Introduction

Patrick: [00:00:49] My guest this week is unique, and so requires a short story. I met our guest, Michael Mayer, because of Twitter. I followed and enjoyed one of several pseudonymous accounts that he maintains to experiment with ideas. His various accounts have wide followings. I think many of the best accounts on Twitter are anonymous or pseudonymous. And I've always made a point to get to know the ones that I like best.

As it turns out, Michael was also an entrepreneur. He'd been building a new company and was raising a small amount of outside capital. I didn't personally invest, in part, because he raised it so quickly after I spoke with him. Ever since I've gotten to know him better, I've followed his company Bottomless with interest. You know that I'm always hyper-transparent about any potential conflicts of interest so it's worth noting that while I'm not an investor in his company, I expect to be at some point in the future.

The topic of our conversation is both his social media activity and his company. I'm a coffee fanatic, and the problem he is solving is one that I live. I order a weekly bag of coffee beans, but I often have too much coffee or run out. Bottomless solves this by shipping you a simple scale, which you can keep wherever you store your coffee. You connect it to your WiFi and set your bag of coffee on it. It automatically orders new coffee for you at the right time. Thus, the name Bottomless. If you like this conversation, check out bottomless.com and their product.

With this podcast, all I'm really trying to do is find, meet and learn from interesting people. Michael certainly qualifies. I hope you enjoy this unique episode.

Maybe we can begin by you telling me why you choose to write to pseudonymously online.

Michael: [00:02:09] So a lot of people assume that I do this because I want to protect my real life identity from anything offensive I might say. That's not really the case. I mean, that's part of it. It is useful to have this firewall between anything you say online and your real personality. But the crux of it is that I just don't think who I am is that relevant to people out there in the world, in terms of what I'm saying. Why should people care about picture of my face, a link to my LinkedIn? It's noise. I think the best way to be successful with publishing online is to have the best signal to noise ratio. So you go to my account, you see nothing about me, you don't see a picture of me, you directly get to the content. So, yeah, I don't think my identity is the most relevant thing in the world.

Patrick: [00:02:58] What have been the most positive things to come out of that, that you didn't expect? Maybe if you had assumed it was going to be a success, and I'm sure you didn't assume that. But what have been some of the positive things that have happened as a result?

Michael: [00:03:08] It, surprisingly, doesn't matter whether my public identity is connected to my account at all. I don't think I've seen any difference. I myself am not a famous person because of it. But it's somewhat easy to transfer that social capital into the right places when necessary. So, it's somewhat overrated to accumulate social capital only under your own name. I still have control of the various accounts that I have. They're still my own capital. It's like a Swiss bank account or whatever. I can transfer out of it whenever I want.

Michael's Backstory

Patrick: [00:03:42] Yeah, really interesting concept. So, let's talk a little bit about your personal background before we get into some of the tech and entrepreneurial type stuff. So, just tell me your backstory. I found it really fascinating when you and I first talked. I think people will appreciate it as the fertile soil for what's going to come next.

Michael: [00:03:59] The most interesting thing about my background is probably less where I am today and more about my speed of progress in the last five years and where I was five years ago. A snapshot of where I was five years ago; I dropped out of school, it was an unranked university, by the way. I think if you went and found my university on the U.S. News & World Report, it says unranked.

Patrick: [00:04:19] It's not there.

Michael: [00:04:20] I had dropped out of that because I couldn't hack it there. I was working at a restaurant washing dishes and doing prep work. I was working with a lot of Spanish speakers. They had a nickname for me. They would call me tortuga, which means turtle in Spanish. So I wasn't even particularly succeeding at that. If you flash forward to today, I consider myself a fairly productive person. I have this company with venture backing that's growing fast. I have a lot of distribution for my ideas online through various accounts. So, it's been quite a rapid clip of improvements since then. I would probably put almost all of that down to just getting my habits right. I can get into that if you're curious.

It's been quite a rapid clip of improvements since then. I would probably put almost all of that down to just getting my habits right.

Patrick: [00:05:02] Yeah, I'd love that. So, what was the first step?

Michael: [00:05:03] The first step, there's just a constellation of habits that are useful. Some of them are removing things and some of them are adding things. I honestly think that the removing of things is easier; removing unhealthy food, removing binge watching TV shows, removing junk information, junk food, just removing as much junk as I could, removing very unhealthy people from my life. Whenever I hear people say that, I cringe. It sounds cruel. But in reality, it's very helpful to remove people from your life that don't really want to see you happy because we're social creatures, and we want to make the people around us happy. If the people around you don't want you to thrive, then you actually subconsciously internalize that and don't really want that for yourself, either.

So if you surround yourself with people that actually are happy when things go well for you, suddenly, it starts to create this incentive in your social mind, or whatever you would call it, to do better. So removing things and then adding things, exercising regularly, personal health. Your physical and mental health is upstream of your productivity - trying to be productive first is virtually impossible no matter how ambitious you are. I was ambitious at the time, and I just couldn't find a way to hack it because I didn't have the right foundation.

Patrick: [00:06:18] Was there a catalyst for - so, obviously, a lot of these things are unsurprising in the sense that they're so obvious - but what I'm often interested in is the original point of trying to do this. Was there an epiphany or some catalyst that caused you to realize at some moment that you had to change course? Or was it a slow process?

Michael: [00:06:39] I don't think there was a seminal moment, to be honest. I think it was just the momentum flipping from the wrong direction to the right direction. Once things are going in the right direction, habits really can compound. So once you have a solid foundation, you can start layering better habits on top. So, I probably didn't quit binge watching TV shows until after I got a solid exercise habit and was eating well, and I felt well enough to do that.

Patrick: [00:07:03] So, on top of those habits, where is the origin point for your interest in technology and business and startups? Where did all that get laid?

Michael: [00:07:13] It just seems to me like technology is obviously the most fascinating thing going on right now. Human beings are tool builders. Everything around us was a technology once before. So building new technology is just really building the future. I think it's the best way to make an impact. I don't think any of that's particularly surprising, either.

Patrick: [00:07:29] I don't know if you got it from the same book, this book, Algorithms to Live By. But I've seen you write about explore-exploit, and chapter two of that book, here and there. So I'd love to hear your thoughts on that concept, where the prescription of that chapter, if you will, is that if you're early on in your journey, you should basically spend all your time exploring.

Michael: [00:07:46] Oh, yeah.

Patrick: [00:07:47] So, you have come to some conclusion of the exploration process because you're doing something very literal, and you're starting a business. But talk about what your thoughts are on maybe over exploring as the right way to approach things. I'd love to hear about that exploration process for you, personally, before starting Bottomless in this case.

Michael: [00:08:04] Almost by definition, you shouldn't over explore, you should explore the right amount. The question is, what is the right amount? Maybe you should over explore because most people don't explore enough. So in their mind, they should over explore. The funny thing is you were asking about how I got into technology. That was just a pure exploration. I had a really simple thing that I wanted. I thought, "Why not just try to build it?" I was shocked at how easy it was to build. I don't think people understand how coding works today. There are tons of massively powerful libraries and technologies that are totally free to use and fully documented online, with tons of examples all over the place for virtually anything that you want to do. People don't understand that. They think that you have to be some sort of wizard or some sort of incredibly educated person. Somebody needs to teach you to code. That's absolutely not the case.

There are tons of massively powerful libraries and technologies that are totally free to use and fully documented online, with tons of examples all over the place for virtually anything that you want to do. People don't understand that. They think that you have to be some sort of wizard or some sort of incredibly educated person.

So, this was like a simple project. It was basically I wanted to show all the show times for the independent movie theaters in my town. Not an interesting product, but I wanted it. So I sat down to build it. I thought it was just quixotic attempt that I would abandon. But I made such rapid progress on it and it turns out, there's an example for exactly how to do this. You can use all of these libraries that already exists completely for free. When I discovered the power of that, and I realized, "Holy crap, there's examples for literally anything that I want to do out there. There's free open source libraries that are fully documented that I can use." That really piqued my interest because it's just like a massive amount of leverage for a curious person that is just available. It's just sitting there open for anyone to figure out.

Patrick: [00:09:42] So, that's the code epiphany, we'll call it. What was the progression like from that to trying to find if the independent movie theater is not an interesting idea of business, to try to find something that you could commercialize, that potentially can be very large?

Michael: [00:09:56] For me, it wasn't a conscious decision. I'm just always thinking about businesses and random business ideas and just obsessed with business strategy or whatever, reading whatever I could come across. So, it wasn't really so much of a progression. It was just stumbling upon an idea that I couldn't get out of my mind that I thought could potentially be very powerful. That's what eventually drove me to leave where I was working at the time to start Bottomless. I was so taken. I was infatuated with the concept and the potential of what the future could look like if we were able to really attack the idea.

Most Powerful Business Concepts

Patrick: [00:10:32] What are some of the most powerful business concepts that maybe at this point in your life you believe in or hold most firmly? The same exact thing you said about open source libraries could be said about things like business strategy, where now there's this incredibly well codified set of books and articles that if you read these things, you're better informed on business strategy than anyone in the world was 25 years ago. So I'm curious in all that exploration, you say you're obsessed with business. Maybe what is the point that you're obsessed with? Or is it just the whole thing? Are there strategic elements to business that interest you most?

Michael: [00:11:06] I think you hit the nail on the head that, basically, the keys to the castle are there and available to anyone. I'm not sure that there's that much that I can add because that's already out there. The market for ideas and business strategy is pretty efficient. Berkshire Hathaway and Amazon.com shareholder letters, they basically give it away for free. The only thing that isn't free is actually executing on that and the judgment to choose the right strategy. I think people get infatuated with certain strategies. You see things that become popular and then everybody wants to do it, both in the business world and the technology world.

For instance, right now, computer vision is really hot because suddenly you can make all of these real world things-

Patrick: [00:11:46] Readable, yeah.

Michael: [00:11:47] Yeah. Fundamentally, the real world has now become legible to computers and to the internet through cameras. But that doesn't mean that every company should be doing computer vision. For instance, Bottomless is using weight-based sensors to restock. If I had a nickel for every person who's told us we should do computer vision, we would already have raised our series, like, C rounds. But fundamentally, I don't think computer vision is going to work in people's cabinets or fridges or wherever it may be. I fundamentally think it's not the right technology. But that doesn't mean you're not going to see everyone trying every possible computer vision thing they can, which isn't a bad thing, right? Who knew-

Patrick: [00:12:24] A lot of consumer surplus in there.

Michael: [00:12:25] Who knew what the right application of the internet or any particular technology was? So it's definitely good for people to explore. But you see these trends - like, for instance, I've heard on your podcast people talking about value investing, and how value investing became very popular after the dot com bubble. But now, there's probably just not much alpha left unless you have some other unique insight. So I think you have to be careful following trends and in particular, contrarian ideas, which is counterintuitive. You would think contrarian ideas are a way to avoid the trends. By the time that you hear them, they're not contrarian anymore by definition.

Patrick: [00:13:03] You need to originate them, not follow them.

Michael: [00:13:05] Yeah. You have to understand you're one of the average people out there. By the time the contrarian idea hits your mind, assuming that you didn't come up with yourself that you've found it online or you read about it, it's not contrarian anymore. It's actually probably less likely to work than mainstream because contrarian ideas, they're contingent, and the opportunity vanishes once people discover it.

Patrick: [00:13:26] Before we lose track of some of the writing stuff because I do want to spend quite a long time on the business and your thoughts on the competitive landscape in this world, I'm curious for your thoughts on the intersection of education and social capital, which you talked about earlier. So if you believe that there's, I think, some way to build social capital through sharing interesting ideas, I guess, is kind of the way that you've done it. But I think you and I share some, maybe we'll call it, a negative view on classical education. I just love to get your take on your own education. You mentioned dropping out of school, and then obviously, your activity has been very counter to that. So talk about social capital, how you think it's most effectively built and your thoughts on education.

Michael: [00:14:10] The traditional ways of building social capital are definitely useful if you can get a Harvard degree or a Stanford degree, of course. But even then, they can only be so valuable because they're being minted. You're still replaceable with another Harvard MBA. If you can create social capital in a totally unique way, you're totally irreplaceable, and that is much more valuable. Even if what you're doing, if you imagine that it's more valuable than a credential that is stamped out, almost an industrial process, you can do something that's actually less impressive than a Harvard MBA. For instance, for me, I've built up a number of pseudonymous accounts. Maybe doing that is actually not as impressive as being like a Harvard undergrad and like a Stanford MBA or whatever it may be.

Patrick: [00:14:56] But it's unique.

Michael: [00:14:56] But even if it's less impressive, it is almost more valuable because you can't replace that.

Patrick: [00:15:02] Super interesting. Do you think there's any rules? I think the answer is no. But do you think there's any kind of playbook for original social capital creation other than just authentic pursuit of what interests you and sharing of it?

Michael: [00:15:18] Are there any rules?

Patrick: [00:15:19] Or maybe not rules, but just things that as you've continued to do it, you would say you've learned or your beliefs about it that have strengthened in the process.

Michael: [00:15:28] I think that originality is important. And in order to achieve some degree of originality, you need to have an original information diet. You're not really that special. If you consume the same information as everybody else, how are you going to come up with unique ideas. Maybe you can. Maybe you're such a fantastic algorithm that you can combine those in a unique way. Fundamentally, it's unlikely. So if you can find a way to try to inject random, unique, interesting things into your mind at a regular point, and also try to regularly inject good content into your mind. I know a lot of people that basically spend 90% of their information consumption on sports and serial TV shows and stuff like that, then they try to come up with unique insights about tech or business. It's like, "Good luck, man."

I started cutting out a lot of what I consider junk information when one day I was in the shower, and I was sitting there and thinking of "Oh, should the team trade x player? Or would it be better if the head coach took this player and put them in the rotation?" I had this epiphany, I though, "Is this what I want?"

Patrick: [00:16:34] Your brain is hijacked.

Michael: [00:16:35] No. Is this what I want the stuff of my thought to be? Is this what I want to be having an epiphany about? Am I adding any value to my life or the lives of other people by having an insight about the configuration of what players play at what time in some regional sports team? I'm really not. To get back to, I think, what the original question was. Social capital formation through like content creation. You're not going to be able to get very far unless you have really good quality information coming into your mind, at which point you are able to transform it into good quality content.

Patrick: [00:17:06] For you, what has been the nature of that information? So not necessarily - you could share individuals' types or sources if you think that's useful - but any thoughts on what those inputs look like relative to the generic content diet out there?

Michael: [00:17:22] I was hoping you would ask me for book recommendations, so that I could explicitly not give any book recommendations because...

Patrick: [00:17:30] Now's your chance.

Michael: [00:17:31] Yeah. Now's my chance. I'm going to answer the question that I wanted to be asked. So, I've noticed on a vast majority of podcasts that guests will be asked for book recommendations, and almost everybody will give lots of book recommendations. I'm guessing that most of those people don't read that many books, or maybe read like a couple books a quarter. That's probably the median.

Patrick: [00:17:50] That's why I stopped asking.

Michael: [00:17:51] Okay, that's good. That's good. Yeah. I noticed there's not much about books in the modern age that would make the quality of information in them that much better than other sources. I mean, obviously, you have editors, but there's a lot of really garbage books out there just because somebody has a lot of distribution. They're famous. Those aren't that great. For instance, I'm a big fan of CS183 notes that Blake Masters took on the...

Patrick: [00:18:17] The Zero to One source material, yeah.

Michael: [00:18:18] Yeah, exactly, the Zero to One source material. I think I've read that three or four times. I bought Zero to One - I actually think Zero to One is worse than the Tumblr blog. There's a lot of great writers out there that write a lot of good stuff. I could name some. There's Jerry Neumann. His blog is really great. It's so good. There's a lot of really, really great blogs out there. Kevin Simler's blog is good.

Patrick: [00:18:40] The best, two past podcast guests.

Michael: [00:18:42] Oh, really?

Patrick: [00:18:43] Yeah. Had them both.

Michael: [00:18:44] Inside Baseball. But to get back to information sources, it's much more important for people to cut out bad information than it is for them to seek out good information. I imagine there are so many people out there that want to learn and educate themselves. They spend 95% of their time just consuming junk news articles online. Then they sit down to read this book, recommended on some podcasts. I'm like, "That's not going to get you very far." You're way better off cutting out all of that other stuff, and then just reading what interests you. Whether that's books or podcasts or social media, as long as you curate your social media feed to remove all the garbage, I guarantee you, you can build a Twitter feed that is very high quality by just removing anybody that tweets about the Democrats or the Republicans or...

Patrick: [00:19:31] Anything topical.

Michael: [00:19:33] Yeah, anything topical.

Patrick: [00:19:34] Pretty much.

Michael: [00:19:34] Yeah, yeah. Actually, that's a good point. Anything that's not going to be relevant in a day is unlikely to be a source of solid insight more than a day from now. It seems really simple. But if it's not going to matter tomorrow, why would you want to consume that information today because the shelf life of the information you consume is going to be so short. The stuff in your brain is just going to be constantly out of date and useless basically.

Anything that's not going to be relevant in a day is unlikely to be a source of solid insight more than a day from now. It seems really simple. But if it's not going to matter tomorrow, why would you want to consume that information today because the shelf life of the information you consume is going to be so short.

Patrick: [00:20:00] One of the things I'm always surprised about - I don't consider myself a particularly controversial person - but the opinion I have, which seems to bring up the most controversy, is not reading the news. When I tell people I don't read the news, they cannot believe it. They think I'm a moron.

Michael: [00:20:14] How do you stay informed, Patrick?

Patrick: [00:20:17] Yeah.

Michael: [00:20:17] Don't you have to be informed?

Patrick: [00:20:18] It's always some version of that, yeah.

Michael: [00:20:19] Yeah. You actually are uninformed if you're reading daily news...

Patrick: [00:20:24] Exactly.

Michael: [00:20:26] ... because it's completely out of date. I mean, news is good if it's weekly, or monthly news, perhaps, or quarterly. That almost isn't news anymore. But there is definitely...

Patrick: [00:20:35] Information anyhow. It's amazing how much you can keep dropping and have zero impact on your life - negative impact - or your business or really anything else you're trying to do, and obviously gain a huge amount of free time to do other stuff.

Michael: [00:20:48] Yeah, if you read something that is still being recommended by people 10 years later. Anytime you get a book recommendation for something that's 30 years old, if it's 30 years old, or 1,000 years old, you should fill your brain with that stuff as much as you can.

Patrick: [00:21:01] Before we lose the thread. The reasons why you think the standard system of education is busted. I'd be interested in that.

Michael: [00:21:08] We have this system of education where essentially everybody needs to learn at the same pace. It makes sense, it's like a batch process. Batch processes are more efficient. So, to the degree the information needs to be coming from the teacher - from one source - then you need to have batch processing of students. Otherwise, it's very inefficient to have one-to-one teacher-student relationships. So, up until the moment where self-education was really possible, you absolutely need to have the system of education that we have today. Otherwise, it's way too expensive.

So you have this system where the teacher teaches to the median, or probably slightly above the median, but they're trying to get as many kids covered in the bell curve as possible by any lesson. Which is okay, it makes sense. You lose the kids at the top and you lose the kids at the bottom. But that's what you have to do.

Nowadays, self-education is absolutely possible. I've completely taught myself code, but it doesn't have to be code. It can be virtually anything. You can have teachers as helpers, while students are learning at their own pace. Now, I know people have wanted to do that for a long time, and that might just be one of those ideas that failed for a long time because there were good reasons why it didn't work but now it might work. I would just recommend people to actually try to learn things themselves rather than pay somebody $2,000 for a course that you just show up to and browse Facebook the whole time.

Patrick: [00:22:33] It's a nice way of thinking about it for me. Now as a dad, this is the thing I think about more than anything else. Since I'm not homeschooling my kids, obviously, that could be one option. That seems a little too daunting. Trying to steer the ship a little bit away from the process, which left me so uninspired and in this kind of wilderness, it is something I think about a lot.

Bottomless

So, let's talk a little bit about the formation of this business and how the business represents all of these interests intersected. What the challenges are, what the advantages are, why it might be interesting. So first, describe what the business is. It's actually pretty straightforward, but has a million really interesting, very deep rabbit holes underneath the simple surface points. So, describe the business and then we'll go down each.

Michael: [00:23:19] Yeah, it's really funny. It almost sounds trivially simple, but it's essentially a smart subscription that figures out the perfect time to ship you your next order using a weight sensor. So, the weight sensor essentially just captures how much you have and how fast you're going through it. So therefore, it's able to ship you at the right time rather than just a typical subscription, which ships you on a monthly or a weekly schedule. We're starting with fresh coffee. So right now, we're a coffee company. In the future, who knows?

Patrick: [00:23:48] I love this idea because the obvious analogy is Amazon had to pick something first to express its core idea. You picked coffee, so I do want to talk about coffee because I think it's an interesting one to start. But the more interesting aspect is the measurement and the scale. So, you mentioned the word legibility earlier, which I think is a critical word - that word specifically - that ties back to this notion of software eating everything. So, talk about that connection and the type of legibility you're trying to create, and why specifically you think that is interesting and hasn't been done.

Michael: [00:24:21] Anytime you can make something important, legible to computers, you can change the world. If you look at all of the revolutions that happened after smartphones got significant penetration, it was because suddenly people's locations were legible, and the world around them was legible with cameras. Suddenly you have, Ubers, Instagram, all these things. So fundamentally, we're making the stock level in your house of coffee legible to us, and then eventually to coffee sellers. Not just your level, but also using those levels, we can figure out how fast you're going through it.

Anytime you can make something important, legible to computers, you can change the world. If you look at all of the revolutions that happened after smartphones got significant penetration, it was because suddenly people's locations were legible, and the world around them was legible with cameras.

So previously, where you have taxis, people's location isn't legible. People have to call to get a taxi and then the dispatcher has to call a taxi driver. It's this whole laborious process. Then with Uber, when you make everything legible to a network, suddenly, it's much more efficient. Suddenly, you see people using taxis a lot more than they used to. Similarly, I think making stock levels and consumption speed legible will actually make subscriptions a reality in a lot more cases than currently exist. It will definitely expand the market if you have subscriptions that actually work.

Patrick: [00:25:31] I guess one assumption that this relies on is that people consume at a rate that isn't really simple and obvious. So, it just so happens that we consume a bag of coffee per week. We consume perfectly one bag of coffee per week, and then maybe every six or seven weeks, we need one supplemental bag. So, I'm as probably as close to the perfect standard subscription service as possible. So, talk about, given the data that you've gathered so far, what the normal customer looks like. Probably, it looks way different than me.

Michael: [00:26:06] Oh, people are totally variable in their consumption. It's all over the map. Even people that consume regularly every day, we'll see inflection points, where suddenly they start consuming faster or slower. Definitely, people will go on vacation. Then we automatically don't ship for you because you're flatlining.

When I started this, I thought that we would have to focus on things that were very variably consumed. Coffee, for some people, you can imagine other examples. I remember I was doing a pitch to somebody for the business. I was telling them coffee works really well but it wouldn't work for something like razors. That's why Dollar Shave Club worked. There was this massive subscription boom after Dollar Shave Club. It turns out only a couple companies actually worked. It's something that you actually can just use until you have to throw it away. That's not true for most things. I was telling him with total confidence that nobody's ever going to want Bottomless for razors. He's like, "Hold on a second." It was a video call, and I was like, where is he going, then he comes back. He has two fistfuls of razors that he got from some subscription service. He's like, "You'd be surprised." I was surprised.

So fundamentally, humans are not predictable creatures. Our consumption of various things are not predictable. Even people that drink coffee at home every day, you might run out of the house one day, you might have guests, it turns out that even very consistent consumers for us, if we were to put them on a set subscription schedule, they would get off very fast - within three or four shipments.

Patrick: [00:27:35] Something would have gone wrong.

Michael: [00:27:36] The timing would be wrong. That's why you don't see subscriptions for all of the repeatedly purchased items in people's houses. I saw this study that the average grocery consumer buys only 13 SKUs every week. So, people are buying the same stuff over and over again, and people go to the grocery stores a couple times a week just to get the same stuff. That's the idea behind Bottomless - it's ridiculous. It's the same products. The information problem in a lot of retail is, do you want a product or not? But with the vast majority of grocery items, the answer is, "Yes, I do want that product. I want it all the time." So that's no longer the information problem. The information problem is when do we ship it. So I think we can solve that for a lot of things.

Patrick: [00:28:17] One, just because there's some recency bias in this because my wife and I just started using Instacart and find it to be a really great, interesting service that, in some ways, attacks the same problem of; we don't want to deal with going to the store all the time. It's all the same crap. We're reordering the same crap all the time. So, when we find out that we need it, it's only two hours away. We can add it in, and we live close enough, I guess, geographically to a Whole Foods or whatever it is, that the food comes to us. So talk about the differences between those two models and where the white space is between them.

Michael: [00:28:49] I think people have Stockholm Syndrome to the way grocery works now. You're like, "Oh, look at this. I only have to fill out a ton of web forms for 10 minutes instead of driving down to the store." Then it arrives two hours later, "Wow, this is great." I actually don't think it's great. I don't like filling out web forms. I think it's only marginally better than going to the store. So I guess I'm biased but I think what we're doing is much better because it arrives before you even know you needed it. So, it removes this whole problem that maybe people don't really know that they have, but once people figure out that they can get rid of this problem, they're really happy.

Patrick: [00:29:25] My wife was asking me what I was going to be talking about today with you. I said, "Well, it's kind of like the that How I Built This Show, but without the survivorship bias because we're doing it very early in the building process."

Michael: [00:29:36] Oh, yeah.

Patrick: [00:29:36] So hopefully, you'll be on that show one day, that'd be great.

Michael: [00:29:40] By the way, people talk a lot of - can I swear on your show?

Patrick: [00:29:42] Yeah, sure, yeah.

Michael: [00:29:43] People talk a lot of shit about How I Built This and what they call entrepreneur porn. I honestly think, yeah, there is some aspect of truth to that. But at the same time, who else would you rather society be worshiping?

Patrick: [00:29:57] Emulating.

Michael: [00:29:57] Yeah, using as role models. Pop stars, sports stars, actors? Nothing wrong with those people. I think entrepreneurs are better to worship. It's better for society as a whole because entrepreneurs take a lot of risk. I'm taking a lot of risk by doing this. The myth of the successful entrepreneur, it's an important myth in our society.

Patrick: [00:30:16] So, I want to touch on some of these, always interesting, early-day aspects. So before we hit record, we were talking about the scales themselves. So, I've done the signup flow, right? So I know how this works. So you sign up, you answer some simple questions. A scale gets shipped to you, you stick it somewhere, you connect it to a WiFi, you put the bag of coffee on it, and the rest is history. The scale, obviously, is a super important, differentiated part of this whole thing. You mentioned how the first couple of them, you were literally putting them together yourself, three hours a pop. So, talk about that discovery process. If sensing and legibility is the theme here, how did you decide on what the scale should be? How you should be sensing? What the options were? Then starting to build them.

Michael: [00:30:58] Okay. So, how did we decide on the form factor? Well, first of all, it had to be easy to build because I'm not a hardware engineer. I don't think I really knew what voltage and current even were when I started this. I didn't really pick something that I knew how to do. I just picked something that I thought could be great and figured it out. So, I used the technology where there's lots of libraries available for doing the basic stuff that you need to do. So, I didn't have to actually know anything, and where there were plenty of tutorials for how to put it together. Then, the main design consideration for us was battery power. If you get a shipment every month, you don't want to be recharging your scale every month.

Patrick: [00:31:33] Right, defeats the purpose, yeah.

Michael: [00:31:35] What's the point of that? So, we had a design goal of having the scales last for at least a year. Basically, the story is, we just started building these things. The first units we put together, as you said, took three hours to put together of painstaking labor. Also, we were painting them to try and make them pretty. We found out nobody cared because we shipped one out that was unpainted, and they didn't complain. So we were like, "Okay, well, let's see how this goes." So, we've not been painting them. Yeah, it's just been a process of; we've turned our apartment into a little mini manufacturing facility every week or so and just build as many scales as we're going to need that week. A lot of hardware companies will try and perfect a product, do a Kickstarter. We've learned so much making different versions of this thing. The version we have now is way better than what we had before.

Patrick: [00:32:21] In what ways?

Michael: [00:32:22] It works 95% of the time instead of 70% of the time.

Patrick: [00:32:25] That's a good start.

Michael: [00:32:27] Yeah, that's important. It's much easier to assemble. We got to the point where eventually the model that I put together was good enough that we could hand it off to an engineer, and they can just draw it onto a PCB (Printed Circuit Board) and get it manufactured. So, the iterative process got us to where we needed to be rather than just trying to dream up something and doing a Kickstarter. A lot of companies die doing Kickstarter, too. I think it's really overrated.

Patrick: [00:32:48] So you got a scale. The scale does what exactly. It is measuring weight and weight only, I suppose at this point. Are there other dimensions that could be measured in the future - I'm thinking somehow measuring freshness or taking other information as part of this idea of seamless replenishment of food staples?

Michael: [00:33:06] One of our fundamental insights is that nobody wants to pay for reordering hardware. We're not the first to come up with the idea of reordering hardware. There's like an Amazon initiative for replenishment services. You have to buy a printer or, I don't know if they have coffee, but you would have to buy a coffeemaker. Nobody wants to do that. You already have your setup. You don't want to pay for reordering hardware. You just want to get your stuff when you need it.

So, one of our most important design considerations was just that we can give this away to people for free and make money off of the actual coffee in the subscription. So, we've kept it purposefully simple. I don't think in the future we're going to add any bells and whistles. It's just one button to do the WiFi sync, and that's it. I think that's how it should be.

Patrick: [00:33:50] We've got this process where you are creating a really interesting disintermediation marketplace kind of model. It's like a lot of the modern tech models all at once. One side of this marketplace is roasters. So, I'm curious what you've learned about that side of the business, and what kind of receptivity you've gotten when you go and say, "I want to sign you up for this thing. All you'll have to do is ship directly, and we're just going to be a customer acquisition vehicle for you for free." It sounds like a good pitch.

Michael: [00:34:21] Yeah. Well, they do give us wholesale price for the most part. So, you do have to convince them a little bit, but it's been surprisingly easy for a couple of reasons. First of all, all of these roasters are e-commerce-enabled already. If you look up a random coffee roaster that you like, odds are they're going to be selling online already. That's thanks to Shopify. So we're riding on top of some waves of e-commerce, fully penetrating into the economy. They're already shipping individual bags. So, it's not that hard to convince them to ship our individual bags. Also, they're very familiar with this problem. They see their customers turn out of subscriptions all the time. They see people complain saying, "Hey, I have five bags, and I want a refund for four of them because they're all going stale."

So, when we go and we tell them about this, they're very excited. A lot of them really want to get onboard because they want their coffee to be super fresh. They don't want people having multiple bags piling up because it doesn't present their coffee well. Part of our value proposition is that we actually ship at the right time, and people can select whether they care more about freshness or not running out. A lot of our customers do care about freshness. So, the roasters love that. They want people to enjoy their coffee at the optimal time.

Leverage and Supplier Power

Patrick: [00:35:29] One of the things I'm always interested in, in any of these strategic power frameworks, whether it's Porter's, or Helmer's, or whoever's is supplier power. I'm curious how you think about supplier power right now and maybe prospectively.

Michael: [00:35:43] I would love to say that suppliers don't have a lot of power in this model. But in reality, they kind of do. Because people are very brand sensitive. So, I could put a cherry on top of that, but we do own the customer relationship. Half of our customers pick just a rotating product.

Patrick: [00:35:57] Like a Pandora.

Michael: [00:35:57] Yeah, exactly. So, we can ship whatever we want. That's a massive amount of...

Patrick: [00:36:03] Leverage.

Michael: [00:36:03] ...leverage for us. So, there's some supplier power. You see a lot of very valuable businesses just own the customer relationship. I think that's something that is under explored from a business strategy perspective. I have been thinking about why that is, and I don't think I really know exactly. I maybe have a couple of theories.

Patrick: [00:36:18] What's your theory because I've been thinking quite a lot about this specific topic lately?

Michael: [00:36:22] I think the person who owns a customer relationship gets to set the default.

Patrick: [00:36:26] How would you define owns, first of all?

Michael: [00:36:28] Amazon owns the customer relationship, not the actual sellers. They've accrued all the value to themselves. A lot of people say that that's because they own the network. That's a big part of it. But I think that there's something powerful about actually just owning the direct customer relationship and being upstream from the rest of the suppliers because every company is sitting on top of a web of suppliers.

Patrick: [00:36:50] Do you think there's some litmus test for this ownership of a willingness to bear inconvenience or something? So, for example, if there was a brand that didn't sell on Amazon, and I was willing to go outside my normal workflow to get them even though I could get a substitute on Amazon, maybe then that brand owns the relationship, not Amazon, in this particular case, or something like that.

Michael: [00:37:09] Oh, you see this playing out a lot. There are brands that won't sell on Amazon. So for instance, one thing we're really proud of is that we just got Philz Coffee, if anybody's familiar with Philz in the Bay area.

Patrick: [00:37:18] Yeah, yeah. Sure, yeah.

Michael: [00:37:19] They just signed up with us.

Patrick: [00:37:20] Does it come with cream, too? That's the signature thing.

Michael: [00:37:24] Yeah, well, not yet. You can't find them on Amazon. That's part of why. But fundamentally, the person that owns the customer relationship gets to set the web of decision making that happens after that. So, that just gives them default leverage because they're upstream. They control where the money flow goes and what the customer decisions are.

Patrick: [00:37:43] They capture inertia, too, in some interesting ways. If people are just, like if it's good enough, you just keep doing it.

Michael: [00:37:48] Yeah, exactly. You see that Amazon are rolling out all these white label brands. They can just do that, and people will buy them if they put them on the top of their page.

Patrick: [00:37:57] So talk to me about feeding that beast. So, I was going to ask you, some of Keith Rabois startup questions. What is the accumulating advantage here? So, one of those is this relationship with the customer. Obviously, if I've been working with Bottomless and I keep adding staples to my Bottomless list, that relationship strengthens over time, and it's harder to tear out. But talk to me about how you intend to deliberately build the relationship with the customer, and then we'll talk about other accumulating advantages.

Michael: [00:38:24] Well, so right off the bat, we have a certain stickiness. Once you have our scale, you're pretty much unlikely to set up a scale from our future competitors. Why would you do that? Then you have to be send out...

Patrick: [00:38:33] Coffee forever.

Michael: [00:38:34] Yeah. You have to send our scale back because we're not selling the scales. We're giving it to you for free. So there's this massive stickiness that you typically would see in SaaS companies that we have, just by default, just by the way that our system works. In terms of accumulating advantages, we can expand from coffee to other products. The really important nut for us to crack is bundling of shipments. If you imagine, you have five, six, seven items, we can figure out the perfect time to ship two or three or four at a time. Once you bundle shipments, the marginal cost of an additional item is virtually zero. Once you have a shipment going to people's houses, there's almost no marginal cost to adding things to that box.

The really important nut for us to crack is bundling of shipments. If you imagine, you have five, six, seven items, we can figure out the perfect time to ship two or three or four at a time. Once you bundle shipments, the marginal cost of an additional item is virtually zero.

So that's, I think, a massive accumulating advantage. That could be really, really huge in the future. If you already have a couple boxes a week going to people's houses, you can actually replace most e-commerce because you can provide things cheaper. Even Amazon with their prime, they still have to break even with the shipping cost. So, all of their products actually account for that cost. But if we don't have to account for the shipping cost because we're already sending a box, we can actually be cheaper. So that's a potential advantage in the future. Amazon would have to rewire the way they work to make that work.

Biggest Obstacles to the Business

Patrick: [00:39:50] You've talked to investors, you've talked to people about this idea a lot, I'm sure. What have been some of the best objections or things that people have held out as reasons why the business won't work?

Michael: [00:40:02] Oh, yeah, that's a great question.

Patrick: [00:40:03] You're not allowed to say like, "Because Amazon will do it."

Michael: [00:40:06] That's the worst. I almost want to just be like, "Okay, well, it was nice meeting with you. Good luck putting all of your LPs money in Amazon stock since you're so bullish on Amazon." I haven't actually said that to anybody. I don't have the nerve. Anyway, the best one was probably, it's a shoutout to John at Founders Fund. He said, "Okay, well, you have a one-to-one hardware to product relationship." That's going to be tough.

Patrick: [00:40:31] Why? Why can't you have multiple sensors on the same piece of hardware? You just ask people to put stuff on the right labeled slot?

Michael: [00:40:36] Yeah. It's a limitation of how Bottomless works now. Not a limitation of how we'll work in the future. So in terms of long-term, there is this big question, "Do people really want to have all of their stuff monitored by sensors and sent to them?" I think skepticism around that is probably the most authentic pushback because, nobody knows. For any company, it's really hard to say. We've been talking a lot about Amazon. So like Amazon, originally, their vision was, "Hey, I went back and I read a lot of their shareholder letters." They said, "15% of all retail is going to be online eventually." They said this back in '98. I think it would have been pretty rational for people to push back and say, "Hey, I don't know if I believe this." They haven't gotten there yet.

Patrick: [00:41:18] It's crazy.

Michael: [00:41:19] It's crazy how much they have to run, but it's also crazy how bullish people are on them thinking they can go from whatever their revenue now is to three times, four times because they've mined all of the most profitable e-commerce opportunities outside of doing something unique, like Bottomless. But I'd sit here and I pitch to people and I say, "I think 15% of all of the repeatedly purchased items in the US, which might be like $500 billion, is going to go on to systems like Bottomless." I think being skeptical about that is rational because that's an audacious view of the future.

Patrick: [00:41:50] Sure, yeah. It's a nice addressable market, though.

Michael: [00:41:52] Oh, yeah. Yeah, yeah. People get excited about that. But there are reasons to be skeptical. Time will tell.

Patrick: [00:41:58] What have been the biggest just operational or personal or emotional challenges? I guess, how far into this are you a year and a half in earnest?

Michael: [00:42:05] More than two years. Bootstrapping a hardware company is hard. It's very hard. I've spent a lot of time just putting together...

Patrick: [00:42:13] Scales?

Michael: [00:42:13] ... literally soldering scales and listening to podcasts and shipping them, taking them to the store and dealing with the ones that don't work. I don't know if it's actually been that hard other than just the exhaustion of it. But it's fulfilling to work hard towards a goal. A lot of entrepreneurs talk about the emotional rollercoaster, and I think it's absolutely true. When you're putting your whole life into something that's so high variance, it's pretty natural to have high variance feelings about it. You have one customer cancel and you think you don't have that much data.

A lot of entrepreneurs talk about the emotional rollercoaster, and I think it's absolutely true. When you're putting your whole life into something that's so high variance, it's pretty natural to have high variance feelings about it. You have one customer cancel and you think you don't have that much data.

Early on, our first pilot was only 15 customers. The results from it were great, but we would lose one of them. I'd be like, "Oh, God, we're screwed. This is definitely not going to work." Small sample sizes can create a lot of anxiety.

Patrick: [00:42:56] What's been the best moment, the best subjectively like emotional moment?

Michael: [00:42:59] Probably the first time I got my retention graph up and running and actually looked and said, "Holy crap, that long-run cohort is now buying more coffee than they did in month six." They're on month 18 or something, and they buy more coffee now than they did in month six. In general, the cohorts will go down, but then they'll flatten out. Just recently, I did it again for our more recent cohorts. We actually see them starting to go up very early. And that's probably because our scales are working better now. So, it's getting even better.

So that's important validation. Customers will send us messages saying, "Somebody told me, it's like having somebody do my laundry. I love Bottomless. It's like having somebody do my laundry."

Patrick: [00:43:41] It's a good compliment.

Michael: [00:43:42] That's a good compliment. Or I'd call a lot of customers and ask them questions about Bottomless. I asked them, like, "How much would I have to pay you to cancel the service?" I think that's a pretty interesting way to figure out the consumer surplus. When people told me $1,000, and I said, "No." We're like, "Really, how much?" They said, "$1,000. You don't get it, before Bottomless, I was getting my car and driving to a coffee roaster. I was running out all the time, too, and you're saving me a half an hour of time regularly. If I play that out for the next couple years, it's easily worth $1,000." That's validation.

Also, people say that you shouldn't care about investor validation, but I think that's bullshit. Investor validation is important. If you can convince smart people to part with their money or their LPs money and believe in your vision, that really means a lot, especially the angel investors that have invested in us and are putting their personal money in. Those are great moments. That's great validation. It's great to bring people on and see that they believe in what you're doing.

Y Combinator Experience

Patrick: [00:44:39] Can you talk about Y Combinator?

Michael: [00:44:40] Yeah, I can. We got into YC. That was a really nice moment.

Patrick: [00:44:44] What was that process like? Anything interesting in that process? Or was it just apply, you got in?

Michael: [00:44:47] Well, the interesting thing is. So, they fly people from around the world to do interviews. But the interview is only 10 minutes - actually, it's only 10 minutes. We were only flying down from Seattle. That's not that big of a deal. But there's people flying from all over the world. At one of the YC dinners, we saw these people from Columbia, and they flew from Columbia to San Francisco to meet for 10 minutes. That's a little insane. I understand why they do it because if it weren't 10 minutes, they actually couldn't interview people face-to-face. That's a crazy part of the process. Then another part of it, they give you the feedback same day, which is really awesome because...

Patrick: [00:45:26] If it's good.

Michael: [00:45:27] Yeah, if it's good, but if it's bad, too, it's good to know.

Patrick: [00:45:31] I guess, yeah.

Michael: [00:45:31] The funny thing is, they call you if you get in and they email you if you don't. We are waiting that afternoon. I happen to have done some food orders. So, I kept getting calls from local numbers. My order got screwed up. So, I got 10 calls from different local numbers. Every time I would answer them, I'd be like, "Hello. Oh, yeah. Did you want the guacamole?" Oh, God. Then like, finally, when I did get the acceptance call, I answered, but I was super nonchalant. I answered almost like in a rude way. They're like, "Oh!" Then I heard who it was, but...

Patrick: [00:46:04] Did they ask any good questions in the interview?

Michael: [00:46:05] The interview is kind of oppositional. My understanding is YC was started by some grad school people. You have to defend your thesis or whatever in grad school. So it's almost like a thesis defense for your business and what you want to be doing. So, I remember leaving and thinking we totally had screwed it up because it was not a positive, friendly - I mean, it wasn't unfriendly, but it also wasn't, they didn't give you the impression that they were super impressed.

Patrick: [00:46:28] You've kept saying we. I want to acknowledge that, unique co-founder story.

Michael: [00:46:33] Unique co-founder story, yeah. So my co-founder is my wife, which is not that common, but becoming more common. I think it's a huge advantage. If you look at a lot of the companies that have been very successful, they've been started by two friends or Stripe, started by brothers. A lot of these people live together and work together in the early days of the startup. The reason is, is in the early days, there are so many decisions that you're making that will determine the path of your company in the future and are very important. They need way more attention than you would be able to put to them during Monday through Friday, 9:00 to 5:00. You have to schedule a meeting to talk; tomorrow at 4:00 to 5:00, we're going to talk about the strategic decision. Good luck making a good decision at that time.

A lot of times, the muse will attack at random moments. So my wife and I will be walking around on a Saturday. Suddenly, we'll talk about something of immense strategic importance for Bottomless, but we have that availability. So, it's very high bandwidth because we know each other very well, obviously, and constant uptime on those exchanges. So, I think it's a big advantage.

Patrick: [00:47:39] What's her background and skillset?

Michael: [00:47:41] Well, I think she's the real founder of Bottomless to be honest. I'm like the type of guy who's just always coming up with lots of ideas that I think are all really awesome. She would always say, "Oh, no, that's horrible. No, no." When I just was obsessed with this for a long period of time, she said, "You know what? I actually think that could be it. You should quit your job and do this. I'm going to continue doing my job. You can try to make this business." If it was just me by myself, I don't know if I would have made the leap. So she's the person I would call like a woman of action. Those people are rare in the world. They're very rare. Maybe more rare than people who have unique insights.

Patrick: [00:48:18] I'm married to one, too. So I...

Michael: [00:48:20] You're married to a woman of action, you're lucky.

Patrick: [00:48:21] Yes.

Michael: [00:48:21] So we're the lucky ones. Honestly, I think that's the case. So, she has a great degree of focus, and that's never been my strong suit. So she's always been in the background, steering the ship in the right direction. Also just making sure that any vision that I have is actually being carried out by me. It's kind of funny, but that's actually the case. So her background, she's originally from Lima, Peru. She was born and raised there. She moved to the US when she was 18, alone, barely spoke any English at the time. She's done a remarkable job of building her life and her career.

Her most recent job, she was working at Workday and was a senior consultant. But her speed of progress in her career has been remarkable. That's why I really wanted her to come onboard at Bottomless full time. People's careers follow an S curve. If you can catch people before that first inflection point, where they're still accelerating upwards, and you can see it. She started at her first job out of college doing temp work and AP work and stuff like that. Just capturing that progress inside of Bottomless is very good for the company. But there is some friction to being a married couple. There's this tendency for people to think that if you're doing something with your spouse that is less serious. That's real. Obviously, for us, that's not the case. We're dead serious about this.

Patrick: [00:49:47] I'm always interested in marketing strategies. I know that you are, in some ways, like if you got 10,000 orders, you don't have 10,000 scales. So, you're somewhat constrained by the supply of scales, which is a good problem to have if it becomes a problem. But talk to me about how you think about growing this thing, and any unique ways that you have plans to grow this thing in what's become this like customer acquisition, consumer rat race, where you read Chamath's annual letter, and if he's right, 40 cents on the dollar of every venture dollars go into Google and Facebook for paid acquisition.

Michael: [00:50:21] Oh, it's absolutely true. I think that's the number one cause of startup death after the early years. Just your acquisition costs blowing up. Actually, you asked earlier what the biggest pushback people have. That's probably actually, the best pushback - it's like, how are you going to acquire customers at scale when this is almost impossible? For us, we have this magical quality where we can acquire coffee customers that break even, and then start pushing other products down the channel. So, we don't have to figure out how to constantly acquire - there's tons of potential for revenue expansion. So, once our customer acquisition cost blows up on coffee, we can really just start pushing out other products at basically zero acquisition cost just by adding more products to the store, and that's free.

I think that's the number one cause of startup death after the early years. Just your acquisition costs blowing up.

Patrick: [00:51:11] That's good math.

Michael: [00:51:11] Yeah, yeah. It's a-

Patrick: [00:51:12] Good long-term math anyway.

Michael: [00:51:13] Yeah, we'll see. That's basically the strategy which might make other people uncomfortable. "Hey, when are you going to start pushing other products down the channel?" I never give an actual answer. I'm like, "When customer acquisition cost blows up, you mean like at 100,000 customers, 1 million customers, 10 million?" I don't know. When customer acquisition blows up, when we can't profitably acquire coffee customers anymore, then you start pushing other things out.

Patrick: [00:51:37] So, I'll take a bookend approach to this. So, we started with some non-business ideas and then did some business for a while, and we'll close with a bunch of non-business or maybe business ideas but not pertaining specifically to Bottomless. So, a couple of the ideas of yours that I've enjoyed, one is this cycle of, I think you called it like the commercial loop or something somewhere, where you move from if something is hard for most people to buy, making it cheaper and cheaper so that you can sell in volume. Then when people are no longer sensitive to the price, you ramp up the, I think you used the word artisan, to increase the interest again at a higher price point, and then it just keeps going in that cycle. So, talk to me about that insight and any other places you've seen that happen.

Michael: [00:52:13] The original insight is just seeing it happen in beer. You had beer originally made in little breweries across the United States. It's basically artisanal. Then you have these big companies come in and produce it at scale for much cheaper. But then once those people with their huge scale economies have one final victory, once they've completely stamped out all of the small breweries, then suddenly, the landscape for the consumer is just boring. It opens that window to start providing more options for people.

So, then you see this boom in craft beer. But then now, you're starting to see mass produced craft beer. The end state is like a consolidation with a couple of the big breweries owning like a much more differentiated product line with all the options that people wanted. Eventually, it consolidates into some big players who have operational excellence. Then once they've won a final victory, there will probably be a new cycle. You definitely see this in the beginning of the marijuana markets where you have these massively distributed brands. You haven't seen the consolidation yet. I don't know. You can probably name an industry - it's just like a matter of knowing that industry well enough to be able to speak to it.

Patrick: [00:53:25] That same loop has happened. Yeah. What were some of the other ideas that maybe you got discarded, maybe your wife vetoed them as stupid?

Michael: [00:53:33] One that I would love to see built, so I'll put this out into the world, is a spaced repetition social network. I think social networks are underrated. I think right now, there's this massive backlash against social networks. I think it's because people haven't actually developed the ability to properly curate their social networks. The Facebook thing of just get all your friends was a great way to bootstrap a first social network, but it's definitely not optimal way to consume information. Then you have Twitter, which is better. But then it's still like follow all these celebrities that you have name recognition from the offline world, still not ideal, but it's getting directionally better than your friends and family. So that's besides the point. I think people are getting better and better at curating information. The problems with social networking is not the form factor of social networking itself, but with people's inability to actually properly handle it.

Spaced Repetition

Patrick: [00:54:22] So, what do you mean by spaced repetition?

Michael: [00:54:24] So let's get back to that. So, I think the big problem with social networking right now is we only have this model of things show up in your feed once, and then they disappear. That has a huge incentive for ephemerality. I almost don't want to see something in my feed that I'm going to want to see again. There's this optimization for one-off pieces of content. You definitely see that on Twitter. People will definitely post ephemeral stuff. I think you could make a much more meaningful social network if you had some sort of repeatability in the content. That's where spaced repetition comes in.

So spaced repetition is this way of, basically, creating flashcards is how people use the concept now to be reminded of something when they're most likely to have just forgotten it. So you'll make a flashcard. People use it for learning languages, for instance. You put the word you want to learn, and then it will repeat to you again the next day because you're most likely to forget it in one day, or be just forgetting it. So, it's an optimal time to be reminded. Then you probably don't need it again the next day. You maybe need it two days later, and then four days later, two weeks or whatever it may be. There are different algorithms. But basically, you start spacing out the repetition. That's a way for, basically, putting things into your mind forever, which I think is very powerful idea and makes things legible to yourself forever when you decide you want it to be in your memory forever.

So, it's like people spend a lot of time talking about putting computer chips in people's brains so that we'll remember stuff, but that's silly, Sci-Fi crap. But if you use spaced repetition, you can do that already. So, I do this all the time. I have my own spaced repetition app where I make notes to myself.

Patrick: [00:56:11] It kicks it back to you at some cycle?

Michael: [00:56:13] Yeah. I'll see it again tomorrow, and then I'll see it agan two days later, and then a week, and then eventually, I'll only see it once a year. I moderately remember these things.

Patrick: [00:56:22] Yeah. You actually choose what goes into that cycle?

Michael: [00:56:25] Yes. So the way my app that I use works is I just make notes.

Patrick: [00:56:29] Did you build this app or did somebody else?

Michael: [00:56:30] No, it exists. I think it's called just spaced or something. You can search spaced repetition, and there's a lot of apps. Spaced repetition itself is not really a new concept. I think it's just underused because you have to actually create these cards for yourself. Who wants to do that?

Patrick: [00:56:44] Yeah. It's like the habit formation you talked about, but for intellectual concepts or something.

Michael: [00:56:49] Yeah, it's like, I don't use it as much as I would like, because I have to remember to open this app and put these concepts downs. It sucks. So, I've always imagined this social network where you see a feed just like it's Twitter. But if you like it, or the equivalent of retweet would be putting into your own spaced repetition, so it shows up in your feed into the future. So, it creates a whole another incentive structure for people to try and put things out that are actually durably relevant for people. I imagine for my own various Twitter feeds, if I was trying to put something out that people would use in this app, the content would be different. It would totally be different. It would probably be better because people are trying to create something that people actually want to remember.

Patrick: [00:57:27] This makes me think of like - have you ever used if this, then that?

Michael: [00:57:30] Yeah.

Patrick: [00:57:30] Yeah. So, for example, you might say, I don't know if it's possible in that app. But if you say, "If I like a tweet, then put it in this app and put it in my spaced repetition loop or something."

Michael: [00:57:39] Oh, man, I should do that.

Patrick: [00:57:40] So, you can hack your way to it. It raises this interesting thought that ties back to Bottomless, which is a lot of these networks, we'll take Twitter, for example, are amazing in the sense, like the creators of the platform of the toolkit don't know the ways in which their toolkit is going to be used. Then their success is, in part, the ability to recognize interesting use-cases and magnify them. So, like the @ symbol in Twitter, or whatever.

I'm wondering if there's any application of that, where you think of Bottomless as somewhat like a marketplace platform business. Most businesses that have that feature that I just described are marketplace businesses. Have you thought about that at all, where somehow it's like... It's almost like there's a developer kit or something that allows them to take your relationship with the customer and the legibility concept, and then do whatever they want with it?

Michael: [00:58:31] Yeah, there's some potential future for Bottomless, where we create an API that uses our prediction technology that other people can build on top of. I totally think that would be a very valid route for us to scale. We have people all the time selling random stuff through subscriptions, saying, "If I have this, how much better would my business be?" You can sort-of hack on to emergent properties. In general, I think, business strategy, depending on emergent behavior is overrated right now. I know that that is something that Twitter did. But there's this saying in Silicon Valley that Twitter is like a clown car that fell into a goldmine. So maybe emulating them isn't the smartest thing. Amazon and Bezos basically had this business plan for exactly what they were going to do. I think that is a smarter thing to try to emulate. It's really popular to try and iterate your way to an epiphany, but you're just as likely to iterate your way off a cliff. In general, I think it's way overrated to think that you're just going to iterate your way into a billion dollar unicorn business.

Amazon and Bezos basically had this business plan for exactly what they were going to do. I think that is a smarter thing to try to emulate. It's really popular to try and iterate your way to an epiphany, but you're just as likely to iterate your way off a cliff.

Fundamentally, if you can reformulate a problem as an information problem, you can tackle it with information technology, if you get the right data. That's what we're doing at Bottomless. We're reformulating this problem of repeatedly purchased items as an information problem. We're getting the relevant data. You have lots of technologies that are maturing right now. It's not like we've seen that process stop. We've talked about; computer vision is coming to maturity, the cost of anything in the smartphone supply chain is continuing to collapse. So, there are a lot of waves that can be served to improve the world and improve people's lives. I think this concept that we're anywhere close to optimal or that somehow, these big tech companies are going to be the ones to surf those waves. I think that's just...

Patrick: [01:00:15] Doesn't really happen that way.

Michael: [01:00:16] Yeah, that would be historically anomalous.

Patrick: [01:00:17] Well, I love this idea of - I'll add it to my find an untapped supply source as one great business starting point - to find a previously unlegible or non-legible information source that's relevant to something broad and interesting. It's a great way to think about business opportunity. I'm curious, in winding down, if we go back to where we started, if you've learned anything about Twitter, specifically - I'm obsessed with this platform as really an incredible tool, it really has made a massive impact, at least, on my life.

I know, on a lot of people I know's lives too, in a pretty deep and meaningful way even though it is ephemeral and silly. All the jokes about it are fair enough. Anything you've noticed about building a presence on Twitter and elsewhere where no one even knows who you are in the accounts that you think speaks to what works there? So, is it pithiness? Is it being unique and different? Is it straddling the known and the unknown? What is in common if I sorted all your, let's say tweets, by the most liked or the most recognized on down? What would those things share in common?

Michael: [01:01:22] A snarky answer would be that the best tweets...

Patrick: [01:01:25] They're the least interesting...

Michael: [01:01:26] Well, no. The best tweets are the ones that seem contrarian, but are actually consensus because everyone's like, "Oh, look at this great contrarian idea." They boost it. There's a couple ideas right now that are becoming very popular. You shouldn't compete because that was in Zero to One, and everybody and their grandma read Zero to One in the startup community. So, all of the ideas in there are just actually very popular now.

Patrick: [01:01:48] Picked over.

Michael: [01:01:49] They're popular, but they still feel...

Patrick: [01:01:52] Like the person is the one discovering them or something.

Michael: [01:01:54] Yeah, they still feel contrarian. So like, don't compete, never compete with anyone. I see those tweets really blowing up all the time. So yeah, if you want to consciously try to build a big following, I would definitely suggest finding these ideas that feel contrarian but aren't. So I guess, the converse would be; the tweets that don't get a lot of play are the ones that are contrarian, but don't feel like it. I guess that would be the advice for how to try to consciously make good tweets. For me, I don't really try to do anything other than to be interesting. I don't actually try to tweet things that are true because it's almost besides the point. I'm not going to be convincing anyone of my version of the truth, I'm mostly going to be feeding people food for thought that might change them on the margins.

So, I think people don't optimize enough for saying interesting things. They're too scared to be wrong. There's a lot of interesting things that are wrong, but make people think. I think those are very undermined opportunities. You want to have followers that are interested in your content. So, you should be very unscared of losing followers if you think what you're saying is interesting, it's actually good to shed followers, it keeps your account interesting to the people that eventually will follow you.

Patrick: [01:03:08] Sure, sure. So two closing questions for you. The first is for what you think is your most controversial opinion, just generally speaking.

Michael: [01:03:17] Yeah, that's fun. The concept of equality is very important. It's generated a lot of progress in the world. We used to be a very unequal world. Obviously, you want to give everybody equal access to opportunity. But it gets confused with equal access to outcome. So, I think the idea of equality is something that people can't criticize. People are way too scared to criticize that. So, anything that's advanced that might make the world seem like a more equal place is generally backed by plenty of people, who are willing to advocate for it, and very few are willing to go on the record to resist that.

So, in general, any very popular idea that's probably good will be applied in a lot of ways that are really bad. I can come up with a couple examples of equality that are bad. We talked about education. Giving everybody an equal education means you're going to drop people at the bottom and people at the top. So there are ways to make education less equal that might actually improve outcomes for everybody on average. In the healthcare space, we have this idea that everybody needs to have healthcare which I agree. You shouldn't have people who have no access to basic healthcare in a rich society like this.

But we have this aversion to lower quality health care for people that we give them for free. I don't understand why we couldn't have health care for everyone, but more basic, but that doesn't exist because we have this aversion to anything that feels unequal. So, we need to pull everybody at the bottom up to slightly above the median. If you keep pulling people at the bottom up to slightly above the median forever, that just creates constant cost inflation. It's literally impossible when you think about it. When you think about it from the fundamentals, you can't constantly pull the bottom above the median in anything.

If you keep pulling people at the bottom up to slightly above the median forever, that just creates constant cost inflation. It's literally impossible when you think about it.

Same for income inequality, for instance. There's some degree of optimal income inequality. Who knows if we're there? I don't really want to weigh in on that. But I've definitely seen people advocating for increasing the wages at the bottom part of the distribution. I think that's a good thing. That's a very noble cause. But you can never get the bottom above the median. I actually was talking to somebody and be like, "You can't have these people making average salaries." They said, "Why not? They should be making average salaries." That's a road to destruction. Economically, it's just trying to do things that are literally impossible and mandating them from the top. So I think that's a very controversial opinion. I think it's something to explore. That's widely applicable to anything that in general is a good sentiment, it will be overrated.

Patrick: [01:05:58] My closing question for everybody is for the kindest thing that anyone's ever done for you.

Michael: [01:06:02] Yeah. So, I spoke earlier about the importance of having people who genuinely want you to thrive. I didn't have a lot of people like that in my life up until I met my wife. I didn't actually know that was a thing. When she came into my life, I suddenly had this incentive because somebody else wanted me to do well. I think wanting somebody else to do well, regardless of how that reflects on you, is fundamentally kind. There's a lot of people in my life that have helped me. Obviously, I'm very appreciative of that. But there's very few truly selfless acts. A lot of people help you because it helps them. But truly, wanting somebody to thrive and do better is very selfless and is something that I'm truly thankful to have in my life.

Patrick: [01:06:50] Well, I look forward to hearing you on How I Built This someday, hopefully. I appreciate you coming out to do this. This was great.

Michael: [01:06:57] Thanks, Patrick. I had a good time.

Created by @domcooke