Why Your Product Launch is Overrated with Superhuman Founder Rahul Vohra

In fact, I no longer believe in a traditional launch. Let's say you've built a new email client or a new task manager or a new calendar app. Now the surface area for these products is absolutely massive, bigger than almost any domain that you can think of. And what that means is you also get a massive surface area for bugs, as well as massive variability in how users will want to use the product. Now most companies would launch their app, and because demand for these products is so generally high, they would quickly get tens of thousands of users. But guess what? These users will find thousands of bugs and the company would quickly get overwhelmed. They would not be able to fix the issues fast enough. And so these users will become disappointed. They'll churn out of the product and then they'll tell everybody about their experience. And that, of course, is the very definition of a net detractor.

Welcome to a new episode of the Unicorn Bakery. My name is Fabian Tausch, and in today's episode, we will have a look at the recipe for success of Superhuman. Superhuman is a premium mail provider founded in 2015 by Rahul Vohra, who previously sold Reportive to LinkedIn and by now also invested in more than 100 startups, including AngelList, Descript, Supabase, and many, many more. I think I'll link the fund in the description because there are a few great other investments too, and we will talk about this soon in the, or at the end of the interview. Rahul and I will explore Rahul's approach to company building, focusing on product building, myths about fundraising, hiring, and also have a look into his efforts as an angel investor, as I mentioned. We will also talk about AI strategy as Superhuman recently got awarded the number one AI email provider by US News, for example. So I think a lot of topics to talk about. Rahul, welcome to the Unicorn Bakery.

Well, Fabian, thank you for having me.

Let's start quite easily with an easy introduction. One thing that you did when we— when you looked at what you built next after Reportive and your time at LinkedIn is email. And email for a lot of people is something that they use for free and they use and are like, it's a very, it's a commodity. Everybody uses email. A lot of them, a lot of people and me including for a long time used email as a free product, for example, Gmail, but a lot of other services and providers as well. And the first question that I had in mind when you thought about what do I do next is it's taken for granted by so many people. Due to all these free providers. So how do you define and test the need for a premium mail service? Because it feels like an interesting decision. Everybody thinks, okay, it's, it's free. Why should I pay for this? And then you coming around and saying, hey, I can do this better.

Markets where the solution is considered a commodity, actually, especially when it's free and powered by incumbents, are often the best markets to create a premium solution for. And it's actually precisely because people people take Gmail and Outlook for granted, and, and these products have to cater to billions of people, that we have an amazing market for Superhuman. So for those that don't know, Superhuman is the fastest email experience ever made. Our customers get through their inbox about twice as fast as before. They reply to their important emails 8 to 13 hours sooner, and they save 4 hours or more every single week. Many of them actually see inbox zero for the first time in years. And to understand where Superhuman came from, and how I tested into the idea, we actually have to wind the clock back by almost 13 years. Because in 2010, I started a previous email company called Reportive, and we built the first Gmail plugin to scale to millions of users. When people emailed you, we showed you what they looked like, where they work, their recent tweets, links to their social profiles. We grew rapidly, and 2 years later, We were acquired by LinkedIn. Now, during those 4 years, I developed a very intimate view of email. I could see Gmail and Outlook getting worse every single year, becoming more cluttered, using more memory, consuming more CPU, slowing down your machine, and still not working properly offline. But on top of this, people were also installing plugins like ours, Reportive, but also Boomerang, Mixmax, Clearbit, Yesware, you name it, they had it. And each plugin took those problems of clutter, memory, CPU performance offline, and made all of them dramatically worse. So we decided it was time for change. We imagined an email experience that is blazingly fast, where searches are instantaneous, where every interaction is 100 milliseconds or less. An email experience where you never actually had to touch the mouse, where you could do everything from the keyboard and fly through your inbox. An email experience that just worked offline so you could be productive anywhere. And an email experience that had the best Gmail plugins, but built in natively and which was still somehow subtle, minimal, and visually gorgeous. And so with those insights, we built Superhuman.

The question that I have for a lot of people out there who are maybe now thinking, as you said, okay, markets that feel like commodities or are perceived as commodity are the best to build premium products. The question that I have in mind is what are specialties about building a premium product in a market with a lot of competition and also free alternatives where the product is currently a commodity? What are the specialties that I have to think about as a founder to build a product that people actually are happy to pay for? Because if we look at Superhuman, we will find a lot of people, and we will come to that later, a lot of people happily advocating for Superhuman as well. So You somehow made it.

Superhuman is the extreme case of having a lot of competition. Literally everybody has an email address. Everybody has an email client, whether it's something they download or the web interfaces to Gmail and Outlook. So we had to differentiate from what people already had. And from my own experience with Gmail and Outlook, I knew that there are 1 billion professionals in the world and that we spend on average 3 hours a day in our email. That's a study that McKinsey did not too long ago that showed on average the 1 billion professionals in the world spend 3 hours a day on email. That's 3 billion hours every single day, or 1 trillion hours every single year that go into email. And yet before Superhuman, we were all using one-size-fits-all tools. I knew that nobody wanted to spend 3 hours a day in email, and I knew that nobody wanted to miss important emails because when they do, they miss opportunities, they block their team, and they damage their reputation. So we built a product that solved for these problems. Superhuman is blazingly fast, helps you reply to important messages way faster, and saves you 4 hours or more every single week. Now, those are big claims, and there's a lot of things that come together to make that possible. First of all, the UI itself is blazingly fast. As I mentioned, everything happens in 100 milliseconds or less. Secondly, our workflow helps you focus. Imagine, for example, being able to split your inbox into streams that matter most to you. Streams for your team, for your VIPs, and your most used tools, whether they're Google Docs or Notion or Figma. And third, Superhuman is the one email experience that makes the best use of AI. For example, imagine being able to instantly respond to any email with just a thought or a phrase. Our users now do this more than 25 times per week, and I can tell you it's pretty game-changing.

I haven't tried it, to be honest, even when I resubscribed to Superhuman, as I now have enough email flow to also user experience benefits. But let's talk about the product building side of it because it's a commodity. Everybody is used to having an email client. So I can imagine that building the first MVP and the first version that users will use and then also change from their previous provider to Superhuman can be very hard to build.

We were not building in a space that was very amenable to user feedback. And after all, we had to build most of an email client before users could reasonably switch over. Sometimes people ask, is this a minimum viable product? Is this a minimum lovable product? And between viable and lovable, we definitely leaned towards lovable. And I was user number one. In the beginning, my process was very subjective. I was trying to make the product create very specific feelings. Feelings of joy and satisfaction. And actually, when we sold my last company, Reportive, to LinkedIn, I explained this process to the CEO, Jeff Weiner. I said something like this: we make products like their games. The challenge is this: find the magic. It starts with a hunch. People think you're crazy, but you push to a prototype. Then the hunch becomes a heartbeat, faint and faltering. Some days you don't feel it at all. But then come the days when it all rushes back, beating so strong that it pounds in your chest. You cut, refine, and polish, go back and do it some more. And then eventually your prototype grows up. Now you can't stop playing it. Everyone who sees it wants it. Everyone who plays it loves it. You just found the magic. And once I'd felt the magic, once I knew that the product created feelings of joy and satisfaction, then we started to put it into the hands of users. But of course, that alone isn't enough. We had to make sure that Superhuman was also highly reliable and had enough functionality for people to actually switch over from Gmail.

Two questions. On the one hand, how long did it take until you felt this magic? And then also, yeah, let's start with that.

It actually took several years, which was one of the hardest things about Superhuman. And also, in retrospect, one of the things I'm most appreciative for because it led directly to the development of one of our other big contributions, which is the product market fit engine. In the summer of 2015, we started much like any other software company by writing code. In the summer of 2016, we were still coding. And in the summer of 2017, we were still coding. And I felt this intense, incredible pressure to launch both from the team, but also from within myself, because after all, my last company had launched, scaled, and been acquired in less time. And yet here we were 2 years in and we still hadn't launched. But deep down inside, I knew No matter how intensely I felt that pressure, a launch would go very badly. I did not believe we had product-market fit, and I didn't believe that we had enough of the magic. And although I knew it, I couldn't just say that to the team because these are super ambitious, hyper-intelligent engineers. They'd poured their hearts and souls into the product. So I needed a plan. And in April of 2017, I started my search for the Holy Grail, for a way to define product-market fit. For a metric to measure product-market fit, and for a methodology to systematically increase product-market fit. I searched high and low. I read everything I could find, spoke with all the experts, and I came across this guy, Sean Ellis. Now, Sean ran growth in the early days at Dropbox, LogMeIn, and Eventbrite. He coined the term growth hacker, and Sean found a leading indicator of product-market fit, one that is benchmarked and predictive. Just ask your users this: how would you feel if you could no longer use the product? And measure the percent who answer very disappointed. After benchmarking hundreds of startups, Sean found that the companies that struggle to grow always almost get less than 40% very disappointed, and the companies that grow easily almost always got more than 40% very disappointed. In other words, if more than 40% of your users would be very disappointed without your product, you have Initial product market fit. Now this metric is more objective than a feeling. It predicts success way better than net promoter score. And it's not only the best way to measure product market fit. You can use it to develop your very own product market fit engine. And with this engine, we now have a methodology for systematically increasing product market fit. It even automatically generates our roadmap for us, and one that is guaranteed to increase products market fit. So there's a lot more there than we can easily cover in the time we have. What I would say is if the audience hasn't come across this products market fit engine before, simply Google for Superhuman products market fit, and you'll find it.

Happy to link it in the show notes. It's a very good article and definitely a great guide to really understand, okay, how can we build this for our own venture? Quick question on the, on the product market fit engine. How does it scale when the company scales? Like, is the framework the same? Do you have to adjust it? You're in now at a completely different stage than in the early days. Do you still use the same measurement? So how does the product market fit engine scale and age over time?

Yes, we are indeed still using it. I will say it's most useful when you're starting a new product, when you're still shifting your market at the same time as shifting your product, but it also continues to be useful throughout the entire lifecycle of any product. And as you grow, it can be tempting to ignore the process. And just iterate on feedback. That's very risky because as you grow, that feedback will start coming from parts of the market that strategically probably don't make sense. So you always need a way to figure out who to listen to and who not to listen to. And that ultimately is what the product market fit engine gives you, a methodology to figure out who to listen to and to systematically increase your product market fit score.

And when we, when we look at that, because I think the question that comes to mind that we had right now, you said, okay, you build for a few years, you were your first customer, you had to be satisfied with the solution before you wanted to give it out to others. How would you differentiate? And then you build Reportive before, which you scaled and launched and scaled very quickly. How would you, as a founder, evaluate, hey, I am in a position where I need to build for 1, 2, 3 years, or I'm in a position where I can build, launch, iterate very, very quickly.

I think it all boils down to the complexity of your product, the ubiquity of the use case, and the surface area for things like performance, quality, and reliability. Reportive, I was able to build the first version single-handedly in about 2 to 3 months, got it to 10,000 users overnight and 30,000 users over the course of the next 2 or 3 weeks. And that's because it was a very simple product. With a ubiquitous use case, with a very low surface area for bugs and complexity. Of course, building an email client is entirely different. And I've seen a lot of companies, especially other productivity and collaboration tools, get this very wrong. In fact, I no longer believe in a traditional launch. Let's say you've built a new email client or a new task manager or a new calendar app. Now the surface area for these products is absolutely massive. Bigger than almost any domain that you can think of. And what that means is you also get a massive surface area for bugs, as well as massive variability in how users will want to use the product. Now, most companies would launch that app, and because demand for these products is so generally high, they would quickly get tens of thousands of users. But guess what? These users will find thousands of bugs, and the company would quickly get overwhelmed. They would not be able to fix the issues fast enough. And so these users will become disappointed. They'll churn out of the product and then they'll tell everybody about their experience. And that of course is the very definition of a net detractor. In my experience, it's significantly better to do what we did, which is onboard customers at a measured pace each week. And that way you have the bandwidth to fix any issues they might find and then to focus on making them exceptionally happy.

I think that's interesting. So, um, for everybody who I didn't follow Superhuman for a while, it was the case that I had to sign up for a waiting list and either get invited by someone who already uses the product or wait a little until I am able to join and use Superhuman. And I think you did this until 2023, if I'm correct. I'm not 1000% sure if it was 2022 or 2023 until I was able to sign up myself.

That is a great question. I believe it was around the 2021, 2022 era. When we finally retired the waitlist. And a lot of people think about the waitlist and take away the wrong lesson from it, thinking that perhaps we did it to create the fear of missing out or to create the nightclub effect where you see a long line of people wanting something and therefore maybe you want the same thing, perhaps just a little bit more. But that actually couldn't be further from the truth. That's not the reason why we created a waitlist. The reason we created a waitlist is because we were simultaneously working on things like product functionality, product quality, product performance, product reliability at the same time as onboarding people. At the beginning of your product, you'll attract early adopters and early adopters by definition are okay if you're not 10 out of 10 on some or even all of those factors. But eventually you'll have to cross the chasm. Eventually you'll be going to the early mass market and then eventually the mass market, those folks are far less forgiving. And especially for something like an email client, it's mission critical. Let's say for the sake of argument, you drop one email, you send an email and it doesn't arrive, or even worse, perhaps you send an email and it arrives twice. Well, now you've lost trust. And I think there's this famous saying, trust takes many, many years to build up, but can be lost in an instant. We were very aware of that all the way along. And so it was so important for us to make sure that the application experience was fully dialed in and phenomenal. So with the caveat that we never really launched Superhuman and that it was just gradually happening over time, there were many things that we looked at as we started to peel away the waitlist. Things like product quality, which I just mentioned, were we staying on top of issues? Things like product sentiment more generally, including subjective measures of speed and power and reliability. Things like user activation, were users starting to use the product? User retention, were users staying in the product, user virality, were users referring other users, and then ultimately dollar retention and dollar expansion. Were users continuing to pay for Superhuman and were they growing their teams? The year when we finally peeled away the waitlist, I think we'd took care of the last two main reasons why people couldn't use Superhuman, which was number one, Superhuman for Outlook, and number two, Superhuman for Android, both of which now exist.

Okay, fair enough. I think what's interesting here is, so you also raised money from venture capitalists and sometimes it feels like VCs and investors want to see as much growth as possible as early as possible. That's a bit, I know that's a bit superficial right now. And knowing that you were a second-time founder or at least serial founder who already sold a company and experienced a lot, maybe there's a different trust than in a first-time founder. But did you, how did you communicate it to your investors that you decide, okay, we will onboard just a certain amount of people at the moment because we think it's better than letting everybody in? Is it that investors understood or did you also have to manage expectations there and convince or at least overrule in quotation marks some other opinions?

I never really had to explain or convince why we were doing what we were doing. To our investors. It just made sense. It made sense to all of us that email is an extremely, extremely hard product to build. I mean, look at every single other email startup. The last several decades is littered with the corpses of basically every email startup that has ever raised venture funding. And so we all knew that we had to do things differently. And this was just one of many things that we did differently. The second thing, of course, that we did is pick our investors. We picked extremely patient investors who have a very long-term investment horizon, who are interested in absolutely massive outcomes. And oftentimes the way that you get there is the combination of hustle. Yes, you have to go fast, but patience, patience, because you have to do something special. You have to do things different and you have to build a product that has stickiness and staying power. We're now at the point where 8 or 9 years into the Superhuman journey, the product itself is this incredible moat. So much energy, so much hours, so much effort has gone into building the thing and it's so unique and it's so different that it would be basically impossible for another company to rebuild what we've built, at least without putting the effort in. It's not necessarily rocket science, but it does certainly take a very long time.

And what happened from the day that you, in quotation marks, officially launched? How did the growth change? Did you convert the waiting list and people were like, yeah, okay, we're all there. Did something unexpected happen when you then publicly announced, okay, hey, everybody can use it now?

We never actually made a public announcement that everybody can use it. And even in switching over from waitlist to non-waitlisted, from concierge VIP one-on-one onboarding to what we do now today, which is self-service by default, but anyone can opt into the concierge onboarding. All of those things were gradual. For example, if we look back at 2023, we launched a lot of things, including Superhuman AI, Superhuman for Windows, Superhuman for Android, AutoSummarize, Superhuman for Teams, and more. But I wouldn't say that we quote unquote launched in 2023. Building a startup is always a constant process of improving, iterating and launching. It is never this binary effect.

Do you, or what do you think are the biggest misconceptions about launching a product that many founders have? You're investing in a lot of founders, so you might hear quite often, hey, we're launching tomorrow. So what are the most common misconceptions about the launch?

I think the biggest misconception is the launch is a thing that will bring in this influx of users that will then turn into an ongoing stream of users. At best, it's a one-time spike. Let's say you're lucky enough to get press coverage. And even that spike usually has very low-intent users. They may be early adopters, and early adopters have pros and cons. The pros, of course, are, as I mentioned, they have generally lower requirements for things like reliability and performance. Although the flip side is they really love it when products have those things. But the other side of the coin is they're also quite fickle. Just as they moved on to your new shiny thing, they may well move on to the next new shiny thing. And many founders don't realize that. So to do a great launch, you need to ideally have a product that really solves a real problem, is high quality, is highly performant, and will retain those users over time. And then in addition, you need to build a company with go-to-market or distribution channels that continues to acquire new users over time as well, because a launch tends to be a one-time affair.

I see this quite often. I think, was it Brian Chesky who said, okay, we launched so often because nobody cared about our last launch that we just decided, okay, we can launch Airbnb again?

Yes, I think I've heard that in podcasts of his. And I completely agree. No one will care for most of the things that you launch. So what do you do? Well, keep trying, keep trying to get as much noise and attention behind every launch that you do. But the corollaries fundamentally do not rely on launches to be a growth strategy. Do not rely on launches to bring in high-quality users. If you're lucky, they'll do some of both of those things, but they are not a user acquisition strategy.

Fair enough. One thing you said, okay, you have to have the magic and to feel it while you use the product. And you are a game designer by heart. And I often, when I talk to founders, I often hear, okay, yeah, we have to gamify our product a bit more and we have to make it more engaging. And when I listened to a lot of your podcast interviews and I listened to many, many of them because I wanted to dig a bit deeper in some questions. Just to start, what is the biggest difference between game design and gamification? And what is relevant for which use case?

Well, I would start by saying this: gamification does not really work. Game design works, but game design is not gamification. This is not simply taking your product and adding points or levels or trophies or badges. And to understand why gamification does not work, we have to understand human motivation. And there's a very interesting study from Stanford that demonstrates the difference perfectly. In the 1970s, the Stanford researchers recruited children aged, I think, about 3 to 4 years old. And all of these kids were generally interested in drawing. Some kids were told they would get a reward, a certificate with a gold seal and a ribbon. And some kids were not told about any reward, so they did not expect one or even know of one. And each child was then invited into a separate room to draw for 6 minutes, and afterwards they would either get the reward or not. Now, over the next few days, these children were observed to see how much they would continue to draw by themselves. The children with no reward spent 17% of their time drawing, but the children who expected a reward, well, they only spent 8% of their time drawing. The reward had halved their motivation. So what's happening here? Well, researchers differentiate intrinsic and extrinsic motivation. With intrinsic motivation, we do things because they are inherently interesting and satisfying, whereas with extrinsic motivation, we do things to earn rewards and achieve external goals. And that's the problem with rewards. They massively undermine intrinsic motivation. And that's why gamification does not work. When gamification does look like it's working, I actually think it's because the underlying experience was already a game. So the real question becomes, what is game design? And as you know, I've been obsessed with game design my entire life. As a kid, I learned how to code just so I could make games. And before I was a founder, I worked as a game designer. And as a founder, I've gone deep into the principles of game design. And as it turns out, there is no unifying theory of game design. To create games, we need to draw upon the art and science of psychology, mathematics, storytelling, and interaction design. And at Superhuman, we've identified 5 key factors to consider. These are goals, emotions, toys, controls, and flow. And across these, we've identified many principles of game design. One principle would be make fun toys and then combine them into games.

It's— can you name a few examples? Because I think it's super hard for people to understand, okay, how do I use game design to make emails better? And I think of course one way would be using it, but at the other hand, maybe we can throw in a few examples right now.

Sure. Well, let's take the principle of making fun toys and then turning them into games. Let's actually start with this question. Are toys the same as games? Because they do seem different. We play with toys, but we play games. A ball is a toy, but football is a game. And as it turns out, the best games are built with toys. Why? Because then they're fun on both levels. The level of the toy, and the game itself. In Superhuman, a favourite toy is the time autocompleter. This is the thing you use to snooze emails and set reminders. You type whatever you want, it can be gibberish, and it does its best to understand you. For example, 2D becomes 2 days, 3H becomes 3 hours, 1MO becomes 1 month. And the time autocompleter is fun because it indulges playful exploration. What can it do? When does it break? How does it work? It's not long before people start doing this. Hmm, I wonder what happens if I keep typing 10? Well, it turns out that's October 10th at 10:10 PM. Or how about a sequence of 2s? That's February 2nd, 2022 at 2:00 PM. And then you start typing more complex inputs, like in a fortnight and a day. And it's not long before you find pleasant surprises. For example, timezone math happens without you thinking about it. 8:00 AM in Tokyo is 8:00 PM Eastern. And then most people are delighted to find out that if you really want, you can snooze emails until never. I.e., you can literally type in never and the email will never come back. So I would say to our listeners, Consider the features of your own product. Do they indulge playful exploration? Are they fun even without a goal? And do they create moments of pleasant surprise? If so, congratulations, because you have a toy and you're well on the way to building a great game.

Does every product have to be a game and therefore has to include several toys that I can use? Or is it only for, let's say, email providers.

Oh, it's definitely only not for email providers. I would wager that the best products that we use, the ones that we have the greatest emotional connection to and make the biggest impact on our lives, those have been designed with game design in mind. And toys are just one of the factors to consider. We also strongly consider goals. Games have great goals and there's definitions for what makes good goals. Emotions, controls, and flow. And if you think about any great product, goals, emotions, toys, controls, and flow have all been taken into consideration.

Fair enough. I think it's just helpful to, to open that. I think, would you say it's still relevant if I'm thinking about like game design, if I'm thinking about super, super specific B2B software for the enterprise level of customers, or is it because you started with one person in mind first and then build out into teams just because people and founders might have in their head like, yeah, but it's not applicable to what I'm building.

I don't think it matters whether you're building for teams first or for individuals first. But maybe some factors to consider are how much time people spend in your product. The, for Superhuman, 58% of our users spend 3 hours or more every single day in Superhuman. It's a lot of time. It's more time than they spend in any other product. So for that reason, making it an enjoyable experience, making it fun to use, making it playable almost is very worth the effort. It may not be true for something that perhaps you only spend 3 minutes a day in.

Fair enough. I think that's a very good measurement to figure out, okay, how much effort do I have to put into the product development? Anyways, it's always customer first in mind and then figure out from there what is necessary and, and, then also for an MVP, for the lovable product, and then for really the proper solution. Talking about proper solution, I think one of the big questions, and you talked a bit about it already when you said, okay, we were only ready to publicly launching or making it available was when we also launched Superhuman for Outlook and more of the market widening steps so that people can use it in a broader distribution. Instead of only focusing, okay, new features, new features, new features so that the Gmail people and users could use it better. How do you trade off solution deepening versus market widening when you build your product? When do you have to focus on solution deepening and when is the time for a broader market? And therefore, for example, in your case, not only delivering Superhuman for Gmail, but also for Outlook.

Great question. And I'm very aware that we're using internal Superhuman jargon here. So I'll go ahead and define it for folks. Solution deepening is when you make your products better for its existing users, but you don't make it accessible to more users. And market widening is the opposite. You're taking your existing product and you're making it available to more users, but you're not making it any better. It doesn't do any more. Now, in the book, the authors, Gabriel Weinberg and Justin Mez of the book Traction, have a great rule of thumb for this. Spend roughly 50% of your resources on solution deepening and roughly 50% of your resources on market widening. And we codified this in our product market fit engine. If you follow the engine, then pretty quickly you'll deeply understand why users love your product. And you'll pretty deeply understand what holds users back from loving your product. To increase your product market fit score, I think you should spend half your time doubling down on what users love. In Superhuman's case, that means even more speed, even better shortcuts, even more efficient workflow, even more powerful AI, and an ever-better design. But just as importantly, you should spend half your time systematically addressing what holds users back. Because if you only double down on what users love, as vision-driven teams tend to do, you will not increase your product-market fit score. Whereas if you only address objections, as data-driven teams tend to do, then a competitor will eventually overtake you. So by following this process, you can actually automatically create a new roadmap, one that writes itself, balances a market widening with solution deepening, and one that also increases your product market fit score.

Would you say there's ever the point where product, the product market fit score is like, let's say, 100, or is there always something to do?

Oh gosh, there's always something to do. I think Andrew Chen called this the law of shitty metrics, which is every metric you have, no matter how great they are, will always come down. And if you read the original product-market fit engine thesis, it's pretty clear to see why this is the case. As you expand your market, as you market to widen, to use that phrase, you'll be encountering users who aren't necessarily early adopters. Perhaps they're not quite the perfect use case for what you'd originally built. Perhaps they need, or they have more challenging environments or more challenging circumstances. In our case, we had all of those. We were going from Gmail users to Outlook users, people who are on average more technically savvy to people who are on average less technically savvy, from iOS users to Android users, a homogeneous environment to a heterogeneous environment, from early adopters to the early mass market, people who are highly tolerant of faults to people who are not tolerant of faults at all. And there were many other reasons beside. So no matter how great you think your metrics are, every good metric will come down. And building a company is the grind of constantly working on all of these fronts.

I think it's on the one hand that you broaden your audience, so therefore new customers, but also when we look at recent events, it doesn't feel recent anymore, but with the launch of ChatGPT and therefore new technology coming into the market, this also defines the playing field completely newly and you have to adapt to that. So also the existing customer set wants different things because of new technology. Or could be of service due to new technology and also disadjusts your product market fit score once in a while. It doesn't happen every day, but I think also that has to be taken into account.

I agree.

Yeah. So given that US News, I mentioned it in the beginning, selected you as the number one AI-powered email client, how did you approach the topic of AI? When did it come to your mind the first time and when was it the the moment that you said, okay, now we have to implement it. And then we are again in the loop of, okay, we have to implement it, but we have to do it in a minimum lovable way. So we also have need some time to develop it. How long did it take?

Well, we were actually relatively late to the party, but once we got started, we got moving very quickly. We started work on superhuman AI in February. Of last year. And so if you recall, I believe ChatGPT had already reached 100 million plus users at that point. GPT-3.5 Turbo was already out. Lots of incredible things were happening. Amazing products and experiences were being made. And we'd literally just started to build. Now you might think, gosh, isn't that really slow? But when you look at the landscape of email clients and email applications, because these things are just in general so hard to build, we actually were still the first email experience to come to market with these latest AI technologies. And there were a number that we were building at the same time. So for example, with Superhuman AI, you can now write entire messages effortlessly. You just jot down some phrases and we'll turn them into a fully written email. And best of all, the email sounds like you do. Superhuman AI matches the voice and tone in the emails you've already sent, and it applies that to everything it creates. I know you mentioned you haven't yet used it. I highly recommend you try. Our average user is now using this 25 times every single week. And so it's saving people vast swathes of time. As part of Superhuman AI, we also launched autocorrect, which fixes typing errors as you go. Extra characters, missing characters, transposed characters, missing punctuation, incorrect capitalization, to name just a few. Our autocorrect now beats Gmail and Outlook across multiple benchmarks. And the net result is you type faster, much, much faster. So with Superhuman, you can actually type 30 to 50% faster than before. Last year, that was just in English. Recently, we launched English— sorry, Spanish and French. And we now have German in private alpha, which should be coming out in the next few weeks. And most recently, we launched Auto Summarize, which is our first AI feature that you don't have to remember to use. You'll see a one-line summary above every conversation. And as new emails arrive, that updates instantly. AutoSummarize has actually changed how we work. We almost always read the one-line summary first, and sometimes we don't even need to read the email. So again, as a result, we're moving faster than ever before.

I love that. But I think one of the— like, good to hear that German is coming because that helps me a lot. Then I can test it even better. I'm of course writing emails in English, but still. One thing that I am interested about The one hand is how you approach AI features for Superhuman. I think that's something we understood and something that I really like. I think the more interesting question will be for a lot of founders, how do you think will AI influence how we build products in the future? How do you use it inside the teams to maximize productivity and creativity? It's not only productivity. I think it also helps on different other aspects of life as well. So how do you supercharge the Superhuman teams by letting them use AI, not only in the Superhuman products, but also in the product building phase?

Oh, that is a great question. I think it boils down to creating a company culture where you talk about uses of AI and you celebrate its clever application. So we have Slack channels that are dedicated to this. We have one, for example, called AI-Ideas, where we see and put the best uses of AI that we've seen in other products into that channel so we can share them. When someone in the company does something particularly cool with AI, again, we share it and we talk about it. And we've made AI one of the biggest parts of our company, one of the pillars of what we're building. So that again, the culture is to use AI in the work that you do, not just in the products that we're building.

What are your personal favorite AI tools besides the features that, Superhuman launched for, for email?

There are so, so many. I think one that I'd call out is you've mentioned creativity, the ability for ChatGPT to creatively help with writing. So outside of work, I still am an avid game designer and I still do a lot of gaming. In particular tabletop roleplaying, and I'm part of a very longstanding D&D group. As part of this D&D group, one of the things that my character had to do was be very proficient in the lore, the history of the world in a very deep way. And so I actually, actually had to write it. In this case, it was 10 different ancient dragons and in detail all of the information that you could imagine for those. Now, that would have been a lot of work. To creatively come up with 10 different unique personalities, 10 different legendary powers, 10 different abilities, weaknesses, rumors, and so forth. But with ChatGPT, it was just half an hour of work. Of course, I had to edit it and tweak it and make it feel less derivative, but the scaffolding was provided for me immediately. Another example, and this is again a ChatGPT example, would be the analysis mode. Where I've used it for many work use cases where I download a spreadsheet and I upload the spreadsheet and I work with it as an analyst. And I say, well, you know, these are the insights I want to get. Yes, I could fiddle with it in Excel or in a SQL query tool, but it's way, way faster just to converse in natural language with an agent and have that working in the background as I'm multitasking on other things.

Yeah, I think there are so many great use cases and also so many specific tools for specific things. I'm the only, the thing that I'm very curious about is like what tools will become standalone tools and one will be, which ones will be integrated as features into some products and are currently single products, but still might be features at some point. I think there's, because you are an angel investor, like with, with your fund and doing so many investments, you also see a lot of pitch decks and therefore probably by now at least tons of them, not to name specific numbers, but tons of them have AI somewhere listed there. And I'm very curious to see which ones will be standalone products, which ones will be delivering solutions enhanced by AI, but AI will just be part of it instead of being the thing, because a lot of companies currently advertise AI as the thing that makes their product assumptions and also USPs. And I guess This will be very, very interesting to see who survives as a standalone product and which companies will turn into features either by being bought or by others just developing it and integrating it into a product with more mass distribution.

I agree. From an investing perspective, I think there are relatively few new AI ideas that are possible that were not possible before. And that most of the value will accrue to the foundation models, to the large organizations that are creating those models, and then to applications that have some other way to get distribution. At the end of the day, it's not cash that is king, although of course that is extremely important. It is distribution that is king. And I think Superhuman is a great example of a product that had a fair amount of distribution, has a fair amount of distribution, and is able to leverage these new technologies to build something entirely new. Uh, and so it was actually really funny. I think Forbes called us an incumbent in the, in the email space in a recent article and used us as an example of a company where AI is not creating a new product opportunity. It's making an existing product even more sticky and even better.

Yeah, I think that's a good example. I have one topic that I never heard you really talking about, and I think is, um, interesting. You mentioned that you wrote, for example, the product market fit framework and the engine. And I guess this helped you to a lot of credibility and also personal brand in the ecosystem. Not that you didn't have it before by building and selling Reportive and a lot of other things, but you are yourself, in my opinion, a very renowned personal brand in the ecosystem. And not by oversharing and being loud all the time, but by being very precise and concise in what you share and what you don't share. What are the positive and maybe also negative effects of becoming a personal brand as a founder?

Gosh, well, I can't really think of any negative effects. Fortunately, I'm not, not well known outside of a few very specific circles. If I walk around San Francisco or I go out in San Francisco, I do get stopped out and about. But again, it's not a problem. People just generally want to say hello, or they want to say thank you for the product-market fit engine, or maybe they have a question, or maybe they want to talk about Superhuman, or my favorite, they want to give me product feedback, in which case I'm 100% all in, all ears, because that stuff is invaluable. But really, outside of the Bay Area, it doesn't really happen. So there are very, very few negative outcomes to that that I can think of. And of course, there are plenty of positive outcomes. You know, investors are more likely to believe that you can do what you'll say you'll do. It's easier to hire candidates. It's generally speaking easier to get press. And zooming out, I think this is, and you know, there are plenty of other people who've done this far better than I have. I'm but a small example, but it's a good example of power of compounding effort. I knew I wanted to be an entrepreneur since I was 15 years old. And almost every single thing I've done in my life, but I've had a few detours, like I started a PhD and dropped out, but even that was to start a company. All of the things were in pursuit of being a founder. And so when you pursue something doggedly, the effort really does compound over time. And so in my case, it's turned out, one of the valuable things has turned out to be personal brand. Uh, but for, I think for every founder, it's different. And every founder finds a way to make their efforts compound.

What are the measurements for yourself to decide, okay, do I put out content about this or not? Because I think once you had an article like the product market fit engine, that pretty much everybody who starts a company, especially in the US, but also all over the world starts getting recommended and reading. And I think it. For me, sometimes it creates like a different scale, super small and niche here, mostly in Germany, sometimes Europe, US, but with the podcast or something. If I, if I see, okay, a podcast performs super well compared to others, it sometimes feels like there is a bit more like a bit more weight for me to create the next piece of content that might achieve such success. Do you have that when you think about Okay, how do I present a new framework? How do I present a new topic? Because in most of the podcasts that I listen to, everybody wants to talk about the product market fit engine, for example. Then there is gamification versus game design. So there are a few topics, but how do you think about what topics to position yourself with?

It was really those areas where I genuinely felt like I have a extremely unique insight or a way of presenting or codifying the information to the technology audience such that they hadn't seen it before and that it would have massive impact. And I think both the product market fit engine and the game design work are examples of that. They're very unique. I happened to be a professional game designer before I was a founder. And so I was able to translate that world into Silicon Valley language. And I don't know how I stumbled across or invented the prompt market fit engine. I mean, it was really just bashing my head against the problem and applying my computer science mind to it. But once I'd figured out the algorithm, I was like, wow, this actually really does work. And we can start to turn what has historically been an ineffable art into a science. That's the bar. It's not really starting from the perspective of, well, you know, let's build something that every founder will read. That was a very valuable side effect. Really, the primary motivating factor was I've got this thing and everybody needs to read it. So we have a few more of those things in the pipe. I'm not going to share yet exactly what those are because they're still ongoing experiments internally. But if the internal experiments do work, then I think that we have some exciting frameworks that are yet to come.

Lovely. That sounds great. As you mentioned, you're thinking in frameworks, and I really like that because you structure your thoughts very well and you are very good with that. One question that I have, because you're also hiring for Superhuman, or hired a lot and are still hiring, you're always trying to figure out how the people you work with think, what makes them tick. What their correct characteristics are. So what are your favorite questions to ask to uncover personality traits and characteristics of a potential hire?

Interviewing is perhaps the area where I'm least rigorous. Over time, I've just realized that I'm a very intuitive interviewer and that there are other people in the company who have a much more rigorous approach. And it's the combination of those that has allowed us to build the team that we have today. For me, I'll often just know when a candidate is likely to be great and when they are likely not to be great. And I have learned to listen to that sense. I think that alone would be dangerous. Of course, it can lead to things like bias, and which is why we pair that with a very rigorous set of interviewers in the rest of the company. But one interesting question that I like to ask is, what is other people's biggest misconception about you? Because 9 times out of 10, it turns out to be true.

That's a good question. I think I had to answer that once in a questionnaire, and I think what I wrote was that people think I'm extroverted because I get to talk to all these founders on the podcast and meet so many new people. And I said, hey, I'm probably situationally extroverted, but most of the times also need the time for myself, and I'm like also an introvert. So thinking about the question helps you understand yourself, sometimes a bit better because I think most of the people who listened to this question right now were like, yeah, what is it? Because they haven't thought about it yet. So I know how long it took me to come up with an answer that I was happy with that was not like just a superficial thing, but a genuine answer. So I really like the question. When you interview and you say you're an intuitive interviewer, are there red flags that are things that you're like, oh, these are things we definitely don't want to have in the company and we we can't work with?

Plenty, and many of them are in fact obvious. For example, if somebody is displaying a victim mindset, let's say, about their previous role, they will more than likely act that way at your company. But some red flags are less obvious, and some are just a reflection of us and how we want to build our organization. If somebody says the main reason they want to work at Superhuman is because they really value working with fantastic people, Well, depending on the role, that can also be a red flag, but a less obvious one. Now, don't get me wrong. We have fantastic people and we are blessed that folks want to work with them. But for certain roles, I found that candidates are much more likely to succeed when their primary motivation is the mission itself. When that person will live and bleed for our customers and for their problems and for the things that our customers are trying to do. In some roles, that's more important than the love of working with fantastic people. And so I get pretty deep into those factors. And that's one of the things that I look for, uh, depending again on the departments that we're hiring into. Building a company can be rewarding, but building a company is almost always very hard. And what I've found is that the best performing candidates always have a deep and an enduring passion for the mission.

Do you only bring in people that have a passion for the mission, or do you have to make exceptions sometimes, or is there other different grayscales here?

Oh, there are definitely grayscales. Like I said, it heavily depends on the role and the department. It depends on the seniority and it depends on the hiring manager. I'm talking about me as a hiring manager. So the people that work for me, I need to be absolutely mission-driven, but that's just me as a manager and me as a leader. We have hiring managers in the company who don't require that quite so much as I do. And again, that's okay. It really comes down to knowing your own needs and your own preferences as a leader and as a manager of people.

Fair enough. And quick roundup because we will run out of time if I ask all my questions, but going to your job as an angel investor from time to time, we already stated that you're a framework person. So the question that I have is how do you approach investing in startups? What is your framework? For investing, evaluating startups? How do you look at all of those?

I look for 3 very specific things. First, I look for founders who have the following magic combination of they know how to make something people want and they know how to make people realize they want it. Because if a startup only has one of these, we've talked so much about distribution, it will unfortunately not be able to succeed. Secondly, I look for founders that demonstrate exceptionally high value of grit. And I think of grit as the combination of passion and perseverance. Passion means that the founder will not easily get distracted with new interests or goals. And perseverance means that the founder will follow through on things despite very difficult and hard challenges. And the founders that are both persistent and relentless in moving forwards are the ones that make their startup successful. And then thirdly, I look for the possibility of a billion-dollar outcome. Not all products and not all markets support that. And many times when we end up passing on great founders with a good business, it's because we were not able to get to conviction that the business could support a billion-dollar outcome.

I also think that's a question that often founders have to ask themselves because if they want to raise money, they need to deliver that. And often you don't get money if you're not able to show that you're trying to deliver it. But still, If I talk to founders on, hey, why do you want to take investment? They are often not aware that they have to or should build such a company when they really raise money, especially from venture capitalists. So super important question for everyone to ask themselves. And out of personal interest, because throughout the years, and I think there are prominent examples also as podcasters who did it, people told me, hey, you should start a fund, you should start investing into startups, probably more as an angel fund, less than a like concentrated portfolio with competing for the lead in a round. If you were in my position, and we will not judge if the deal flow that I would have or anything is, is well enough, but more out of my position with there is distribution, there is probably the possibility to get capital for a fund. What would be your questions to think about for me starting an angel fund and also your tips and best practices from investing in 100+ startups?

Well, the first question I would ask is, do you feel like you have an unfair advantage in being able to deploy capital? And I think we've seen from folks like Harry Stebbings and many others in the industry, that having a well-respected, widely downloaded podcast is indeed a big advantage. And if you can parlay that into deal flow, especially given your location, you probably have access to deals that other people don't necessarily have access to. You might be able to make some really interesting investments. So let's take that for granted. Let's assume that you actually do have access to unique deals. Let's also assume that you are good at picking because for whatever reason, you're a good founder yourself. You stay up to date on how to pick and where the opportunities are. The only other thing I would point out is raising a fund then requires time, significant time. And the more that you raise, the more time it takes to deploy. The only way that I've been able to be an investor is to do it with somebody else. My full-time job is running Superhuman. It is more than a full-time job. It is, of course, being a founder, it's very busy. I love doing it. I love the product. I love the market. I love the customers. And so in order to be able to invest, I partnered up with a very good friend of mine, Todd Goldberg. Now, Todd Goldberg and I had been angel investing for a long time. I'd sold my last company, he'd sold his last company, and we kept seeing each other in deals. And we realized that we had mutual taste when it came to picking companies and that perhaps we should pool our resources, go raise money, and deploy larger amounts of capital. So that's precisely what we did. But whereas I spend most of my time working on Superhuman, Todd spends most of his time being an investor, whether it's the angel funds that we run together or other investments that he does. He also, for example, has an NFT fund that he runs with a friend of his who works in crypto. So that arrangement makes it possible for me and able to do in a way that I'm not shortchanging either side, I would encourage thinking about a structure like that.

Thank you. I think this will conclude today's interview because we spent a lot of time running through a lot of circles, a lot of different topics, and getting an insight into Superhuman, into your mind, and how you approach problems, product building, and also investing and more. So Rahul, I'll definitely link to your LinkedIn. I'll link to Superhuman for everybody who hasn't tried it or just wants to get a bit more of a glimpse what we were talking about. I'll link to the angel fund so that people can have a look if that would be a great investor to reach out to. And I'm super thankful for you taking the time. I will also link to a few content pieces that we mentioned. And I guess if I may, at some point, um, I'll be happy to invite you again and ask you more questions. But, um, let's see when and if this happens.

Great. Well, I will be excited to be back. And again, thank you for having me on the show.