Thinking beyond frameworks | Casey Winters (Pinterest, Eventbrite, Airbnb, Tinder, Reddit, Grubhub)

[00:00] every new person in the product team is acting like they work at Google and have these infinite resources and infinite time. [00:06] to make sure everything is perfect. [00:08] And there became such this focus on [00:11] the right way, [00:12] of doing product management that no one's taken any risk. [00:16] And I felt like, oh, am I responsible for this? Like I've created a bunch of frameworks on Reforge. You know, I'm onboarding these UPMs. Like, is this my fault? [00:25] But you know, you know, at Reforge, we're building frameworks that are tools in a toolkit. [00:31] You pull them out when relevant. [00:32] They're not a coloring book. [00:35] to stay inside the lines of. [00:41] Welcome to Lenny's podcast, where I interview world-class product leaders and growth experts to learn from their hard-won experiences building and growing today's most successful products. [00:51] Today, my guest is Casey Winters. Casey was one of the first ever guests on the podcast and the first to make a return appearance. He's a legend in the growth and product community, having worked with or advised companies like Pinterest, Reddit, Canva, Airbnb, Tinder, Thumbtack, Grubhub. [01:08] and many more, [01:09] He recently led product at Eventbrite for just about three years. [01:13] and recently returned to full-time advising and he's exploring new opportunities. [01:17] In our chat, we cover something he calls the Zero Interest Rate Phenomenon Product Manager, and how to avoid that. [01:23] What he's found works best when interviewing product managers, what impact he expects from GPT-4 on the role of product management,

[01:31] when to trust your instincts versus your team's insights and instincts. [01:35] What are all the different kinds of network effects and how do you effectively leverage them? [01:39] plus a great story about what Grubhub missed that led DoorDash and Uber eat their lunch. I always learn so much from chatting with Casey, and I am 100% sure you will learn a lot from this episode. With that, I bring you Casey Winters after a short word from our sponsors. [01:56] This episode is brought to you by Amplitude. If you're setting up your analytics stack but not using Amplitude, what are you doing? Anyone can sell you analytics, while Amplitude unlocks the power of your product and guides you every step of the way. Get the right data, ask the right questions, get the right answers, and make growth happen. To get started with Amplitude for free, visit Amplitude.com. Amplitude, power to your products. [02:22] This episode is brought to you by Eppo. Eppo is a next generation A/B testing platform built by Airbnb alums for modern growth teams. Companies like Netlify, Contentful and Cameo rely on Eppo to power their experiments. [02:37] Wherever you work, running experiments is increasingly essential, but there are no commercial tools that integrate with a modern grow team stack. This leads to wasted time building internal tools or trying to run your experiments through a clunky marketing tool. When I was at Airbnb, one of the things that I loved about our experimentation platform was being able to easily slice results by device, by country, and by user stage. Epo does all that and more, delivering results quickly,

[03:04] avoiding annoying prolonged analytic cycles, and helping you easily get to the root cause of any issue you discover. [03:10] Epo lets you go beyond basic click-through metrics and instead use your Northstar metrics like activation, retention, subscriptions, and payments. And Epo supports tests on the front end, the back end, email marketing, and even machine learning clients. Check out Epo at getepo.com. [03:28] eppo.com. [03:30] and 10x your experiment velocity. [03:33] Casey, welcome back to the podcast. [03:39] Thanks, Lenny. Great to be here. [03:41] You're the first ever return guest [03:43] to my podcast how does that feel i feel honored [03:48] Perhaps a bit unworthy? [03:50] But, you know, I'll go with it. [03:52] Well, you're both worthy and I'm honored as well. [03:55] So thanks for joining me again. [03:57] I have a lot of stuff that I want to chat about. [03:59] But first of all, I'm just curious, what are you up to these days? I know you left Eventbrite as CPO. I know you're doing a few advisorships with startups, but [04:06] How are you spending your days and what do you think is next for KC winners? So I'm still spending some time with Whatnot and Eventbrite as advisors. [04:14] I stepped back from the CPO of Eventbrite in October, but still working on some [04:20] long-term marketplace strategy stuff, some growth strategy stuff. [04:24] I'm on the board of a company called Beak, which is like Netflix for audio in Latin America. [04:30] And I'm doing some angel investing in marketplaces and what I call tech debt as a service, which is

[04:35] You know, what are startups that are building out things that were hard for my teams in the past to like build or maintain inside previous companies? [04:43] The other thing I'm working on is revamping the product strategy program for Reforge. So I got some things keeping me busy, but definitely not as busy as I was last year. [04:53] This is like a rare free agent Casey Winner's time. [04:56] Should people reach out if they're interested in maybe working with you? What's your advice for people listening? I always love seeing how I can help companies. So everyone should err on the side of reaching out and seeing if I have time. [05:08] to help if there's something they think I can help with. I just, [05:11] I love talking to startups. So yeah, absolutely. I can't promise anything, but... [05:16] Yeah, I just love talking to people about this stuff. [05:19] Awesome, and we'll point people to how to get in touch with you at the end of the episode, and I'll be in the show notes too. [05:23] But talking about the CPO role, so you're at Venbrate, your CPO, [05:27] It reminded me of a post you wrote about [05:29] how hard the CPO role is. And there's some quotes that I [05:33] recall in your post about how one if you ever ask a cpo chief product officer how they're doing [05:38] No one's ever going to say it. I'm crushing it right now. Yeah. [05:41] And then there's this other quote about how [05:44] You basically, as a CBO, just try to put some points on the board before you inevitably get fired. [05:48] Why is that? Why do you find that? And do you still believe that's a feature after? [05:52] leaving that role. Yeah, I would say I still believe [05:55] those things I said, it was funny. Um, [05:58] a product leader who will remain nameless. [06:01] emailed me when I hit year three at Eventbrite, and they said, congrats, you didn't get fired after two years. What's your secret? I think it's very hard to keep product market fit as a CPO within a company for a long time.

[06:14] The needs of the business shift over time. And of course, we all have different strengths and weaknesses in our game. Like no one's perfect at everything. [06:21] and [06:22] In general, as an executive, it's just impossible to do everything perfectly. [06:28] Any little misstep, mistake or something you just missed can blow up in a major way. [06:33] Frequently, CEOs have visions that can change and get misaligned with the product leader. [06:39] and also with leadership roles. [06:41] You don't get put on a pip, right? Like, [06:43] If a CEO loses confidence, it's over. [06:46] immediately. [06:47] So nothing like that happened at Eventbrite. I... [06:51] started to see that the areas of leverage for a product leader were just less in my wheelhouse over time. And I just talked to my boss about it. And, [06:59] you worked out something where I continue to advise on the things that I'm uniquely good at and find other people who are better at some of these other things that were [07:06] maybe more important at the time, [07:08] But... [07:09] Yeah, it's a really hard role, a really... [07:13] challenging role, you know, can be a lot of fun. And I, you know, I learned a lot from doing it. [07:20] But yeah, I still believe, and I think I've still yet to hear someone say they've really crushed it. [07:25] Is there anything that you learned from that experience to survive in that role for a [07:31] Three years, you know, I imagine... [07:33] You have a lot of skills and you're very valuable to the company, but I don't know any for someone that's in that role. Maybe right now, just like here's something that I did that maybe you should do. The first one I can think of is. [07:42] A big change from being a

[07:45] product leader to a CPO or something equivalent. [07:50] is you're actually a comp exec first and a product exec second. [07:55] So I think some mistakes I made early that I corrected successfully [08:00] is. [08:01] you want to show that you are caring and paying attention [08:05] to the overall business first before just taking care of your own product designers, product managers, researchers, whoever. So that's advice I found myself doling out to other product leaders a lot more now that I've [08:16] you know, gone through it, you have to, [08:18] not only truly care but create the perception that you care about sales about marketing about you know eagle etc [08:25] Because first and foremost, you're expected to leave the business. [08:29] at that level. So I think [08:31] That's something that I think is really important. And I think something else is... [08:37] really trying to diagnose [08:41] without as much bias as possible, where are the strengths and weaknesses of your team, where are the strengths and weaknesses of your product? [08:48] Lay that out with your peers. [08:50] and say, here's where we are. Here's where I think we need to go. [08:55] Here's the timeline on which I'm going to work to get us there. [08:58] because, [09:00] I think products such a confusing... [09:03] discipline for people who aren't in it like a sales leader many ceos you know marketing leader [09:09] that [09:11] They don't necessarily know what great looks like. [09:14] and the things they know about great leadership.

[09:16] they don't know if you know those things. So you have to make it really clear [09:21] Yeah, I know our OKRs are not as quantitative as they should be yet. That's because this team isn't ready for X, Y, and Z. [09:27] This is where we're going to get them to, but it's going to take some time. [09:30] So one of the things we did at Eventbrite that I thought was really helpful [09:33] is when you go public, you tend to rotate over your executive team and get public company execs versus startup execs. [09:39] So a few of us were new and it's just like, hey, [09:43] Let's do a deep dive on product. Let's do a deep dive on customer support. [09:47] Where is it at right now? [09:48] What do you feel like your job is? What do you feel like are the issues you're facing? Where do you want to take it over the next three to five years? [09:56] And now let's talk about it. [09:57] and just forcing you to write it all down. [09:59] is really helpful, but then it forces you to explain it. [10:03] Um, [10:03] And, and, and. [10:05] It helps the rest of the executive team get better company execs because [10:08] you know, now they really understand product a lot more deeply or understand finance a lot more deeply, whatever the function is. [10:15] Speaking of gaps and opportunities for [10:18] people in the company to level up, you've kind of [10:21] been highlighting this trend that I thought was really interesting. [10:25] something you call the zero interest rate phenomenon product manager. [10:30] Can you talk about what that is and what you've been noticing? Yeah, sure thing. [10:34] So, [10:35] This is something I started noticing while managing PMs and product designers. [10:40] at Eventbrite and it's now coming up with whatnot as I hire [10:44] help them hire new PMs.

[10:47] is the environment in which a lot of product managers have come up is very distorted from when you and I started doing this kind of work. [10:55] I think we used to always describe PMs as this gang of misfits. [11:00] We all started in different functions. We all had different strengths and weaknesses. [11:04] And the PM team, you know, gained as a whole from that diversity of skills and previous experiences. [11:12] So what I noticed is when we started doing [11:14] Product reviews at Eventbrite. [11:16] if there was any sort of like uncertainty in a problem or solution [11:21] The product manager, instead of shipping to learn, would talk about all this research they have to do. [11:26] to really learn the problems, really learn the solution. [11:29] You know, and my feedback would be like, no, like we [11:33] user research is a scarce resource. We have to reserve it [11:37] for the areas that have extreme uncertainty and high leverage for getting to certain [11:43] So if you're redesigning the login page for Eventbrite, [11:47] You don't need to use that resource to learn what to do. Just go see what Fang and the top unicorns did. [11:53] They probably did a good job. [11:55] And, you know, even if they didn't, [11:58] All of our customers also use those products, so they're going to be familiar with us copying any approach that they've taken. [12:05] So I think we sort of forgot in the industry that many times the fastest way to learn is to ship. [12:12] So then. [12:14] If you actually get them to skip research and just go look at competitors,

[12:17] Another thing I was noticing is I get like a list of options other companies have done. [12:22] But there would be no analysis of why those companies chose different options and what's most applicable. [12:29] you know, for us. [12:30] So it got me thinking like, okay, every new person in the product team is acting like they work at Google and have these infinite resources and infinite time. [12:39] to make sure everything is perfect. [12:41] And there became such this focus on [12:44] the right way, [12:46] of doing product management that no one's taken any risk. [12:49] And I felt like, oh, am I responsible for this? Like I've created a bunch of, you know, frameworks on Reforge. You know, I'm onboarding these UPMs like, is this is my fault? [12:58] But, you know, you know, and Reforge, we're building frameworks that are tools in a toolkit. [13:04] You pull them out when relevant. [13:06] They're not a coloring book. [13:08] to stay inside the lines of. [13:11] So, [13:11] At one point I got so fed up I wrote like an internal blog post and [13:15] I called it. [13:17] on best practices and breakfast rituals. [13:19] And I talked about how there were these articles about the morning habits of the most successful people. [13:24] And if you try to do all those things, it take you like six hours every morning. [13:29] And there was this article, I think it was on Business Insider or something about this futurist at Google. [13:34] And he talked about all the pills... [13:36] and food heats for breakfast every morning to try to live longer. [13:40] And the reporter estimated that cost him one million a year to have these habits. [13:45] So I was like, hey, I want to make it clear at Eventbrite, we don't have that much money.

[13:49] We don't have that much time. [13:51] There's no framework that allows you to not use your brain at this company. [13:56] and just follow some sort of process for success. [13:59] in this profession. [14:01] I adopted this thing into a public blog post that's on my blog now, [14:06] Now that I've had some time outside of Eventbrite, [14:08] I feel like I at least partially missed the mark on what's really going on. [14:12] I see this with like whatnot, because one of the things I'm helping them with is interviewing ICPMs again, which I haven't done in a while. [14:20] and [14:21] It's fascinating interviewing a PM or managers early in their career, because what not? It's a startup. [14:28] And you know better than most of us as a former founder of one. [14:32] startups typically require us to wear lots of hats you have to write sql you have to talk directly to customers you have to [14:38] prep marketing and sales. [14:39] And most importantly, you have to make a lot of decisions under uncertainty. [14:44] which you wouldn't necessarily expect the PM to do, say, at Google. [14:47] But it all boils down to... [14:50] using your brain in different ways. [14:53] you know, uh, [14:54] Nisa, who's on our legal team, she called Eventbrite a public startup. [14:57] because the pandemic basically [14:59] erased our business and we had to build it back from scratch. [15:03] So now that I'm doing these interviews, and whether they come from a small startup, a unicorn, or a public company, [15:09] They all sort of look the same. [15:11] because there's been so much funding to all these companies. Every company's been acting like they're Google with Google margins. [15:17] Meaning a lot of engineering support, a lot of design support, a lot of research support, lots of analysts around them.

[15:23] And actually seem pretty ill prepared for a real startup or even a public company with some uncertainty around it like Eventbrite. [15:30] So, [15:31] you start getting these weird responses in the interview process. You ask them to solve a problem. [15:35] And they'd say things like, [15:37] I can't even begin to come up with solutions until I see all the data and talk to customers. [15:41] And I'm like, yeah, I get that that's something you would normally do if you took the job, but like, [15:46] You don't have the data. You can't talk to customers. Make a decision now. [15:49] What would you do? [15:50] I want to see how creative you are. I want to see how much you're intuiting about the real problem and solution. [15:54] And they can't really answer. [15:56] And then, you know, my follow up, which I don't ask, but what I really want to ask is like, so when's the last time you used your brain? [16:02] versus follow the process someone else designing your company because like [16:06] I want the former, not the latter. [16:08] Wow, amazing. [16:11] There's a lot of things I want to dig into here. And I was going to ask why you call it zero interest rate phenomenon, but [16:15] I think what I'm hearing is it's, [16:17] When interest rates are zero, everyone's got a lot of money. Things are going great. Everyone looks like they're killing it successful. Everyone thinks they've got it all figured out. [16:24] And then when that goes away, it's like, oh, shit, maybe maybe not. And also, I think zero interest rates allowed every startup to operate like it was a public company with billions of dollars. Yeah. And, you know, that's going to go away over the next five years. [16:37] And it sort of takes us back to... [16:40] what product management used to be like, which was, you know, you didn't have all these resources around, you didn't have all this time to figure it out because [16:47] If you don't figure it out now and make it work, you may run out of funding, right? [16:51] Yeah, like a very concrete thing that's changed is a lot of research teams have been laid off because

[16:55] They grew really large and companies kind of found maybe we don't need the like of all the things we can cut. It's probably an area we can cut. [17:02] Yep, for sure. [17:04] There's probably PMs listening to this wondering, shoot, am I one of these? Am I just following a bunch of frameworks? What's your advice for someone that may fear that, oh, shoot, this is maybe who I am and I don't want to necessarily be this? [17:15] and what could they maybe do? I always advise going to companies where you can learn quickly and try things, right? [17:23] And I think it's certainly okay to learn a bunch of frameworks. There's a reason I've invested a ton of time in Reforge, because I think it actually really does help people. [17:31] to see how other people have built things in a way that they can [17:34] scale to their other companies. [17:36] But. [17:37] You have to... [17:38] understand that the job is not to follow the process. The job is not to [17:43] learn every framework possible. The job is to figure out how to add value to customers that translates into value to the business. And [17:49] And, you know, just reorient your North Star if you've gotten away from that. [17:53] when I joined Eventbrite, [17:55] you know, there was this team that [17:57] Didn't ship anything in a quarter. [17:59] And, you know, I went to the designer who I knew. [18:02] And, uh, [18:04] I know that he likes to [18:06] likes to get stuff out there and likes to get feedback. [18:09] And I was like, dude, you didn't ship anything. [18:11] Like, how could you have gone the entire quarter, [18:14] not shift any of your designs and felt that that was okay. And if you didn't feel it was okay, why just censor yourself? Come to me. [18:21] Tell me why this team is not letting you ship. [18:24] there's just no way you can get better as a designer there's no way you can have impact

[18:28] on the company there's no way you can have an impact on our users so [18:32] Um, [18:32] You know, those are some things that come to mind to that question. [18:36] So if you're a PM on a team that maybe isn't shipping, [18:39] maybe he's overusing research. Do you have any advice of just like when [18:43] it makes sense to invest in research. [18:45] Say you have the resources, is there kind of a rule of thumb you have of like, okay, let's actually spend the time on this for some time? [18:50] I think about it a little bit. [18:53] depending on the type of job you have and the types of customers you have. [18:57] So, you know, [18:59] the more scale you have, [19:03] tends to be you have less sophisticated customers because you're generally like a consumer thing where the customers [19:09] are rational, they're not experts. [19:11] And that's generally where you just try things. [19:15] You measure the impact they have. [19:17] And you only really bring in research when the impact seems confusing, right? [19:22] But you can run experiments, like you can get data really easily. [19:25] So bias towards getting stuff out there and seeing what [19:28] what users respond to. In consumer products. Right. If you're super enterprise, it's kind of the opposite, right? Each customer, first off, is sophisticated and tend to be rational. [19:37] They're paying you like six figures or more. [19:40] And you can talk to the customers and sales. She knows the customers really well directly. [19:44] and they tell you what they want, they tell you what they're willing to pay for, you build it, they pay you, bing, bang, boom. It's not obviously as simple as that, but... [19:53] you're basically doing the research directly with the customer. [19:57] and with the sales team and translating that to things that are strategic to the business.

[20:02] And I would say those are the two extremes probably most people are familiar with. [20:06] But we now have a lot of companies in the middle. [20:08] where they have more customers those customers are more sophisticated they're consumers they're maybe employees at companies or their small businesses [20:17] And that's where you have to be more nuanced. And this is, of course, where Eventbrite was on. Hey, when do I really lean in on research? When do I lean in on data? When do I pay attention to internal feedback? [20:27] So, [20:29] For Eventbrite, we really tried to focus research [20:31] on the B2B side of Eventbrite, where the problems seem big but not well-defined enough. [20:36] or that problem is well defined but we just really don't feel like we have the right solution yet. [20:42] And, you know, [20:44] if research is gonna be [20:46] a scarce resource inside the company, which I think it will in most places, [20:51] you have to figure out where it's a high leverage area, which means you have a really big problem, [20:55] We don't understand it well enough. [20:58] Or... [20:59] We know the problem is big, we understand it well, but we're just not sure if our solution is going to at any way. [21:04] Like, you know, hit the mark. [21:06] Makes me think about some of the best researchers I've worked with, and they often tell me, like, we don't need to do research on this. We have enough information. We have other things we should be doing. I love it when they say that, yes. [21:16] This would have been a really good segue to another topic I want to talk about, which is around gut instincts versus team expertise, which we're going to get to. [21:22] I have a couple more questions along these lines. Sure. You talked about interviewing PMs. [21:27] at Whatnot. Can you describe Whatnot just because you mentioned a couple of times just so people know what we're talking about? Sure. So Whatnot's a live streaming marketplace.

[21:34] mostly focus on the collectibles market. So [21:37] you know sellers whether it's like you know baseball cards or women's handbags or sneakers [21:42] You know, they'll go live on video and show you the products they have that they're selling. And people who are watching the stream can engage with the sellers and bid on, you know, different items. So kind of like Twitch meets eBay. [21:52] would be a good way to describe it. [21:54] Awesome. And I'll just mention I'm a tiny angel investor in that company. Oh, that's why you brought it up. Okay, I get it. [21:59] I just wanted to make sure people understood what this word was because it's like a [22:03] a fun word, whatnot. [22:04] But what I wanted to actually get to is you said you were interviewing a lot of product managers for them. What is your approach to interviewing PMs? What do you find is a really good signal for someone that's going to be successful? [22:14] I feel like the whole thing's gotten so performative. It's like, [22:18] interviewing is handing out Oscars based on who's prepared the best tell me about a time response. [22:23] versus actually assessing who can do the job we're hiring for. [22:27] So it's like most PMs are better PM interviewers than PMs now. [22:31] There's this quote from a movie called The Way of the Gun. I don't know if you've ever seen it. Nope. [22:35] But Benicio del Toro says in it, [22:38] It's like these days they want to be criminals more than they want to commit crime. And I think about that quote a lot when it comes to interviewing. So. [22:47] I would say I'm a bit cutrarian here in my approach. [22:50] I don't ask about your work history. [22:52] I don't care about your perfectly practiced answers. I'm going to give you real scenarios, [22:57] that I expect from the role, [22:59] I want to hear how you'd approach them. [23:01] And if you can't come up with a few reasonable ideas, figure out how to test them quickly without analyst support or research.

[23:08] I'm just not interested. [23:09] And I'm not saying this is a perfect approach. Interviewing is a very lossy format. [23:14] But the more I can see them do the job we're hiring for with questions, whether it's a presentation, a prompt, [23:21] That gets me the most comfortable that they know the job they're signing up for. [23:24] that they've shown in a practice scenario they can do it and they actually enjoyed it in some way. [23:29] Are there red flags you look for in these interviews? One is maybe said, "I can't even begin to answer this without research and data." Is there anything else? [23:36] Some things I pay attention to is if, [23:39] They're talking about... [23:40] solutions that are going to take a long time to get signal from. [23:44] where it's months and months of engineering time before you actually [23:48] you know, see the impact on users or the business. I think that's always great. [23:53] If they're not factoring in, [23:54] The amount of time it will take in general is not a good sign. [23:59] and not think [24:01] a bit more holistically about the types of metrics they expect to... [24:06] improve versus track to make sure like they haven't gone down. Those are some things that I pay a lot of attention to. [24:11] How much do you think they're intimidated by being interviewed by Casey Winters? [24:15] How much do you think that factors in? I think you overestimate my importance or awareness in the community. I think most people [24:22] I have no idea who I am when they start talking to me. [24:25] Well, I don't know. [24:27] Another thought is there's this meme that GPT-4 or 5 or 6 or 10 is going to replace product managers. And anytime some new feature comes out, people will put out these videos of, [24:36] Oh, look, they're doing all the PM's jobs.

[24:38] What's your take on the future of PM and AI? [24:41] if any. [24:42] I think if you thought the PM job was just filling in the latest Reforge or Shreyash framework, [24:49] and then getting that automatic saying promotion every year and a half. [24:52] then yeah, you're going to get replaced by AI. But I think the real PM job is the least likely to get replaced over time. [25:00] Because you need real subject matter expertise. You need to be trading off a lot of different types of things and making good decisions for the company and for your customers. [25:09] So in terms of using AI now as a PM, [25:13] I'd actually be cautious in... [25:16] like the current iteration of the cycle. [25:18] for PMs. It's a tool that's trained on sounding smart rather than always necessarily being smart. [25:26] I was at Eventbrite the other day and someone was telling me how they were loving the new Notion AI integration. [25:32] And she asked me if I used it. I said, no. She's like, oh, you totally need to. Hey, go ask your bio. It's really cool. And I said, all right, let's just do it on your screen. [25:40] So she did, and my bio said I started my career at Google and worked on Google Trends. [25:47] and a bunch of other products, none of which happens. Like, it's just complete nonsense. [25:52] It did have some stuff that did happen. It knew I worked at Eventbrite and it knew I worked at Pinterest and stuff. [25:57] But like half of it was just completely made up. It sounded plausible. [26:02] But it wasn't actually true. [26:05] I think where I'm more inclined for PMs to try to get leverage out of something like GPT-4,

[26:11] is a lot of the tedious work that's maybe not their specialty to begin with so [26:15] If you're not great at Excel or Google Sheets and you need to model something, [26:19] You can ask it how to format something. That'll probably be perfect. Or like how to get some Zapier integration to work. Like it probably knows how to do that really well. [26:27] So I think at this point, it's a pretty good no code, low code tool. [26:32] And it's obviously going to become a lot more than that. [26:35] But for a lot of these other use cases that people are touting, [26:38] there's a decent chance confidently tells you the wrong thing. [26:42] And that would scare me as a PM. [26:45] I've had the exact same experience. I had it describe what I've done and it was like 70% mistaken. [26:52] Yep. [26:53] Is there anything you've found value in yet with GPT of any kind in your work? Or is it still kind of just poking around in novelty for you? Yeah, I've had a couple of those things where it's like, [27:03] "Hey, how do I actually get this weird thing done?" [27:06] and it's been like [27:08] clear bullet points, take these steps. [27:11] And I'm like, oh, great. Like, this was actually really hard to find on Google. [27:15] So things like that I actually quite like it for. So that's where my head goes to like more confidently use it, because of course, I'll try it. And if it's wrong, [27:24] I'll learn where it's wrong. So it's low risk. [27:27] But it's been right on some of those things. And it's like, oh, cool. I couldn't figure out how to do that. [27:32] There's something on Google Docs. [27:33] that I need to figure out how to do that was kind of fancy. I asked ChatGPT. [27:37] It told me how to do it. And I was like, oh, great. I had a similar experience where I just needed a formula in Google Sheets. And I just told it, like, I need this thing. And it just gives me the actual formula. Yeah, that's so magical when it happens.

[27:47] Yeah, I don't know. I'm not that good with Sheets formulas. [27:51] So shifting course a little bit, [27:54] You also wrote this really interesting post on founder intuition versus team expertise, and there's a lot to it. [28:00] And I think it's a really interesting topic because it's this classic [28:03] discussion that startups have. How much should a founder [28:06] trust their gut and how much should it be top down [28:09] here's what we should be doing versus the team kind of bottom up. [28:12] figure out what to build. [28:14] And so I'd love to spend a little time here. So maybe just to start broadly, what is this? You kind of came up with a framework of how to think about when founder intuition [28:21] should overrule team expertise and how that changes over time. So maybe just talk about that. If founders are [28:28] Lucky slash good enough. [28:30] to find product market fit. [28:31] they've usually built up all this intuition about their customers about their product and their business [28:37] that are really hard to explain to others and may even be like subconscious [28:42] So when they start to hire people, [28:45] These people will come in with a lot of excitement, ideas, maybe even really relevant experience, especially senior leaders. [28:53] And I've seen some founders are like, cool, [28:56] I need to let these experts [28:58] start to own these areas and get out of the details. [29:02] And I think that's actually the opposite. [29:04] of what you want. [29:05] Because none of these people know the business as well as you get. [29:09] You actually need to direct them [29:11] until they show you they've really got it and are making better decisions than you would. [29:16] So this is not a founder example, but I remember

[29:19] when I hired Erica Warren to build out our loyalty program at Grubhub. She now leads growth at change.org. [29:25] And after a month or so, [29:27] She was like, dude, you're in all my meetings. I got this. [29:30] And that let me know she was in a different park. [29:34] of the situational leadership framework than I had put her in. [29:37] So I backed away and she crushed it from there. [29:40] founders will put a lot of new leaders in that delegate. [29:44] bucket too quickly. It's like, oh, it's theirs. They got it. They know more than me. [29:48] When founders actually know the right answer and just tell them. [29:51] And this can, of course, change, like in the case of Erica, you know, if you hire the right people, they do get smarter than you. [29:57] you can and should start delegating. [30:01] And many founders don't notice the signals where their founder intuition has been usurped by team expertise, or maybe it's just ebbed in effectiveness over time. [30:09] And that, of course, makes sense, right? Founders have all these different things they're dealing with, and your employees can really dive deep in certain areas. [30:16] So, [30:17] I encourage employees to proactively signal both directions like [30:21] "Hey, I need direction from you," or "Hey, I got it from here. Trust me on this one." [30:27] Interesting. So, [30:29] You're saying as a founder, [30:31] The heuristic should be if you feel confident about a decision. [30:36] generally you should be clear like hey i think i know what we need to be doing here and don't just like make people feel better necessarily by just saying okay you tell us what [30:44] You want to do, and then on the flip side is a [30:47] as an employee at the company, if you feel really confident about something.

[30:50] make it clear to the founder, like, I really think this is the right move. [30:53] Is there anything more you want to add there? [30:55] Yeah, well, I mean, there's obviously the feedback loop of that, of like, you know, [30:59] what happens when you do either of those things. [31:03] So, you know, depending on the founders response to that or the employees response to that, there's kind of different approaches. [31:10] you know, you can take. [31:12] And part of this [31:14] Yeah, this cool chart of just like over time. [31:16] the founder expertise becomes less and less relevant. And I guess is that because they just don't spend as much time with the customers because they have other stuff going on or they hire people that are smarter? [31:24] And then over time, team expertise goes up. If you're lucky enough to scale a company, there's just more and more things going on. [31:30] that, [31:31] All will reach the founder. [31:33] in some way, but it means more breadth and less depth on any particular issue. [31:38] And the reverse is true for people you hire, right? They're able to get really deep into things that maybe you were really deep into two years ago. [31:46] um but you know you just can't you can't stay deep in anymore so [31:51] I built this chart on [31:54] or I guess I should say table. [31:57] on the different [31:58] phases of a startup, right? And when you're finding product market fit, like, [32:03] everything goes through the founders. You cannot outsource that. [32:06] Um, [32:07] And then, you know, when you start scaling the company, because you found product market fit, it's the first time that founders run into this, like, [32:14] classic problem of [32:16] We got you here, won't get you there. [32:18] What got you to product market fit was iterating on product and doing things that don't scale.

[32:22] Guess what? You found a product that works. Don't do that anymore. Make it scale. Don't come up with new products. You found the one that works. [32:28] The reason I came up with this framework originally is, [32:34] When I was at Grubhub, we were scaling pretty nicely and fairly organically. [32:40] Mike and Matt, the founders, had intuited these phases of building a company pretty well. [32:45] But we acquired a competitor and I saw how that [32:49] company we acquired operated. [32:51] And even though they were largely in the same phase as us, they still operated like they were founding the company, like everything was going through the founder. [33:00] And intuitively, I was like, oh, this is why we're acquiring you and not the other way around. [33:08] So I think it's really hard. [33:11] for founders to get those signals organically. [33:14] And I think it's up to us as employees to help. And, you know, the founders can listen to the signals or not listen to those signals, but. [33:20] That's part of why we're here. But we should also give those signals when we're confident we really get it. [33:25] not like coming day one being, I know what your sales strategy should be. You're doing it all wrong. [33:29] Chances are the founders weren't doing it all wrong. They knew something you don't know about yet. [33:33] what is it that you think made it such that you won and they didn't is it because they didn't invest [33:38] They didn't shift to scaling and delegating to their employees as much as the founder just telling everyone. Correct. Right. Like the... [33:44] Everything was not moving as quickly because everything had to go through the same person to get done. [33:50] So it just allowed them to not launch as many markets, not sign up as many restaurants to not figure out as many.

[33:56] growth channels as we hit that. [33:59] Got it. So they just bottlenecked themselves and they weren't able to move as fast. [34:03] because the founder thought they needed [34:04] to make these decisions and they were still the ones that knew everything. [34:08] Yep. [34:08] That makes me think about most companies. [34:11] maybe not most, but many founders [34:13] Just like... [34:14] believe they have the answers like they are confident they know what to do right and it's rare they get to a place of like no i actually don't well i would say [34:23] No one gets to it super intuitively. [34:26] Mmm. [34:27] Any advice for either as an employee at a company with a founder like that or leader? [34:32] I've just had a [34:33] push back and help the founder understand maybe you should let go and trust people to make decisions? I think the first thing is you have to understand, [34:41] That it's their company, not yours. [34:44] and that founders have impossible jobs. [34:47] They're not going to scale perfectly with the needs of the company. It's just something's going to be... [34:53] off. [34:54] And it may be because there's a whole lot of hubris by being the one of a thousand startups that made it. [35:00] or they just have personal styles [35:04] But they can run the company any way they want. [35:08] And I think it's sometimes surprising that, [35:10] when founders actually [35:12] want to run things suboptimally? [35:14] Like, I think we've seen examples where founders just want to build cool shit versus focusing on what customers want. [35:20] or they're too obsessed with the design. They're always rebranding, redesigning things in ways that confuse customers. [35:25] And, you know, of course, that's going to happen. And those are some of the worst cases.

[35:28] um [35:30] But I think, you know, if you do state your point, [35:33] You've actually built up the expertise, you're confident, and you are refuted for whatever reasons. [35:39] You have to find ways to be contrarian and prove you're right. [35:42] And different founders have different forms of feedback they listen to. [35:47] It might be talking to other CEOs. [35:50] or other operators or trusted advisors, certain customers. Some might be very data-oriented, so you just run the experiment and ask for... [35:59] you know, forgiveness after you prove it works, whatever. Like everyone's got like a different thing they respond to. [36:04] So you try to figure out what actually... [36:07] influences the CEO. [36:10] and build that case. [36:11] whether it's through an experiment, the right customer intro, the right external advisor, [36:15] And if they rebuke that as well, then you got to disagree and commit. [36:19] or find a new company to join or start your own. Like I said, it's their company. They can make the call. [36:24] I remember when I joined Pinterest, [36:28] I was just coming off, you know, 100x increase in user growth from Grubhub. [36:32] And I was ready to drive the same sort of impact at Pinterest. [36:36] But, [36:37] no one knew who i was i didn't come from facebook or google like most of the other people are pinterest [36:42] No one gave a shit about Chicago startups at the time. [36:45] So, you know, when they didn't trust [36:47] My proposals. [36:49] I just went and talked to the heads of growth at all the startups they did respect. I talked to Dropbox. I talked to Facebook. [36:55] And when they said all the same things I was saying, [36:58] But I said that they said it, not me.

[37:01] That was sadly more convincing. [37:03] But I didn't care because I still got what I wanted out of it, which is to do the right thing to grow the company. [37:09] And now you're one of those people that people go to. If Casey says it, it's probably the right way to do it. Whether I'm correct or not, it is a role I sometimes play in the ecosystem, I will admit. Is there an example where you were convinced something was the right way to go in the founder at one of the companies you worked at? [37:22] had a different opinion and they were they ended up being right. [37:25] When I was at Grubhub, one of the ideas, one of the founders came up with, was to build an app for delivery drivers. So at the time, [37:32] you know, [37:33] Restaurants had their own delivery drivers. They did not work for us. [37:36] And we wanted to build an app for the delivery driver so you could see where the food was along the way. So you could know if it's five minutes away, whatever. Domino's was the first to have done this at the time. [37:46] And we thought that would make a bunch of sense. [37:48] And I was just like, [37:49] how am I supposed to convince the delivery driver who doesn't have a relationship with us to download the SAT from Grubhub? [37:55] To actually turn it on, [37:57] to create anxiety for the consumer of like, they maybe didn't take the right turn or whatever. I was like, I don't think this makes sense. No one's going to use it. [38:04] And, you know, Mike and Matt like overruled me and pushed us to do it. [38:09] And. [38:10] In the long run, it ended up being quite necessary because of the innovations around DoorDash and Postmates and later Uber Eats into having their own delivery network and us needing to counter that. [38:20] That's, of course, a key piece of technology that needs to have parity. [38:24] with those new services. So that was an area where perhaps they were thinking a little bit more longer term than I was. [38:30] And, you know,

[38:31] I was not able to kind of get where they were going on the longer term time horizon of how this would be. [38:36] actually ultimately adopted in youth school. [38:38] This is an awesome segue to the next area. But real quick, how did they get drivers to start using this app? Was there anything really clever they did? [38:45] I think it was... [38:47] fairly low adoption for a while and leveraging basically restaurant, [38:53] Authority. [38:54] So if we were able to push on our restaurant customers effectively to say like, hey, [38:59] if you use this. [39:00] It'll get you more orders. We'll get you prominence in the UI, things like that, to then push down. [39:06] you know, to their direct relationships with the drivers to get them to adopt. [39:36] plus get advice on how to fix those errors. You can have it automatically browse your website's internal and external links and get actionable insights from your backlink profiles. And you can learn what keywords your website ranks for and see how you stack up against your competitors. Visit hrefs.com slash awt and start improving your website's visibility. [39:56] That's hrefs.com slash A-W-T. [39:59] Okay, so shifting to our last topic that I was excited to chat about.

[40:04] is around network effects and marketplaces and SaaS and kind of how those things connect. [40:09] I feel like you have the clearest way of thinking about and explaining network effects. So maybe [40:14] just to start can you just like simply explain what is the network effect and then [40:18] the different types of network effects that exist. Yeah, no pressure after you say that, right? Yeah, so at its core, a network effect is when [40:26] a product or a business gets better with more users or customers using it. [40:31] And there's three types that I tend to focus on. [40:35] So the first and probably the most well-known is called direct network effects. [40:40] And that's when [40:41] every additional user makes the product better for all the existing users. So when someone joins WhatsApp, [40:48] other people can talk to them that they that wouldn't be able to talk to them before. And that makes WhatsApp valuable, more valuable for everyone. [40:55] And then there are cross-site network effects, [40:57] where there's two distinct types of users, [41:00] And adding an additional user on one side of the network makes it more valuable for all the [41:04] types of users on the other side of the network and vice versa. [41:07] So when a restaurant joins Grubhub, [41:10] It creates more selection for users to order food. [41:13] And when more users start ordering more food, Grubhub becomes more attractive for restaurants to join so they can make more money from delivery orders. [41:21] The last network effect I focus on is data network effects. [41:24] And that's when the quality or cost of something [41:27] improves as more data is collected through the product. [41:31] So when you save content to Pinterest,

[41:34] to a board, [41:36] It gives Pinterest signal on the quality of that content, [41:40] its relevance to other pieces of content since they're [41:42] shared to the same board. [41:44] and more signal on your preferences. [41:46] And this allows Pinterest to better recommend more content to you as well as to other people who look like you and share similar interests. [41:53] So, uh, [41:55] One thing that's [41:56] Tricky about. [41:58] Peace, Network Effects. [41:59] is most people talk about network effects in the context of social network, not marketplaces or SaaS, like my background. [42:05] And the thing that's really tricky in understanding social networks is they're traditionally described as these direct network effect businesses. [42:13] But all of them. [42:15] have to become [42:16] either cross-eyed, [42:18] and or data and network affect businesses over time. [42:21] The only companies that stay direct network effect businesses are these pure communication tools like Messenger, iMessage, WhatsApp. [42:28] And there isn't a clear path to make money with those, but you want to sell ads, you're becoming a cross-site network effect business. And also creators versus consumers creates a cross-site network effect business. [42:38] If you want to get better at personalization, [42:40] Guess what? [42:41] You're recommending content by leveraging data network effects. [42:44] I love that. You nailed it. [42:46] I imagine marketplace founders listening to this may be wondering, hey, how do I, can I add on these additional network effects? Can I evolve towards one of these? And what you're saying is not only can you, you need to eventually. [42:58] I agree, right? I think [42:59] you know they're a great form of defensibility first off so [43:03] Do they make sense for every business? No. But in a lot of cases, in order for you to evolve properly and continue to grow,

[43:08] they can become a necessity. [43:11] Are there any other examples of marketplace companies that come to mind that did a great job? [43:15] evolving or adding one of these network effects or just like interesting [43:18] So let's take some inventory, right? [43:21] Um... [43:22] So I think Zillow is really interesting. [43:25] in that [43:26] a real estate listing site is not [43:28] particularly an interesting product [43:31] But, [43:32] The ability of saying, very rarely do both sides of this network, people who are buying homes and selling homes, need to connect. [43:39] But I can find a way for the brand to connect with both of them by giving real time pricing information on the value of homes. [43:47] And that creates a much stickier product for both sides. I think that was a [43:51] a pretty genuine innovation that other people have tried and rarely been. [43:56] successful with. So that's why I [43:59] that [43:59] definitely sticks with me off the top of my head. [44:04] Great. [44:05] We've been talking about Grubhub a lot, and Grubhub had some pretty strong network effects. Yep. [44:09] and large part things of the work you did, [44:11] And one of the benefits of network effects is it creates a barrier to entry. It's hard to replicate and compete with a company where they already have all this [44:19] network effect. [44:20] But, [44:21] Famously, Grubhub got disrupted by DoorDash and Ubreads. [44:24] And I need to talk about how that happened and why that happened. And it'd be cool to hear just like, what do you think? [44:29] they could have done differently in what happened where their network effect got [44:32] Eating for lunch, pun intended. [44:34] Sure. It's a great question. And I think people do mistake any form of network effect.

[44:39] as this perfect form of defensibility. [44:42] They're the best form of defense ability, but that doesn't mean they're immune to disruption. [44:47] And I think the main way this happens with cross-side network effects that we typically talk about for marketplaces, [44:52] is when a disruptor dramatically expands selection. [44:56] So I want to be clear. I left Grubhub at the end of 2013 before said lunch was eaten. [45:02] So what I'm communicating is more like public knowledge of watching these amazing companies compete rather than like being inside the fire at the time. [45:11] But one important thing to remember is Grubhub was an asset light marketplace model. [45:16] Restaurants did their own delivery through their own delivery drivers that they hired. [45:20] And I wrote this post for Andreessen Horowitz a few years ago. [45:24] about supply strategies and marketplaces. [45:27] So in that, like, you really need to be comprehensive or have exclusive inventory. And [45:31] Grubhub strive to be comprehensive. [45:34] We show every restaurant that delivered to you [45:37] even if we didn't work with them directly for [45:39] online ordering. [45:41] So, [45:42] What happened about a year before I left Grubhub was that these competitors started to raise their seed rounds or series A's. [45:48] to build delivery networks. [45:51] Postmates and DoorDash were the [45:53] the first thing. [45:54] Two, that rose to prominence. [45:56] Uber Eats would come later, but they actually started with a... [45:58] pretty different business model. [46:00] and [46:01] This is an extremely different business. You're hiring drivers. [46:04] You're managing logistics. [46:06] So we call this a heavily managed marketplace model.

[46:10] as you are now facilitating the transaction versus just connecting buyers and sellers [46:14] and taking payments like Grubhub did. [46:17] So on the one hand, you have Grubhub, this extremely high margin business, [46:22] And these businesses coming up are actually negative margin businesses. At least they were for a long time. [46:28] So in 2013, Grubhub acquired Seamless, its closest competitor at the time. [46:34] And Seamless still had a lead on Grubhub in New York, [46:37] partially because it had this corporate program. [46:40] where law firms and banks and consultants got lunch stipends, you had to redeem by ordering the food through Seamless. [46:48] So Grubhub operated in these dense cities and suburbs like New York. [46:54] Because those were the only places that had restaurants that had delivery. [46:57] DoorDash, when it started, operated in less dense suburbs. [47:01] and worked with restaurants that had never done delivery before and they would provide you know the delivery drivers themselves. [47:07] So this allowed DoorDash. [47:09] to grow without much competition in the early days. [47:12] And these companies also took some real gambles. [47:16] One of the things they did is they delivered from restaurants they didn't have an agreement with. [47:20] And that caused controversy and lawsuits. [47:24] But what it meant is when these companies did launch in the cities, [47:28] to go after Grubhub, [47:29] It felt like DoorDash and Postmates and Uber had dramatically larger selection of restaurants to choose from. [47:35] And Grubhub was definitely no longer near comprehensive. [47:39] And, you know, back when we talked about cross-site network effects, the more selection of restaurants, the more attractive it is to users and vice versa.

[47:46] So at that point, [47:47] Grubhub's put in a very peculiar position. [47:50] It had gone public after raising $80 million in venture capital and raising $100 million in an IPO. [47:57] telling investors it's an extremely profitable [48:00] high growth asset light marketplace. [48:03] but now it's got disruption pressure from operationally intensive negative margin delivery services, [48:09] that were raising $100 million every six months from the private markets. [48:14] So what Grubhub assumed is that these businesses were just structurally unprofitable. [48:19] and that VCs would stop subsidizing them eventually. [48:23] Of course, that didn't happen. They kept raising more money. DoorDash eventually raised, you know, [48:28] multiple billions of dollars in this market. [48:31] And part of what they use that money for. [48:34] was to lock up agreements with national chains, [48:37] which Grubhub never worked with. It was always like the local mom and pops. [48:40] and promotions and discounts on the demand side. [48:43] So all of this is happening and then the pandemic hits. [48:46] And all of those negative margins turn positive for Gordash for the first time, if I understand things correctly. [48:53] And on top of that, all this corporate ordering that we got from Seamless no longer mattered because no one goes to the office anymore. [48:59] So this gambit really paid off and DoorDash just took the market. [49:04] Grubhub then tried to copy the approaches of DoorDash and Postmates, and they wrote kind of this famous letter saying that they still thought the approaches were stupid, but they were... [49:12] Their hand was forced. [49:13] which is kind of funny to read in retrospect. [49:15] um

[49:16] So, you know, a lot of armchair thinkers are like, how could Grubhuff have been so stupid to let [49:22] DoorDash take the market that they built. [49:24] And look, I'll admit, Grubhub made some mistakes. [49:27] But I think it would have been extremely hard [49:30] for Grubhub to come out on top here. [49:32] So let's say, [49:33] In 2014, after you IPO, [49:36] you do actually think what DoorDash, Postmates, and Uber are doing are smart. [49:41] What do you do? [49:42] Go to investors that you promised high growth and high profits and say you need to raise more money in debt to compete with negative margin upstarts. [49:50] in a way that will destroy all that potential profit you mentioned, the stock would drop 90%. [49:56] A lot of your employees would leave because their stock is now worthless. [50:00] And on top of that, [50:01] Building a delivery network is a heavy operations play that matches none of your core competencies. [50:07] So to me, that feels like a death sentence. [50:09] So I think the only play here... [50:12] was to buy DoorDash as early as possible. [50:15] and let doordash and their operationally heavy culture [50:18] Eat Grubhub from the inside out. [50:21] I think Grubbump probably could have acquired them multiple times, [50:25] for whatever reason it ever happened. [50:27] It would have seemed pretty risky had Grubhub done that. [50:30] investors probably would have hated it. [50:32] But it would have been their Netflix moment where, you know, Netflix bet it all on streaming and they bet right. [50:38] And this is all... [50:39] incredibly easy to say in hindsight. [50:42] But I think what I've learned now that I'm more senior in my career, [50:46] is

[50:47] during existential threats. [50:49] You know, it's like when Nassim Taleb says the only rational reaction is overreaction. [50:55] Mm. [50:55] Unless you have a real viable reason to assume otherwise, you've got to assume the disruptor is right. [51:01] and base your strategy on them playing an optimal game. [51:05] Whereas I think Grubhob assumed the Disruptor was wrong, [51:09] and that it would all play out eventually in their favor, and it clearly did. [51:14] Wow. [51:15] I have never heard that full story. That was amazing. Thank you for getting into so much detail there. [51:19] I don't know how much of it is right. [51:21] That seems right from the outside. [51:23] And it's interesting that it's like a version of Innovator's Dilemma. [51:26] Usually, innovators dilemma is people come and do something cheaper at the low end of the market. [51:31] And in this case, it was more they were doing something that they didn't believe with scale. It was just much more inefficient, even though it was a better experience. [51:38] And is that [51:39] Is that Innovator's Dilemma or is that some other version of Innovator's Dilemma? Because to your point, they couldn't almost go after it. [51:45] I think it is in spirit for sure. The takeaway there is, [51:49] that you've [51:51] learned is [51:52] Don't underestimate someone. [51:54] that's trying to do something [51:55] that's going to eat your lunch eventually, even though it feels like a terrible business model. [51:59] Right. Like if the market's rewarding it, the market will probably find a way to make it profitable. [52:04] And, you know, it's essentially when everyone's find like a more efficient network effect for demand. [52:11] on the marketplace side, [52:13] Thank you. [52:14] You probably need to copy that as soon as possible.

[52:17] And if you can't copy it, own it, right? Like buy it. And I think we've seen this a few times, like Rover and Wag was another interesting example of this. Um, [52:26] where they found a more frequent use case that fit the same supply and demand. [52:29] And, you know, Rover just copied it as soon as as soon as possible. And that ended up working out for them. [52:35] in a way that it didn't as much for Grubhub. [52:37] I just did a talk recently on marketplace growth strategy, and someone asked me a really interesting question. If you're trying to go up against the DoorDash, [52:45] or an Uber today. [52:47] what would you do? And my answer is just like money was so freely available at that point. [52:51] that they were able to [52:53] Like a lot of this was spent. They had so much money raised. You said they're raising $100 million every six months. [52:58] like is it even possible in today's climate where you can't actually raise that much money to [53:02] have any sort of chance to eat at DoorDash or Uber. [53:05] Yeah, I mean, clearly they took advantage of the zero interest rate environments that we talked about earlier. That is not a replicable strategy in 2023. So any approach to compete with them. [53:17] has to be competing on a very different angle because they can spin you into the dust. [53:24] So, you know, we'll see if anyone's able to build [53:26] a way that creates new supply in market. I think you and I have seen a couple of earlier stage examples of that. We'll see if any of them scale for something that feels [53:35] cheaper or more fun on the demand side. That would be like another way to potentially disrupt things. [53:40] I think another thing that was working pretty well to... [53:42] before the pandemic, [53:44] in building disruptors was focusing on the pickup use case. [53:47] which is the most important thing.

[53:47] you know, [53:48] DoorDash cannot do as easily. They've tried, right? But their whole value prop is the delivery network, which you don't need for pickup. [53:55] But of course, the pandemic put a lot of those startups grinding to a halt. [53:59] as well. So, you know, there's always angles, but you're going to need to be incredibly clever because, [54:04] just attacking with a war chest is not really going to work. [54:09] Yeah, I've noticed a lot of startups launching that are trying to do white label delivery for restaurants so they don't have to really rely on DoorDash. [54:16] But... [54:17] Maybe that's a wedge to get in and eventually they do DoorDash sort of thing. Yeah, and I would say I'm pretty skeptical of those because they're just not really building. [54:24] the network effects on the demand side. They're just more like SaaS businesses. And [54:28] I think they'll struggle to be multi-hundred million dollar in revenue businesses. Because the take rate is typically a third or a fourth of what a DoorDash can charge. [54:39] Perfect segue to my next question. I only have a couple more. [54:43] There's this concept and I think pull from marketplace founders to add a SaaS tool eventually and have both business models. And then there's often the reverse where a SaaS company wants to add a marketplace. [54:57] And I know this question is something you get a lot. [54:59] And so just a question for you. [55:01] Does this work? Are there examples where [55:04] they've added this additional way of making money and which direction do you find is most often successful. [55:10] and not successful. [55:11] The canonical example that a lot of these founders use is OpenTable. [55:16] That always felt [55:17] really unfulfilling to me because it's very old at this point. Probably the younger people listening to the podcast are like, what's OpenTable? And I think that business has incredibly underperformed the market it operated in. So it had a good exit. It sold for $2.6 billion, I think, to Booking.com.

[55:36] But Booking subsequently wrote that value down below a billion. So that's not really the outcome I think we're looking for traditionally in marketplaces. And maybe explain what it was initially, the SaaS tool versus the virtual marketplace. Oh, sure. So OpenTable promised... [55:49] Both a SaaS tool that would help you understand like what tables are open and where to seat people. So for like the host in the front of the house in a restaurant. [55:57] but also bring in additional reservations. [56:00] to fill up your restaurant right because [56:03] I think what [56:03] people misunderstand about [56:06] restaurants are these incredibly low margin businesses and it's like yeah they are [56:11] because they're paying the rent no matter what happens. [56:14] So once you're already paying the rent no matter what happens, you're incentivized to pump as much food out the kitchen, [56:20] and to get as many people seated in house as possible. [56:23] So people would say like, oh, Grubhub and DoorDash, you're charging too much. [56:27] And it's like, no, [56:29] Like delivery orders and catering orders are basically pure margin to restaurants. [56:33] The chefs are already there. [56:34] the rent is already being paid, [56:37] The more things we pump out of the kitchen, the more money we make. [56:41] which is why restaurants are willing to pay higher fees. These restaurants are not stupid. They're not being like, [56:47] misinformed right so open table well before there were any delivery startups was like [56:53] We're going to fill up your tables and we're going to give you the software that helps you manage your tables most effectively to make the most profit. [57:01] Got it. [57:01] Okay, great. So keep going. So that's OpenTable. So look, this is definitely the strategy we're working on with Eventbrite.

[57:08] where we started with something that was more SaaS-like, [57:10] enabling easy payments and certain tools to make event creators' business more efficient. [57:17] And we're layering in more of the traditional marketplace value prop of demand, you know, driving more sales for these event creators. And, you know, I would say it's got a ways to go to be working really well. And [57:29] What does working really well look like? [57:31] It's when the marketplace model unlocks those cross-side network effects that make it easier to grow. [57:36] So, you know, [57:38] Event creators, when Eventbrite started, they would just go do their own marketing to bring people to Eventbrite to transact. [57:43] on their events that they had listed there. [57:45] And now Eventbrite drives 25% of the ticket sales itself. [57:49] And that percentage is growing faster. That's amazing. Yeah, it's growing faster than the rest of the business, which is great. [57:55] But it probably needs to be closer to like 50% to unlock those real cross side network effects. [58:00] And that would mean that creators start selecting Eventbrite, [58:03] because of all the demand of Empright has versus the quality of the SaaS tools. [58:07] So that's obviously something we're working on. Fair is a company. [58:12] I advised more recently that I think is [58:14] unlock this model in like the reverse. [58:16] where [58:18] Fair... [58:19] is a wholesale marketplace between independent retailers and brands so that you can sell products at these retailers, like a boutique. [58:26] in your local neighborhood. [58:28] So Farah built the Marx place first. [58:30] and then it later built a sas tool that allowed the brands to onboard their current [58:35] you know, retailers into the platform and get better terms.

[58:39] on payments and free returns. [58:41] And it didn't charge anything to use... [58:45] the SaaS tool for any rebookings that happen through that platform. [58:48] But what that did is it onboarded more independently. [58:51] retailers that they could cross-sell to other brands. [58:54] without needing to use a sales team. [58:56] And when they cross sold, of course, Thayer did. [58:58] you know, take their percentage on those transactions. [59:01] So I think that model's worked out really well and FAIR has grown quite quickly. [59:05] And so the examples you gave, Eventbrite was a SaaS tool trying to add a marketplace. Yep. Fair is an example of a marketplace adding a SaaS tool. [59:14] And is the takeaway there that it's in your experience, it's generally more [59:18] You're going to have a better time if you're [59:20] a marketplace adding a SaaS tool versus the other way around. [59:23] Yeah, I think that's right. So I think [59:26] I don't know when it was, it was maybe like 2015 or whatever, a bunch of investors started writing about it. [59:31] SaaS to marketplace transitions, [59:34] And like, that's the transition of Empirite's going through. It's going to work. [59:37] but it's hard and it's not replicable for a lot of other SaaS businesses. [59:42] You need to have a direct relationship with your customers' customers, first off. [59:46] Those customers need to have needs beyond that single supplier. [59:49] Usually it's like a discovery value prop. [59:52] you know, for other suppliers. [59:54] And you need to build like this totally new skill set [59:57] to build tools for your customer's customer, who is very different from your current customer. [1:00:03] Whereas, you know, marketplaces that add a SaaS component, [1:00:06] It's a customer acquisition tool like there.

[1:00:09] or a workflow or retention tool [1:00:12] to increase retention or reduce disintermediation. [1:00:15] I think you'll see that be a lot more common approach and much more replicable. [1:00:20] Is this just rooted in the fact that marketplaces are incredibly hard no matter whether you start with them or whether you add them? [1:00:26] Is that the root of the issue here? I think that's a good way to frame it. So if you're building a startup in 2023 and you're like, oh, I'm going to build the SaaS business. [1:00:34] And then five years later, I'm going to build this marketplace business on top of that. [1:00:38] It's kind of like, [1:00:39] It's going to take a long time. [1:00:40] for you to de-risk the hardest part of that strategy. Whereas if you get the hard part done up front, [1:00:45] It opens up a lot of amazing, you know, adjacencies, different ways to grow. So I think there's definitely truth in that statement. [1:00:53] So kind of the takeaway here is if you're trying to add a marketplace, [1:00:57] It's going to be very hard. [1:00:59] You shared a couple things of [1:01:01] A couple attributes that kind of tell you that maybe it might work. Maybe just share those again real quick. So take Square or Substack or Patreon or like Eventbrite, right? [1:01:11] All of these are examples of SaaS businesses, [1:01:14] that in providing the SaaS service, [1:01:16] deal with the customer and the customer's customer right [1:01:20] That makes it a lot easier because you already have a relationship. They already know your brand. Whereas like Stripe. [1:01:24] When we buy things, we don't know if we're buying it through Stripe, right? Like it's just a SaaS tool that's white labeled in the back end. We have no idea. [1:01:31] So I think that's the first element, right? There's a lot of businesses that look like that, but it is a subset. [1:01:35] SaaS businesses that look like that. [1:01:38] And then secondly,

[1:01:39] Those customers that you're building a relationship with, [1:01:43] They. [1:01:44] need to have a reason to transact with more than one of your current customers on the supply side. Right. So if you're always going to use that same supplier, [1:01:54] There's no need for me to show you the rest of the inventory. [1:01:57] And basically all marketplaces win on, like we mentioned, selection. [1:02:01] and you're needing selection. [1:02:03] And then there's, of course, the marketplaces that win with standardization of like, [1:02:08] I can get it at any time. It might be from different people, but I know the Uber car is going to be there in five minutes or whatever. I don't know. [1:02:13] care as much for the driver. So, [1:02:15] One of those has to be there. And you just find in a lot of these examples, like, [1:02:21] There isn't really that much of a discovery need for... [1:02:25] for who the demand side would be in this theoretical marketplace when you dig into it. [1:02:30] It might be helpful just to define like what makes it a marketplace. In my eyes, it's you're driving demand to supply. That's like the nuance, right? Because otherwise you're just a tool that they're using to do something. Yeah, exactly right. It's like the primary value prop of why supply signs up. [1:02:45] is that you're going to bring them extra business. [1:02:48] and [1:02:49] If you have [1:02:50] you know, customer buyers and sellers, but the primary reason supply signs up [1:02:54] is for tooling or payments like how eventbrite started [1:02:58] That's not really a marketplace. You know, I call it a SaaS-like network. [1:03:02] Some people might just call it a SaaS business, a payments business, whatever the case may be. [1:03:05] But it doesn't really have those cross-side network effects that we associate with marketplaces.

[1:03:10] What this makes me think about is Patreon and Substack. [1:03:13] And interestingly, when I was looking at early Patreon days, [1:03:17] what i understand is they want it to be a marketplace they want it to be a place where artists come [1:03:22] They collect payments and then people can discover artists. [1:03:25] to patronize. [1:03:27] And what they found is nobody needed. That wasn't a problem. I'm not looking for. I'm not looking for other people to donate money to. Same with GoFundMe, right? [1:03:35] And so with that story in mind, when I was chatting with the Substack people early on, I was like, this is exactly what you're going to run into. No one's like sitting around looking for more newsletters to subscribe to. [1:03:44] But, [1:03:45] Shockingly, they've actually pulled it off. They have a really wild network effect going now. I've been very impressed with what they've built. I don't know if I'd call it [1:03:54] marketplace yet, but the [1:03:58] I think they were able to abstract away the version of what you just said to something that was actually useful, which is yes. [1:04:03] I don't want more emails to subscribe to. [1:04:06] But I do, I am fundamentally interested in more articles relevant to my interests. [1:04:11] and [1:04:12] no one has really solved that super well twitter was a very hacky form of that you know where i follow you and then i see when you post a new blog because you [1:04:21] You tell everyone on Twitter you post a new vlog, but it isn't built for that. [1:04:25] Whereas the Substack Reader, [1:04:28] will now allow me to be like, oh yeah, sure, I'll subscribe. So I really like what they've done. It'll be interesting to see how big a business that is and how it all plays out. But [1:04:36] I think their execution has been really solid. [1:04:38] Awesome. I feel the same way. And it is, I also wasn't sure whether it makes like, it's worth calling a marketplace, but.

[1:04:44] For whatever it's worth, 80% of my new signups are coming from Substack's network. [1:04:49] the recommendations feature and their onboarding things like that. But do you think that's unique to you because you are already a top one percenter sub stacker? [1:04:56] Or is that a meaningful number for people? [1:04:59] newish players. Like if I finally switch my email list to Substack, which has been on my list for a long time, uh... [1:05:07] Yo. [1:05:08] That probably will be the same percentage for me, right? Absolutely not. Yeah. No, I think there's definitely if you're there first and people know of you, more people will recommend you. [1:05:16] But I think over time, more and more people start to recommend you. [1:05:21] I think they released a stat that [1:05:22] Something like 40% of their new users are coming from a recommendation or new sign-ups or something like that. So it's pretty widely... [1:05:30] happening but I don't know how much of that is just the top 10. [1:05:32] So it's pretty incredible. It was really innovative. [1:05:35] Good job, Substack. And we had... [1:05:37] They're at a product on this podcast talking about that exact feature for a whole [1:05:41] for a whole hour if you're interested. [1:05:43] Awesome. OK, final question. And then we have our very exciting lightning round. And this is a round. [1:05:47] consumer companies and consumer subscription companies [1:05:51] I know you work with a lot of founders that are building consumer subscription companies and just consumer startups in general. [1:05:57] Yep. [1:05:57] you often tell me how they're incredibly hard to build into thriving businesses. Can you just talk about why consumer subscription [1:06:05] Startups are so hard and products are so hard. I feel like you're making this quite a downer podcast, Lenny. This is people need to hear the truth. You know, there's a lot of happy talk out there, a lot of posts. Here's how you do this.

[1:06:16] And sometimes you can't. [1:06:18] I think in order to understand these businesses, you have to start with, why do investors like [1:06:22] B2B SaaS. Why do they like B2B subscription? [1:06:26] And I think the way most people respond to that question is like, oh, predictable revenue, right? And it's like, ah, sure. I mean, that's cool and all, but I, [1:06:33] I think that's actually not the most important. [1:06:36] So I'd argue there's [1:06:39] to great attributes of B2B SaaS. [1:06:42] One is that businesses are more predictable in how well they're routine. [1:06:45] They're rational. You can understand who are good versus bad businesses for your product, as well as which ones are going to grow versus go out of business. [1:06:52] And more importantly, the second thing that's great about B2B SaaS is net dollar retention. So [1:06:58] What is net dollar retention if you don't work at SAS? Well, [1:07:02] As a SaaS company, [1:07:04] some of your customers are gonna churn. [1:07:06] Some of your customers are going to stay. [1:07:07] and [1:07:08] normally outside of potentially current macroeconomic conditions with all the layoffs. [1:07:12] When customers do stick around, they tend to spend more. [1:07:15] Either they buy more seats if you're a seat-based model, or they use the product more if you're a usage model. [1:07:21] So the SaaS company just makes more money in year two, year three, et cetera. [1:07:26] So, [1:07:27] Consumer subscription just doesn't have any of these benefits. [1:07:30] Consumers are way less predictable. [1:07:32] They tend to retain worse than businesses. [1:07:35] And they also don't have net dollar retention characteristics. [1:07:38] So if [1:07:40] the user retains paying you in year two, you're probably making the same amount that you made from them in year one, not more.

[1:07:47] What that means is you need higher... [1:07:49] user-based retention than a B2B SaaS businesses with more unpredictable users. [1:07:55] And it's a lot higher than I think. [1:07:57] founders tend to think. We're talking annual retention that needs to be north of 60, perhaps even 70%. [1:08:03] So you look at who's actually been able to do that at scale. [1:08:07] And it's a really small list. [1:08:09] Netflix in the US. [1:08:11] Amazon Prime, [1:08:13] Spotify, Duolingo, I think is emerging as one of these players that's making it work. [1:08:19] End. [1:08:20] when you look at how they do it, [1:08:22] They're either doing it with [1:08:24] massive OPEX and economies of scale, [1:08:27] or through a network effect or some other bespoke growth loop that's not that easy to replicate. [1:08:33] So Duolingo has a strong data network effect. The lessons get better the more people use it. [1:08:38] Beak, the company I'm on the board of, [1:08:41] has a cross-site network effect between creators who create the content and the listeners. [1:08:45] and the creators bring a lot of distribution from their existing [1:08:49] you know, social networks to bring new people in the app to listen. [1:08:53] Netflix and Amazon. I spent billions of dollars on content. [1:08:57] So, [1:08:58] I think the default path of like, [1:09:00] I'm going to spend money on paid acquisition. I'm going to retain half of my audience year over year. [1:09:05] That's just a path to going on. [1:09:06] eventually. [1:09:07] And I think what's interesting about these businesses is you can model it. [1:09:11] So you can learn when it's going to happen. [1:09:14] you can learn like [1:09:15] you know, when the retention dips and when you can no longer let your profit, profitably acquire users.

[1:09:20] And if you want to look at like a model in real time, [1:09:23] Just look at Blue Apron. [1:09:24] The company raised $300 million in an IPO that valued it at $2 billion. [1:09:30] It's worth $50 million today on the New York Stock Exchange. [1:09:34] And I think people understand now if they didn't before that, [1:09:39] paid acquisition tends to get worse as you scale. You target the best customers first, they have great conversion, great retention, [1:09:45] And as you expand your targeting of new customers, [1:09:48] every one of those metrics gets worse until it's no longer profitable. Maybe two years from now, it may be $5. [1:09:53] years from now, but eventually it'll be no longer profitable. [1:09:57] And what network effects allow you to do is they allow your product to get better faster than the customers you target get worse. [1:10:04] normally through increased selection, like some of the examples we gave with DoorDash and [1:10:09] Now there's. [1:10:10] Such a cool topic. I actually have a post about this that I recall now as you're talking about it. [1:10:15] And just to kind of double click on the retention point, and it's like, [1:10:19] Freaks you out when you really get into it. [1:10:21] Let's say your retention is like 70% like cohort retention for a year. [1:10:25] I forget the math, but like every three or four years, you basically have to rebuild your entire user base. [1:10:31] because it just keeps trickling out. So your growth just has to continue. It's mind boggling. You run out of humans. [1:10:36] This is. [1:10:37] like it's yeah it's really hard so like [1:10:39] If you can retain people incredibly well, meaning those people just never churn, [1:10:43] Yeah, you've got a great business, but. [1:10:46] Look at the effort the companies that have done that are doing to do that. Spotify does make profit. Netflix, Amazon, they spend so much money to try to do that. And how replicable is that for a startup? I'd rather say like...

[1:10:59] Let's try and get some network effects involved here. Let's get our customer acquisition cost to zero. You know, let's think about other ways to monetize because this alone feels like hard [1:11:08] And this is also why a lot of these companies pivot to B2B. They realize they're not going to get anywhere with B2C. [1:11:14] The other lesson I took away, because I interviewed a lot of these early [1:11:17] B2C subscription app founders and employees. [1:11:20] And I think about companies like Grammarly, Duolingo, Noom, [1:11:24] forget there a few more one of the other trends across the mall is they're all very [1:11:27] efficient and very small for a long time [1:11:30] because, [1:11:31] They needed time to figure out how to make anything work and then just continue to stay small and super scrappy. [1:11:37] I think that's a great point. It was an interesting case study between Calm and Headspace. [1:11:41] Because Calm remained like 10 people forever. And Headspace had gone into hundreds of people. [1:11:48] And Calm basically never lost money. And that allowed them to be, you know, [1:11:54] allowed them to pivot easier during massive changes like app tracking transparency, like through a wrench into all paid acquisition. [1:12:02] so if you're not losing money like cool you have time to figure out how to like grow again but if [1:12:06] That increases your burn by like $100 million you have. [1:12:09] a lot more of a panic. [1:12:10] on your hands. [1:12:12] So maybe just to tie the loop, tie the knot, close the thread on this [1:12:17] topic. [1:12:18] If you're a B2C subscription founder, [1:12:21] what would be some takeaways that you would want to put in their head to think about how to maybe survive? Yeah. So if, [1:12:27] What your plan is, is to use paid acquisition on top of a freemium model to get a percentage of people

[1:12:33] to convert and hopefully stick around forever. [1:12:36] I'd pivot right now. It's like, [1:12:38] I just can't see it working. So how do you pivot? Well, how do you make it social so that [1:12:44] people are bringing in other people. How do you get a supply side to the, you know, content or education or whatever you're building? [1:12:51] and make it in the supply side's interest to refer people. [1:12:55] You have to do something [1:12:57] That's either building some more organic growth loops into the business or building network effects into the business. [1:13:03] Or else like. [1:13:04] We could go spend two hours and model out exactly when you're going to run out of money. Like it's just that predictable. [1:13:09] And, you know, a lot of founders don't like to hear it when I tell them that, but I'd rather them hear it now than learn it three years from now. [1:13:15] Well, we've reached our very exciting lightning round. I'm just going to dive right in. What are two or three books that you've most recommended to other people? Definitely The Goal. [1:13:24] by Elihu Goldratt. [1:13:27] which explains how my brain works pretty well. [1:13:31] Hmm. Uh... [1:13:33] Thinking Fast and Slow by Danny Kahneman. I took a lot of the [1:13:38] behavioral economics classes in [1:13:41] my MBA program and then his book came out like the year after. So I just like, [1:13:45] absorbed it as quick as possible. [1:13:48] Wait, was he a professor or that was just the topic? Some of his pupils were professors of mine. Amazing. Very cool. [1:13:56] And then the third is a book called [1:13:59] Profit from the Core by Chris Sook. [1:14:02] Amazing. That's a new one.

[1:14:03] I love when there's new books being added to the collection. Here we go. [1:14:06] Okay, next question. Favorite recent movie or TV show? [1:14:10] Well, Party Down just came back. I've been watching that. I've never heard of it. It's one of my favorite comedies of all time. Go back and watch the first couple of seasons. That's what I've been doing. [1:14:22] Other things like The Last of Us is great. I really enjoyed the video games. [1:14:28] Severance is great. [1:14:30] Station Eleven is great. I loved the book too, but I thought that was really good. [1:14:34] Movies? They're... [1:14:35] There haven't really been a lot of great movies lately, but I... [1:14:39] I did rewatch Kicking and Screaming from the 90s. [1:14:43] And it really holds up well. I don't know how many of you have heard of it, but [1:14:47] It's about a group of college kids who graduate. [1:14:50] but they just decide not to leave campus and start their lives, and they just stick around campus. [1:14:55] So I really enjoyed that one. [1:14:58] Amazing. We have a drinking game anytime someone mentions Last of Us, so if you're [1:15:02] Driving to work, listening to your podcast on your commute, take some coffee or whatever you got. [1:15:09] Two more questions. [1:15:10] what's something relatively minor you've changed in your product development process or maybe that you've recommended people change? [1:15:16] that has had a tremendous impact on people's ability to execute. [1:15:20] You know, I preach like cross-functional teams and alignment. [1:15:23] pretty aggressively. So how to get people from all the different functions allied on the same OKRs, [1:15:29] and working together. [1:15:30] And I used to say, like, look, if someone in that,

[1:15:34] cross-functional team is playing rank, it means something's fucked up on a team. [1:15:38] But something we started doing more recently at both Eventbrite and whatnot is [1:15:43] just designating the person on the cross-functional team who drives the project. [1:15:47] At Eventbrite, we call it the driver. At whatnot, we call it the DRI, directly responsible individual. [1:15:53] But once that person makes the call, it's disagree and commit time. There's no escalation pass. [1:15:58] Like. [1:15:59] If you're not the DRI and you're on the team and they made the call, all right. [1:16:02] It's time to sign up and go forward on that decision. [1:16:05] And if they made the wrong decision later, okay, we'll learn, you know, [1:16:10] once we share. [1:16:11] Final question. [1:16:12] Something I love that you do in your own newsletter and blog is you share what you're listening to. [1:16:17] What are you listening to these days? Sure. [1:16:20] Kalala recently came out with [1:16:22] her second album, Raven, which is really great. I've been listening to that. [1:16:27] Orbital came out with a new album called Optimal Delusion. That's been pretty good. [1:16:33] And then an oldie but goodie is... [1:16:36] De La Soul's music finally came out on Spotify. [1:16:40] And the stakes is high is like one of my favorite hip hop albums ever. So I've been jamming to that. [1:16:45] Amazing. Where can folks find you online if they want to reach out and learn more? [1:16:48] And how can listeners be useful to you? [1:16:50] I blogged. [1:16:52] not near as frequently as you at CaseyAccidental.com. [1:16:56] You can find me on Twitter, at OneCaseMan. [1:16:59] And yeah, just pay it forward. Help your companies build better products and better businesses. That's all I care about.

[1:17:05] Casey, thank you so much for being here. I feel like we're building some kind of network effect, you and I doing these podcasts. [1:17:11] Hopefully, you'll be back for a third time someday. [1:17:13] Thank you again. Really appreciate it. Yeah, thanks for having me. Hopefully they don't get sick of me. Bye then. [1:17:17] Bye, everyone. [1:17:28] Also, please consider giving us a rating or leaving a review, as that really helps other listeners find the podcast. [1:17:34] You can find all past episodes or learn more about the show at Lenny's podcast dot com. [1:17:40] See you in the next episode.