,000 person company actually wanted. You have to actually go talk to them and find out. And then incorporate it into your view of how to develop the product.

• 9:30Harry Stebbings?

  Now, before we move into that kind of scaling up into enterprise, you mentioned there the five to 50. I do just want to touch on kind of the foundations being kind of product principles. And you said before the show to me that your organization needs product principles. First, what are product principles and what did you mean by this?

• 9:44Speaker 00

  Yeah, product principles are a way of enshrining the culture and beliefs of your product organization into a common language that everyone at the company can easily refer to as a shorthand for making quick decisions and for making qualitative assessments of the level of craft of the product. And so for Slack, I think the why behind why we created these was as a company was scaling and the average person who's working in product development is kind of further away every day from working with Stuart, who's the founder or CEO and very product led. It became harder to figure out how to instill the fiber and DNA of how the product was initially built into teams that were further away from that origin.

• 10:27Speaker 00

  And so what we realized was you couldn't just have Stuart spend all day in meetings with every single product team. How do we scale the culture and how do we make it easy and memorable to reference? So we came up with a couple, things like don't make me think and be a great host and protect the path and take bigger boulder bets. And we can unpack any of them, but.

• 10:45Harry Stebbings?

  Let's unpack them. I'm, I'm going to be deliberately contrarian. They're all quite general, bigger boulder bets.

• 10:52Harry Stebbings?

  I get it, but kind of nuanced. If I'm going to come with an idea that's like bet the company, you're telling me to take bigger boulder bets. Help me understand this because they seem pretty challenging.

• 11:00Speaker 00

  Yeah, I would say I think they mean something to the context of their organization. When you have half these, but we may only say take bigger boulder bets. As a company scales, I think there's a tendency towards incrementalism and kind of local optimization. You have bigger and bigger organizations, you have feature teams, the feature teams may have a single KPI that is the one thing they feel directly accountable for. And the natural tendency is I view my success, my team's success to move this single metric. The easiest way to do that is incrementally with very small experiments. So I can say, hey, we move this metric by 0.6%. And that's kind of antithetical to if you're saying, hey, actually, our product is much earlier in its journey. We're defining an entirely

• 11:44Speaker 00

  new category. No one else has created this space. We're the ones who are creating this space. Yes, we need to do things that are incremental and refinement, but we need to balance that with taking huge swings to push the concept of this category for our customers.

• 11:58Harry Stebbings?

  What do you think are the biggest mistakes startups make with product principles?

• 12:02Speaker 00

  Most companies wait too long to introduce them, whether they call them principles or maxims or design guidelines, it doesn't really matter. It's really hard if you're a product founder and you're a founder led company. But I think one of the hardest things winds up being when you give up enough control that your organization starts building things that you weren't even aware of. And then how do you have enough trust that the people building it are going to build it up to the standard and approach that the company was founded on? That's the leap. But I think most companies wait far too long and then the organization starts going slower and slower. And then the CEO starts complaining, why are things slower when we have twice as many people? I would say probably waiting too long to try and the culture in a way that can scale.

• 12:47Harry Stebbings?

  You said there about kind of speed with scaling teams. I had Gustav Soderfjord from the CPO of Spotify on the show recently, and he said, talk is cheap and so we should do more of it, meaning we should have more internal discussion, more debate around products, and that leads to better outcomes. I bluntly disagree. I think the speed of execution is everything and you should move as fast as humanly possible on non-core items. How do you think about internal product debate and whether it should be fast, slow, and what to get the most out of it?

• 13:15Speaker 00

  Fundamentally, I think it depends on the type of thing you're debating. I think it's Amazon, maybe that was kind of famous for the two class of decisions, like there's the two-way doors and one-way door decisions. I think the thing that we've learned over time is separating the two and trying to empower teams who have context, understand the strategy, have principles, to make as many two-way door decisions as possible locally and not having to have a lot of talk and debate. And then on the flip side, really being clear, what are one-way door decisions that actually do need the entire kind of executive team's buy-in to actually make that call, make that investment? And then I think talk is cheaper than a bad decision, but that should not be the vast majority of product discussions, in my mind at least.

• 13:56Harry Stebbings?

  How do you do product reviews? How often do you do them? Who's invited? What do they look like?

• 14:00Speaker 00

  Yeah, it's evolved a lot over time. I would say where we are now, and you know, for context, Slack has probably around 1,200 people across product design and engineering. So it's a pretty sizable team. Where we kind of are structured now is there's basically product pillars, which are teams of teams that are responsible for areas of the product. So there might be an enterprise product team or the virtual HQ team, that's responsible for huddles and clips and so on. And what we've figured out works pretty well is that we have those pillars, the leads for those pillars, do product reviews or product workshops on a weekly basis within the team. So the leads can kind of weigh in and give feedback and unblock feature developments that's happening.

• 14:41Speaker 00

  And then what we wind up doing is usually on maybe a bi-weekly or depending there are a monthly basis, an exec review with the leads of each pillar, the PD kind of executive leadership, so the head of design, head of engineering, the head of product, and try to focus those discussions less on, okay, let's run through the 15 features that might launch in the next couple of months. Instead be like, what are those one-way door decisions that we need to really focus on? What are things around the product roadmap and the portfolio that we need to refine? Really major launch. Let's go taste the soup. How's it feeling? Is it up to the bar? What's broken? Where do we need to push on the quality bar?

• 15:19Harry Stebbings?

  How did they change in a world of hybrid?

• 15:22Speaker 00

  The biggest change was they became more inclusive, actually. Because what used to happen, I mean, you probably know this dynamic, right? It's everyone is familiar. You're in a room. And once the room gets to a certain size, it feels like you're not in a collaborative discussion. It feels like you're in a performance. And so back in the day, when it was all in a state physical meeting room, it kind of felt like you had to set a hard cap. I don't know if it was eight or 10 or something like that, where you're like, okay, this is a single discussion. And so I think what changed in hybrid, where you could have a much larger room, but not feel like you're doing a performance as much is that the reviews, which I tend to love actually is, you know, if we're reviewing a key product, actually, we have it right after this today about the new information architecture that we're working on, kind of a redesign of the Slack navigation.

• 16:07Speaker 00

  Many of the tech leads, some of the design leads, multiple PMs working on this project, all there in one place. And that doesn't mean that it's completely a chaotic discussion. We're still mostly having a review. But I think the beauty of it is it's more inclusive and people can have a context immediately instead of the typical trickle or cascade of information. In that way, I think it's actually been a real beneficial change.

• 16:31Harry Stebbings?

  You know, you're the only product leader who said a positive about it being a hybrid. Every single leader said it gets worse. The quality of discussion is worse when it's not in person.

• 16:40Speaker 00

  It's harder as an executive. It's less fluid, but I think the team, I mean, even from the team's perspective, I think they actually can enjoy it more and get more out of it. Instead of the typical, it's a closed door, it's six people in a room. Definitely, this is not true of hardware. Talking to friends, for example, at Apple, I think as soon as it was physically safe, they need to be in the physical environment to actually touch the thing. I would say the other thing that's really changed is we do much more of what typically was in a synchronous review asynchronously now ahead of time, whether it's pre-reading or people record clips of demos that people pre-watch. So then the actual review is the part that needs to be synchronous discussion. But so much of the

• 17:21Speaker 00

  Or if it used to be in the meeting can actually be split off when people do asynchronously out of their own schedule.

• 17:27Harry Stebbings?

  How do you do that? Is that with mirror boards? Is that where, what does that look like?

• 17:31Speaker 00

  Yeah, what I would say is the most typical flow would be someone will write a doc that kind of frames what the discussion is, the key context decisions. But I think the big thing that's actually happened is designers and engineers now will record clips with screen sharing in Slack. And it'll be a quick three or five minute clip. And if it's a designer, they'll be like walking through a prototype in Figma. Or if it's an engineer, they'll be doing the same, but with real live code. And they'll actually show you the thing instead of say, hey, let me, you know, the classic meeting otherwise is a 20 minute presentation. Everyone's sitting there being like, okay, I wish I could just go through this quickly on my own. Now it's everyone pre-watch this thing and then come in with your perspective, your ideas, your feedback. So you can get straight into the discussion instead of having half the meeting be a presentation.

• 18:16Speaker 00

  So I think that's made a big difference.

• 18:18Harry Stebbings?

  What's the biggest broken process part of product building today for you and Slack that hasn't yet been resolved?

• 18:25Speaker 00

  What is actually the latest state of the world for all the major areas going on at the company? And I think Slack itself definitely doesn't solve that. Like, yes, you can be in all the channels. But for me, if I just open up a random team channel, it's gonna be really noisy. It's gonna be hard to know like, what is the actual latest? Should I be reviewing this thing? Is this just team discussion? But we wind up solving this like most people do. At the end of the day, you use spreadsheets. So we have a Monday meeting, PD Monday meeting, as we call it. Where like literally we just review in a spreadsheet, super manual process, but the top priorities for the quarter and latest updates. And it gives that kind of operational cadence to the quarter. But that's a solution that is not very elegant and is not very software. It's about as old as spreadsheets.

• 19:07Speaker 00

  But I think a lot of organizations rely on that and software is too rigid often to be able to solve it.

• 19:12Harry Stebbings?

  Indeed, the enemy of most SaaS companies is the spreadsheet. So I think you're not alone in terms of that. But I do want to ask, we've mentioned obviously Stack so many times, you are the masters of PLG. You've been building one of the iconic PLG products for the last five plus years.

• 19:26Speaker 00

  Yeah.

• 19:27Harry Stebbings?

  How do you preserve simplicity with scale and with time?

• 19:31Speaker 00

  Things when you go to the actual mission and vision of the company to make people's working lives simpler, more pleasant and more productive. And so I think everyone here, when we're thinking about how do we build? What do we build? What does the quality of craft bar look like? It does always come back to, is this simple enough? Are people going to be more productive or more confused by the complexity that we're introducing? And I think the other way that we do is through hiring. You know, the vast majority of people who have worked at Slack had never worked in enterprise software before.

• 20:02Harry Stebbings?

  Is simple always better?

• 20:04Speaker 00

  Simple from a, it's easy to understand. That is always going to be an important and good thing. And we talk about this a lot internally, simple doesn't always mean fewer clicks. So for example, I think in a lot of consumer, you'll, you'll say, okay, you know, we want to remove as many clicks as possible. Remove as many options, give less control. But I actually think in enterprise software, often more clicks can be okay because you bring people along. So a guided flow where you're only making one decision per screen can actually perform a lot better. You're helping people feel like they have a master of the software that they're using and feel confident in what they're doing. Simple is about how understandable something is, how comprehensible it is. It isn't necessarily how streamlined or how seamless it is.

• 20:47Harry Stebbings?

  When you review many of the product decisions you made with regards to like the product led growth features at Slack, if you were to choose actually a biggest PLG product feature mistake, what was that? And how did that change how you think about product today?

• 21:01Speaker 00

  Well, two, because one is not specifically necessarily PLG, but I think kind of affects the motion there. So the most obvious PLG one is that Slack always had a really generous freemium model. Like, you know, you have a free plan, you can use it as long as you want. There's no cap on users. The really only significant cap is just on how many messages you have access to, how many files you can upload and so forth. But what we didn't do for many, many years is actually never gave people a rich, valuable trial experience of what the paid product looked like. And we learned this in user, I mean, this maybe sounds obvious in hindsight, but in user research, what we realized was that many people had habitually used their free plan and didn't even realize that paid product existed because we're not banging you over

• 21:41Speaker 00

  Ahead with it inside Slack and then kind of just was like their conception of what Slack was had all the limitations of the free plan. And so they almost viewed as, oh, it's meant to be ephemeral, like, you know, the conversation disappear. You know, this is back when maybe Snapchat was really taking off. So people kind of thought that was the point of it. And so we wound up building a really robust in product trial program, many ways to flow into it. A guided experience help you make sure you're getting the value out of it. And that was probably honestly the single biggest thing we did on top of the freemium model to really push that self-service driven paid conversion. How did it change your mindset like when you reflect on that? I think for a long time, we have a perspective that if you use Slack for long enough, you'll eventually decide to pay for it because it'll be obviously valuable to you.

• 22:24Speaker 00

  And I think what changed, I mean, I remember when we review this and we kind of wrote the maxim is what we want to do is give people such a great taste of the full Slack experience that they never want to go back. And that's a change instead of assuming that they'll discover it on their own, that they'll have the intentionality. It's actually how do you give people a taste of that? How do you walk them through it? How do you make sure they fully utilize it? And then once that ends, they're like, oh, my God, I would never want to go back to the free plan. Like this is so powerful. This changed in my entire organization. So I think that was a big change there. The other thing which I was going to mention, not strictly PLG, but I have this product capability called Slack Connect. Whole idea is take a channel. Now you can share it across multiple organizations. So we could use Slack Connect between companies working on a project together, between vendors.

• 23:09Speaker 00

  We have Slack Connect channels with every one of our customers to do customer success and support. The Stripe's head of sales famously said, you know, once through the Slack Connect channel with a customer, I know the deal is going to get done. The thing that I think we were, you know, maybe overly cautious on is we were really, really slow and deliberate bringing that product to market because we were really worried that people's mental model of Slack was very much that it was this walled garden for my organization. You know, there was no external notion Slack. It was whatever was in Slack was just fully internal. We were really anxious frankly about breaking the walls of that garden with this idea of Slack Connect channels that could connect you to other organizations. And then to your point about simplicity, does people's mental model, what Slack is, get more?

• 23:51Speaker 00

  And I think we probably had it in beta for about two years. It took a really long time to bring to market. Frankly, especially given the competitive backdrop there, I think if we had expedited it into market back before there were more products that, you know, people were starting to use or get for free as defaults. And back when I think the ubiquity of Slack was something that felt like maybe inevitable. This could have created enough gravitational pull where existing customers who love it just forced the rest of the world to come use it with them. And I don't think we missed that moment, but I do think that moment was delayed enough that maybe we missed some of the more like exponential return to having brought it to market.

• 24:31Harry Stebbings?

  You mentioned speed there. I'm asking the tough. I used to be so nice on the show when I was younger. I'm asking the tougher ones. If we're honest, I'm sure you hear what I hear. People say the speed of innovation at Slack is slower today. What are the biggest bottlenecks to the speed of innovation in Slack?

• 24:47Speaker 00

  Where it was most true was 2018 to 2020 back then. But that's when we really made this kind of hard rotation on focusing on the enterprise buyer, the enterprise organization. We did prioritize for a while kind of scalability and solving those really complex Fortune 500 use cases and compliance and security needs. And I think we did take the eye off a little bit, that core pace of product innovation of pushing the boundary of what this category could be forward. COVID was a real wake up call for us because suddenly even our existing customers were coming to us and saying, oh my God, I'm living all day in Slack, but also now have all these other needs and can Slack fill these needs or do I need to look in other places for them?

• 25:29Speaker 00

  Starting the end of 2020 through now, I would say, and this is not a rebuttal, but my perspective of these is that I think that the pace of innovation has kind of increased pretty dramatically. You know, you have huddles, you have clips, cameras, which was just launched, you have workflow.

• 25:44Harry Stebbings?

  And that is because you have clarity of understanding who you're building for, be it the enterprise partner.

• 25:49Speaker 00

  I think what it actually is, is that we got to the place where we realized there was diminishing returns for focusing the majority of our efforts on the enterprise buyer. Once you get through a lot of the blockers. That are just, hey, the CIO is going to say no unless you have DLP and EKM and IDR. I'm not going to go through all the three letter acronyms that will be blockers. But once you do all of those, then you start realizing what we felt all along, which is you want to keep pushing the capabilities of the product for the end users, for the teams who are trying to work in a more productive, delightful way, and those teams need new things. And it felt like, honestly, and Stuart said this publicly too, the pandemic and the shift to this hybrid world,

• 26:31Speaker 00

  I think kind of gave us almost a new level of meaning and urgency to the work that we were doing. Because the level that people were depending on Slack, the expectations they had for what Slack could do had just grown incredibly, basically overnight.

• 26:45Harry Stebbings?

  No, you mentioned those three letter acronyms. Most people have no idea what they mean because generally, most people don't have to scale into enterprise. But many startups do, and many startups, actually, I think respectfully, don't understand quite what it takes. What do you think are the biggest mistakes that startups who start in SMB make when they scale into enterprise?

• 27:04Speaker 00

  It's hard because honestly, a lot of the work is really kind of removing blockers. And I think there, what I would really recommend honestly is like, that's where you do want to hire people who have domain expertise. You don't want a bunch of folks who worked at an early stage startup and never worked in enterprise before to be like, let me try and figure out from first principles what like electronic key management is and why the chief security officer of a Fortune 100 company cares about it. This is not a place where you should re-emit the wheel. You should not take a novel approach. But the biggest mistake I think people make, and it's a little bit kind of reflecting for ourselves, is over rotating on the enterprise buyer where you think that they're hardly more conservative. They don't want features to be shipping every single day to end users because they're, in fairness,

• 27:49Speaker 00

  trying to centrally run an organization where they have enablement and education and so forth. And I think it can introduce, to your point about speed, it can introduce a level of cultural conservatism because the enterprise buyer will say, hey, I have these blockers, unblock them, and please stop shipping things so quickly. And then the small and medium sized business will say, I don't even know what those acronyms are. I don't care. But what else can your product do? And by the way, I'm very value conscious. So I'd love your product to do more for the same price. And I think that becomes a tension. And I think if you over rotate, you can lose some of that speed, that product innovation DNA. And I think that's the thing to watch out for the most is not losing the core or the fiber that got you to build a product that people in enterprise want in the first place.

• 28:33Harry Stebbings?

  How do you know when's the right time to move into enterprise?

• 28:36Speaker 00

  I think there's two paths and there are some product categories where you need to start in the enterprise because there isn't an SMB buyer, things in the like security, compliance, infrastructure space. Those buyers, for good or for bad, are going to be the CIO or the CISO at an organization. To start with enterprise is often the answer and don't sweat that you don't have a bottoms up motion or PLG motion. I think if you start with SMBs, I think the thing to look out for is, are you seeing pockets of teams or subsets of an organization that are starting to use your product kind of independent of each other? And if you start seeing that, so for example, I can imagine, but I've heard interviews too, I've talked to folks at Figma, you know, they started seeing

• 29:18Speaker 00

  different pockets of a large organization start adopting Figma. This is very similar to Slack's story. And then once you see that, you're like, okay, maybe we don't have all the control and administration that a large company needs, but we have that organic demand that obviously people in large companies are getting value. And so that's where you start saying, well, let's go talk to the CIO at Uber. You already have 300 people using it. They paid for it with their corporate card. What if you went in enterprise level agreement? And then they'll tell you the 17 things that are missing and then that begins your enterprise journey.

• 29:47Harry Stebbings?

  How do you determine where to focus? It's so fucking hard because you've got SMBs wanting speed, new products. I want this iteration, that iteration. Your enterprises who want security often. As a product leader, which customer do you serve when you have two?

• 30:02Speaker 00

  Our answer that if we have to choose who we're serving above all else, it's actually the end user who's using the product and living in it for 10 hours a week. Because if we can make something that they love, that they feel like actually next to a working life, you know, simpler, more pleasant, more productive, they're not only going to tell their coworkers, they're going to tell their boss, they're going to tell their friends and they may even advocate for us. With IT. That's SMB. Not so. I would say actually it depends of SMB or enterprise because actually that's the magic of how Slack spreads at most large enterprises. So I think we've said this publicly, but over 85% of our enterprise customers, so customers of over 1,000 users, started in self-service. They started by signing up on slack.com, just like an SMB, sharing it with some coworkers, just like an SMB, putting a credit card down, just like an SMB.

• 30:47Speaker 00

  But at some point that grew enough that they then had a discussion with our sales team about becoming an enterprise level customer. Fundamentally, if we have to choose, we choose the end user who lives in cycle day. But I think we do it balancing what we know are the needs of the organization, the administration, the buyer who sometimes isn't the end user, obviously the larger company. But I think just fundamental to our DNA, build a product, customers love enough to tell their coworkers and their friends, and the rest will take care of itself.

• 31:15Harry Stebbings?

  How do you think about the product decision of short-term product feature improvements and shipping a little bit of what delivers revenue today in all honesty versus longer-term strategic bets, generative AI introductions, you name it? How do you think about that balance?

• 31:33Speaker 00

  Yeah, I mean, this goes back a little bit to the very beginning of the discussion, right? With the Google 70, 20, 10 model, which is like one way of thinking about it. Cause I think there, if you just applied that, you'd say 70% are the first category and 10% are the other latter category.

• 31:47Harry Stebbings?

  What do you think you've done that's in the 10%?

• 31:49Speaker 00

  In that 10% bucket would be the entire Slack platform and the kind of workflow automation system that exists on top of it in that ecosystem, which didn't exist when I first started Slack. Building network on top of this walled garden that existed with Slack Connect.

• 32:04Harry Stebbings?

  I think Slack Connect is so interesting. Did you disagree internally?

• 32:07Speaker 00

  We definitely debated the benefits would be outweighed by the complexity that it introduced. And also, you know, email is what we joke about. It's kind of like your post. It is the lowest common denominator. Everyone can be reached by email. So what would Slack Connect be 10x better than email for? And could you actually deliver on that? Because you do have the post mail that everyone can send. So yeah, we debated a lot. I think the AV stuff. Was another audio visual with huddles and clips and some of the newer coworking capabilities we've been introducing. I think for a long time, we thought Slack was a just primarily tech-based product and adding dimensionality to it, moving into a space that has a lot of other competition incumbents, I think that was a huge bet to take.

• 32:48Harry Stebbings?

  How long does it take for you to know a new product works?

• 32:52Speaker 00

  The thing that we do, which I think has become a playbook for us at least, we start off always with an internal prototype of something that's big, but we make it rough, we make it ugly, it's unpolished. And we see internally, before we do all the refining, the polishing, the scaling, the performance work, are people like, wow, this is super interesting. Like this could change the way that we work internally. Okay, if you do that with a small group, then okay, refine it a bit, scale that to the rest of the company. And then you start actually looking at data of like, what percentage of Slack is using it every day? What percentage of those people are using it the following week, in the week after? What's the depth of usage or those things growing? And then the big leap that we do long before we launch things is that we have a really just rabid customer base,

• 33:38Speaker 00

  many of whom want to pilot new features before they're ever out in the wild. And most of the new capabilities that we launch can't be experiments at a user level because they're social features, right? You can't sit in a huddle by yourself, you have to be able to use it with the rest of your organization. So we have this whole pilot or champion network that we use of companies of every customer segment size in basically every country in the world. And we do these progressive pilots where we'll roll out these new capabilities. We'll measure with surveys by talking to folks and also by looking at kind of adoption metrics as well. Refine the product and then incrementally increase the pilot. And usually by the time that we're launching something that's really big and new to hundreds in the world, we're pretty confident because of the diversity of pilot customers who hopefully became

• 34:22Speaker 00

  totally submitted with this new capability, that it's going to be something that all our customers love.

• 34:28Harry Stebbings?

  What you know as big and new that you thought passed all in time checks and then didn't hit with the public? And what did you learn?

• 34:37Speaker 00

  I want to say this must have been 2018, 2019. We had bought this company that was kind of lightweight wiki text editor, maybe almost like an early Notion-y, kind of prior before Notion existed, very early on. And we tried to incorporate it as a very basic WYSIWYG document composer in Slack. It was called Hosts. We were like, well, obviously you would need Hosts in Slack. It's like a message, but it's richer, more formatted, more visually appealing. For many reasons, this seemed obvious, and yet when we launched it, it really got very low adoption. It never really grew. And I think what we realized, actually this goes back to one of the things I think Google really instills in you, is that the most important feature by far for any product is actually speed above all else.

• 35:21Speaker 00

  That is like, that's like the oxygen for a good product experience. And it turns out to build like a rich documenting experience, if it takes a couple of seconds to load, if there's latency on the key presses, if moving objects around takes longer than you think, you're going to just say, well, I'm just going to go back to the thing I use every day in some other browser tab. I think it took us a long time, and obviously we've launched Canvas recently, or we're rolling it out actually now, and it's a different take on the same space. But I think what we realized was that the bar for just speed and quality for something that is as ubiquitous as a rich text editor was much higher than we're able to hit. And our customers, we could see in the data, like, they seemed obviously valuable, and at the end of the day, they didn't use it because they had good enough alternatives.

• 36:06Speaker 00

  And that was humbling, but valuable to learn.

• 36:09Harry Stebbings?

  I'm fascinated. You said there about that acquisition on Hosts. From a product leader mindset, who are a massive voice in acquisitions, especially so product-centric acquisitions, how do you think about that buy versus build? And respectfully, why buy when often the price is very high?

• 36:24Speaker 00

  To be honest, and this is maybe colored more by our experience at Slack, we've struggled to buy a product that we can then repurpose and incorporate that was faster at the end of the day than if we decided to build it from scratch. So the question is, well, why does that happen? Why do you do it? And I think the reason is often very simple, which is you look at your existing organization, and then you look at your ambition. And when your ambitions are outstripping your ability to scale your organization, you sometimes think, okay, this whole area that we're excited about, can we just buy a team here? Let's say you have the equity or the dollars to do it. And will that be a faster time to market to kind of realize some of the vision that otherwise we just don't have the capacity for?

• 37:04Speaker 00

  And I think it's just fundamentally very, very hard, especially if you're a hyperscaling kind of company, to be able to do acquisitions of products, not teams. Teams are very different. Like, we've done a lot of talent acquisitions and those have worked out amazingly well. But when we're trying to actually buy a product, I think you're right. I think, I think often the thing we've learned is that it doesn't actually decrease time to market. You might bring in expertise and that's a good reason to do it. But that's at least what we've learned for more small scale acquisitions. I think it's different if you're a large company buying like a multi-billion dollar business or Salesforce buying Slack. That's a very different proposition.

• 37:39Harry Stebbings?

  Do you think you're good at integration?

• 37:40Speaker 00

  I think we're good at talent integration. I think we struggle at product and technical integration. This is another reason, I think, if you look at a large company that's very acquisitive, like a Salesforce or Google, there's a reason why they have an entire M and A division that isn't just about purchasing the company. It's about actually integrating the company after. And again, that's where I think at least Slack, when we were still independent for most of our life, we didn't have the capacity, we weren't doing acquisitions regularly enough to build that muscle internally to become great at integration. It was always a one-off and it's hard to get great at one-offs.

• 38:14Harry Stebbings?

  Now, where are you weakest as a product leader? When you do a self-reflection, me and you, whiskey at the end of the year, where are you like, I really need to improve here and what are you gonna do to improve that?

• 38:24Speaker 00

  I think it's really hard to create a team environment within a product management organization where people feel really connected to each other, where they feel energized by each other's work, where they're pushing on each other, where they have a lot of shared contacts. Most product organizations, the PMs feel like their primary team is going to be the engineers and designers that they work with. Those are the people they work with every day that they have the closest connection to. And I'm still trying to figure this out, still trying to crack it. It's like, how do you make a PM organization feel less isolating? How do you create connection between? Between people who work on very different parts of the product to learn from each other, to push on each other, to build product with each other. I've yet to work at a place that has cracked this.

• 39:06Speaker 00

  And I think it's part of the reason why maybe there's so much hunger from PMs to kind of learn externally, to learn from podcasts, to learn from newsletters, to learn from other communities, because it is hard to foster that connection internally.

• 39:17Harry Stebbings?

  What about a North Star metric? And how do you think about effective North Star metric setting?

• 39:22Speaker 00

  Yeah, I think that that works. It's harder to do, I think, in enterprise and consumer because really the metric we care about is, are people willing to pay for the product? But most of the things you do in a product don't directly affect that metric. So it's a little bit harder than I think in the consumer world where it's more of a direct one to one. But I think most PMs are wired to really care about impact above all else. Like, that's the thing that motivates them, impact to the customer experience, to the product experience in the business. And North Star metrics help with that. It's kind of like head, not heart, where they think about what we're obsessed about. But it's different than how you feel connected and energized by your team.

• 39:57Harry Stebbings?

  What piece of conventional product wisdom do you think is BS?

• 40:00Speaker 00

  A lot. The biggest, I don't know if I would say mistake, but there's been like a professionalization of product management over the last five to 10 years. And by that, I mean coming up with what is a very rigid idea of what the role of the PM is and all the frameworks that you could be applying. And I think, no knock, I think it's great to have education inspiration and get that from many different areas. But I think a lot of PMs early on in their career, I think you look for all these kind of shortcuts that are like, okay, this is a framework I can use for a product strategy document, or this is a framework I should use for understanding customer feedback. To me, at least, it's much more of a combination of the arts and the science, the head and the heart,

• 40:41Speaker 00

  knowing things that are like frameworks, but not really applying them par plonche. I think where the role started, my impression, 20 years ago or so, is that kind of started by filling in the gaps in a product development organization where who was the single person who felt responsible for defining the why behind the product, the understanding of the customer, or strategy that's going to be longer term than the next release? And from an operational perspective, To kind of being the glue, being the multiplier for everybody else. And I think that's inherently very fluid definition. What I always kind of see with PMs is I think you have a very rigid definition of the role. I do this, the other roles do that. Where I apply this framework, this is the only framework I should use.

• 41:23Speaker 00

  I think you miss a lot of opportunity to have impact and even just enjoy the job. Take a more expansive view of what product can be. Take a less rigid view of copying and pasting frameworks and advice and figure out what the team needs, figure out what the customer needs and focus on that.

• 41:40Harry Stebbings?

  No, I could talk to you all day, but I do want to move into a quick fire round where I pepper you with questions again, but we just relatively time bound it. So it's not too dissimilar to the last half an hour. But I say a short statement, ready to rock and roll?

• 41:51Speaker 00

  Let's do it.

• 41:52Harry Stebbings?

  Okay, so when is the right time to hire a CPO?

• 41:56Speaker 00

  I would hire, first of all, a head of product before a CPO, because it's always better to start with fewer titles when you're small. And I think the right time to do it is when the product founder CEO, presuming that they are a product founder, feels like the teams are now locking on them for high-quality, fast decisions of direction. So when the development team starts slowing down because the CEO becomes the bottleneck, you probably need someone who can serve as a de facto head of product to start accelerating the organization.

• 42:25Harry Stebbings?

  What other function does the CPO have the most tensional conflict with?

• 42:30Speaker 00

  I think in a poorly functioning organization, I would say engineering or design. In a well-functioning organization, there should be your best friends. I think at an enterprise software company, the answer is always going to be sales, because what you promise to customers and what you can deliver to customers can often be at odds. At a consumer organization, it might be marketing, because that's that's kind of the same dynamic.

• 42:49Harry Stebbings?

  Which product leader outside of Slack do you most admire and why?

• 42:54Speaker 00

  Julie Zhao, who was, I think she was actually a designer by training, but was at Facebook for many years. I think she is the most incredible thinker and writer about product craft, product design, the nature of working between a product development team. I love everything she does. She's great.

• 43:12Harry Stebbings?

  Love that one. Is product more art or science? What's the ratio? And if you were to put a number On it, what would it be?

• 43:18Speaker 00

  On a really mature feature area, it's more science than art because it's about optimization. And you have so much data and scale. So there I would say 80, 20 science, art. If you're at an early stage company or at a really early part of the product development lifecycle for a new feature or new product capability, I would say it's like 75% art, 25% science. And you can be informed by the science and the measurement, but there's so much art in the curation, the editing, the creative approach, the vision that you have to focus on before you have enough scale to get into optimization mode.

• 43:51Harry Stebbings?

  What would be your biggest advice to a PM who wants a promotion today?

• 43:56Speaker 00

  Deliver impact, that's it. Everything else is an input. If you can change the trajectory of the customer experience or the business and you can point to what you did, that is the single most silver bullet you can have for a promotion process because everything else is an input to delivering that impact.

• 44:13Harry Stebbings?

  You can call up a CPO who's starting their role the next day. This is the night before your first day as CPO at Slack, yeah? And you can give them any advice, knowing what you know now. What do you call up that CPO to be and tell them?

• 44:26Speaker 00

  The most generic advice I would say is no one's looking to you to make any rash decisions. In fact, they actually want the opposite. In the beginning, what you should really focus on is understanding the team, the culture, the customers, the vision, the strategy. Focus on the things initially that are guaranteed accelerated to the velocity of the product development organization. Those are the lowest hanging fruit, the things that are about process execution and so on. Focus on team after that because I think you have to spend a lot of time to get to understand the team and the culture, where there's holes, where there's gaps. And then I would focus on the changing the fundamental strategy or mission the last, because that's the thing where you have to have a lot more context and understanding to be informed enough to have a good perspective.

• 45:11Harry Stebbings?

  You can be CPO of any product other than Slack. Which one would you be CPO of?

• 45:16Speaker 00

  Not that I would do it necessarily, but I think the most interesting company right now, that's an independent company, that's not like a behemoth, I would say is OpenAI. And that's, forgetting even the technology side of it, I think strategically for them right now, there's just an incredible exponential set of pathways in front of them. I know those would be really interesting questions to try to think through in an industry that is rapidly changing. So let's say OpenAI.

• 45:40Harry Stebbings?

  Forgetting all constraints, financial, hierarchical, legislative, if you could change anything about Slack, what would you change?

• 45:47Speaker 00

  I wish that there was a way to have both the bandwidth, but also the kind of divergent product experience where we could build for both the kind of like work adjacent and work use cases while at the same time, I think every year in our history, we've always just had that that wasn't something we could pursue and that market has become more saturated. But that would be my answer.

• 46:05Harry Stebbings?

  Penultimate one, navigating and building in the age of generative AI.

• 46:09Speaker 00

  It's a new space. It's changing really quickly. It reminds me a lot of the early days of mobile back in whatever it was, 2008, 2009, where you need to keep such a close pulse on how the fundamental technology and platforms are changing that are enabling new types of products to be built. And that changes every three months. Keeping a close pulse is one of the, I would call a product principle actually from Google back in the day as it relates to kind of ML and search related products that I still think is really relevant now, especially now, is this idea that I think the prominence and the promise that you put into the product experience needs to match the underlying quality and confidence in the data and in the model. And I think that's actually the thing that is hardest right now, the whole issue of hallucination and everything else, but these products are promising a lot to people.

• 46:54Speaker 00

  If you trust them blindly, you're going to get burnt. And I think figuring out how to have a little bit more of a symbiotic experience where you feel like this is an assistant, but not one that you should trust unconditionally and that it's more transparent with you of here's what I think, but you should check this or you should modify this or you should give me feedback on this. I think that's going to be really important for organizations if you're.

• 47:16Harry Stebbings?

  Now, a final one, what recent company product strategy did you look at and go, that was smart, like well done?

• 47:23Speaker 00

  OpenAI, realizing that this thing that was in research labs, this thing that was in a bunch of academic papers, which was the large language models, that the proof of concept, which isn't that much more sophisticated, you know, from a consumer experience, it's just actually making an input box and a text output box, that there was something magical there that would help people envision a world that was a very different model of interaction between people and machines, seeing them. Execute on that. And then obviously, I think, you know, Google has been playing catch up. And I think actually now for such a becoming doing a good job of that. The space is evolving very quickly, but I think OpenAI kind of realizing there's something magical here.

• 48:03Speaker 00

  And for now, the smallest possible product packaging of that to then give the consumer world a glimpse of what the future could look like. I think now the question now this space will evolve is, you know, between the consumer product side, which I think there's not going to be so many different variations that are going to win. And then you have in the platform and the infrastructure side. How do you enable every other company to integrate AI into their products? I think that's going to be the most interesting thing to play out over the next 12 to 18 months.

• 48:29Harry Stebbings?

  No, you've been unbelievably patient. I've just gone pow, pow, pow on peppery with questions. I love conversations like this, though, where it is totally natural. I've so enjoyed having you on. Thank you so much for doing it, and it really has been awesome.

• 48:42Speaker 00

  Thanks so much for having me, Haru. I had a great time.

• 48:46Harry Stebbings?

  I mean, what a fantastic guest to have on the show. If you want to see more from us behind the scenes, of course, you can by going to YouTube and searching for 20VC. But before we leave you today, this episode is brought to you by Linear. Let's be honest, the issue tracker you are using today, it's not very helpful. Well, Linear is different. It's incredibly fast, beautifully designed, and it comes with powerful workflows that streamline your entire product development process. From issue tracking all the way to managing product roadmaps. Linear is designed for the way modern software teams work. Its powerful workflows let you customize and automate processes to fit your team's unique needs. And they have integrations with your favorite tools like GitHub, Slack, and it takes just minutes to set up. Linear is the default tool of choice for high-performance teams.

• 49:30Harry Stebbings?

  From startups to a wide range of large, established companies like Vercel, Retool, and Cash App. So see for yourself what makes Linear magical. Visit linear.app/20vc to try for free with your team and get 25% off when you upgrade. That's linear.app/20vc. And speaking of tools we cannot live without, I need your input on the 20VC Miro board about new guests we should feature for 20VC in 2023. Just head on over to miro.com/20vc.

• 50:00Harry Stebbings?

  VC. Miro is a tool I consider to be truly game changing. It's a visual collaboration tool packed with the right tools, tech and templates to help you think of and create that dream product. That means you can brainstorm the perfect product with your team, vote on the best ones and explore with your customer journey roadmap, all on a Miro board. Whatever you need, Miro's infinite whiteboarding capabilities help you get there. It acts as your team's single source of truth, and now I'm using it to hear from you. Go ahead and add your suggestions for 20 VC guests on our Miro board at miro.com/20vc. And finally, you have to try EPO, the next generation A/B testing and feature management platform designed to help you run reliable impactful experiments.

• 50:44Harry Stebbings?

  EPO saves you time across the entire experimentation workflow. Planning tools help run experiment scenarios beforehand. Best-in-class diagnostics mean you spend less time debugging issues. Their cutting-edge stats engine helps you reduce experiment runtimes. Plus, EPO sits directly on top of your data warehouse and your North Star metrics. So it gives you this real confidence in experiment performance and ease in conducting follow-up deep dives. No more extended analysis cycles to understand results. And that's why companies like DraftKings, ClickUp, Momentive and Cameo all rely on EPO to power their experiments. Get access today at getepo.com. That's getepo.com. As always, I so appreciate your support and stay tuned for an amazing set of episodes next week.