The Exchanges

Q You know, you know what I do, but I actually heard that this intro in particular was more interesting than any other. Brian actually kind of left a dangling carrot. He said that you had some very early, more turbulent years to your career, but he didn't actually go into them. Can you help me understand, uh, what were these more turbulent early career years?

A Yes, Harry, yeah, no, that's totally true. I, I like talking about this because I think people think they need to know exactly what they're gonna do from the get-go, and I'm a great example of someone who, like, fumbled about a lot in my early twenties, and it's still okay. Um, I majored in philosophy, which is arguably one of the most useless majors you can have. I say that in a self-deprecating way. Of course, you learn how to think, et cetera, et cetera, but it's not very useful. I didn't know how to use Excel when I graduated. And All I wanted was to live in New York because my dad had lived in New York and I grew up with like New York pictures all over Brazil. And I just wanted that experience, but I applied to a hundred jobs and didn't get any of them. And then finally I got an internship and I was the only intern who had already graduated. So I was like the loser, older intern, um, at the company and things didn't go quite swimmingly. Um, so for starters, my, my boss, uh, was a heroin addict at the time. Um, So she, we became great friends, and I knew there was something off, and one day I confronted her, so she finally admitted it to me, and then we ended up going to Narcotics Anonymous together, and they asked me to take over for her responsibilities while she was out, um, and so that was one way in which I, I, I grew, because I had to do the job of my boss, um, very q…

AI assessment note: “my boss was a heroin addict at the time”

Q Pause, pause, Gina. You said you go and, you know, work and live on farms, And then Tumblr hired you. I don't imagine that's the normal sourcing strategy for Tumblr at the time. How did that happen?

A Yes. So that, you know, I, I really believe that a lot of the big decisions are made because people don't know any better or they don't have other choices. One thing that I did that was really good throughout my whole career is that I love meeting people and some people call that networking, but I really, truly love meeting people. And when I was in New York and I had kind of these crappy jobs at digital marketing agencies and I had to leave, I decided I wanted to leave all the, meet all the people who I had spoke to on the phone or email that I thought sounded interesting. And I reached out to them and said, like, I'm leaving the country. I would love to have the opportunity to have a coffee with you. And one of the people I met was Mark Cotney. And Mark Cotney had just been hired at Tumblr as their head of media. He was at Newsweek before. Amazing man. We're still friends. Um, and he talked to me and said, I think that Tumblr wants to grow in Brazil soon. And I said, well, good luck to Tumblr. I am no longer doing this. I'm not, I'm not doing the career thing anymore, but you know, I hope you guys have a great time. And I went back to Brazil. And they reached out to me six months later saying, Hey, remember, you know, we were, we're now at the point where we want to grow in Brazil and we remember you. Um, and I had worked with, with Mark on a campaign at one of the digital ma…

AI assessment note: “they reached out to me six months later saying, Hey, remember”

Q Okay, so if that's one with notifications that worked, and actually we saw, like, actually incredibly high investment, sorry, incredibly low investment for great returns, if we think about the flip side, was there one that maybe didn't work, or took longer to work?

A Yes, there is, there is one particular experiment that we did, and that really didn't work, and I'm almost cheating because it ended up being our most successful experiment of all time in the growth team. Um, so what happened was the team really wanted to, to implement badges on Duolingo. One thing that we did as a team that ended up being really effective was that we would have different members of the team go and play, like intentionally play different games that were very popular at the time, and then identify what they thought were growth levers on those games, and then talk about how we could implement those on to Duolingo in our growth meeting. We would do that about once a month. And this is exciting and cool because most people building an education app would think like, Ooh, what's working for education apps? Maybe I'll go copy that. But we always thought what's working for retention and games get retention, right? And education apps do not get retention, right? And why? And so we always looked at games for our frameworks and everyone was convinced that badges was going to be huge because all of the top games have badges, like even from the days of four square. So you do something, You gain a badge, and then you do something else, and you get another badge, and now you're collecting badges, and they're exciting, and then you can show off how many badges you have, and t…

AI assessment note: “Yes, there is, there is one particular experiment that we did, and that really didn't work”

Q Can I ask, Tumblr's a very iconic early internet name. When you think about how that time shaped your mindset today, and any big lessons, are there any that you would really think to straight away from that time with Tumblr?

A Well, yes, Harry. One of the things I learned at Tumblr that I think a lot of people in early stage startups that succeed learn is that no one knows what they're doing at most points in time in life and in business. And so look, Tumblr was an amazing product. David Karp is brilliant. The team was awesome, but a lot of the stuff we were doing, we were fumbling about trying to figure it out. Um, and you know, one anecdote I have to tell about that is Just the memory of standing at a cash register in Sao Paulo trying to take all the money I had in my account in Brazil to pay our contractors because Tumblr couldn't figure out how to wire money from the US to Brazil. And then I ended up borrowing money from family members, and I borrowed money from a well-known startup founder today as well in order to pay those contractors because we were doing this big launch event and the money just hadn't shown up. And if the money didn't make it, we wouldn't have the launch event that I'd been working on for so long. And, you know, I wasn't paid at Tumblr for six months, because they couldn't figure that out, and I was like, I was living with my parents, like in my childhood bedroom at the time when I was 25, and like working out of like my childhood bed, literally. So it was okay, I could eat and sleep, but I, I realized, you know, everyone's figuring every step of the way out, and if they can…

AI assessment note: “One of the things I learned at Tumblr... is that no one knows what”

Q Pause, pause, Gina. You said you go and, you know, work and live on farms, And then Tumblr hired you. I don't imagine that's the normal sourcing strategy for Tumblr at the time. How did that happen?

A Yes. So that, you know, I, I really believe that a lot of the big decisions are made because people don't know any better or they don't have other choices. One thing that I did that was really good throughout my whole career is that I love meeting people and some people call that networking, but I really, truly love meeting people. And when I was in New York and I had kind of these crappy jobs at digital marketing agencies and I had to leave, I decided I wanted to leave all the, meet all the people who I had spoke to on the phone or email that I thought sounded interesting. And I reached out to them and said, like, I'm leaving the country. I would love to have the opportunity to have a coffee with you. And one of the people I met was Mark Cotney. And Mark Cotney had just been hired at Tumblr as their head of media. He was at Newsweek before. Amazing man. We're still friends. Um, and he talked to me and said, I think that Tumblr wants to grow in Brazil soon. And I said, well, good luck to Tumblr. I am no longer doing this. I'm not, I'm not doing the career thing anymore, but you know, I hope you guys have a great time. And I went back to Brazil. And they reached out to me six months later saying, Hey, remember, you know, we were, we're now at the point where we want to grow in Brazil and we remember you. Um, and I had worked with, with Mark on a campaign at one of the digital ma…

AI assessment note: “one of the people I met was Mark Cotney. And Mark Cotney had just been hired”

Q Okay, so we have that story. Why did PR care? Everyone has a story today. Everyone thinks their story is unique. How do we take it one step further?

A So the first thing I would say is not just write your story, but you need to condense that story down into one liner. And that's of course, like what you would do for a startup in general, but one line really matters because when you're messaging someone for the first time, they really don't care who you are. They don't care about what you're writing about. So their attention needs to be peaked immediately. So what I advise founders to do is basically come up with every name or number that might be impressive about their startup or their, or their background and condense it into one line. So for example, You know, I'll, I'll just go with mine, right? Like, like Gina Gotthilf has, um, uh, if I, if I introduce myself to someone, I'll say, hi, Harry, I led growth at Duolingo from three to two hundred million users before working for the Mike Bloomberg presidential campaign, dah, dah, dah, dah, and, you know, coauthored a neuroscience study. Like, I don't know if that's what I would do, but like, these are three moments and three signifiers that Harry would say, I don't know who Gina is and what she did, but I, she just said three things That make me think this is important and I should pay attention. So you need to find those signifiers. And if you don't have them yet, it's okay. You can find proxies. Like for example, I went to this university or so-and-so has used my product, or…

AI assessment note: “condense that story down into one liner... come up with every name or number”

Q Can I ask, Tumblr's a very iconic early internet name. When you think about how that time shaped your mindset today, and any big lessons, are there any that you would really think to straight away from that time with Tumblr?

A Well, yes, Harry. One of the things I learned at Tumblr that I think a lot of people in early stage startups that succeed learn is that no one knows what they're doing at most points in time in life and in business. And so look, Tumblr was an amazing product. David Karp is brilliant. The team was awesome, but a lot of the stuff we were doing, we were fumbling about trying to figure it out. Um, and you know, one anecdote I have to tell about that is Just the memory of standing at a cash register in Sao Paulo trying to take all the money I had in my account in Brazil to pay our contractors because Tumblr couldn't figure out how to wire money from the US to Brazil. And then I ended up borrowing money from family members, and I borrowed money from a well-known startup founder today as well in order to pay those contractors because we were doing this big launch event and the money just hadn't shown up. And if the money didn't make it, we wouldn't have the launch event that I'd been working on for so long. And, you know, I wasn't paid at Tumblr for six months, because they couldn't figure that out, and I was like, I was living with my parents, like in my childhood bedroom at the time when I was 25, and like working out of like my childhood bed, literally. So it was okay, I could eat and sleep, but I, I realized, you know, everyone's figuring every step of the way out, and if they can…

AI assessment note: “One of the things I learned at Tumblr... is that no one knows what they're doing”

Q Okay, so if that's one with notifications that worked, and actually we saw, like, actually incredibly high investment, sorry, incredibly low investment for great returns, if we think about the flip side, was there one that maybe didn't work, or took longer to work?

A Yes, there is, there is one particular experiment that we did, and that really didn't work, and I'm almost cheating because it ended up being our most successful experiment of all time in the growth team. Um, so what happened was the team really wanted to, to implement badges on Duolingo. One thing that we did as a team that ended up being really effective was that we would have different members of the team go and play, like intentionally play different games that were very popular at the time, and then identify what they thought were growth levers on those games, and then talk about how we could implement those on to Duolingo in our growth meeting. We would do that about once a month. And this is exciting and cool because most people building an education app would think like, Ooh, what's working for education apps? Maybe I'll go copy that. But we always thought what's working for retention and games get retention, right? And education apps do not get retention, right? And why? And so we always looked at games for our frameworks and everyone was convinced that badges was going to be huge because all of the top games have badges, like even from the days of four square. So you do something, You gain a badge, and then you do something else, and you get another badge, and now you're collecting badges, and they're exciting, and then you can show off how many badges you have, and t…

AI assessment note: “there is one particular experiment that we did, and that really didn't work”

Q What was your North Star metric at Duolingo when you think back to it? Is it the number of sessions per week? Is it the number of levels completed? What was, how did you define success as a user?

A The north star metric for Duolingo at all times was daily active users. That's it. DAU. And so of course we had some metrics that helped us understand whether, you know, like how we were impacting daily active users, but we cared about daily active users from the beginning, even before we were doing growth experiments with my team and we were just looking at PR. So I was told like PR only matters if we're able to increase the number of daily active users. And if you talk to a PR agency, they would say that's impossible. I don't want to do that. But it is possible because if you're able to get this product in front of a lot of people who are your target users, then they will actually use what, you know, what you're, you're, you're showing them. Um, whereas a lot of people focus on metrics closer to acquisition and growth gets correlated with acquisition a lot of the time, but it doesn't matter how many people you get. If it's not, if your retention isn't there, Then it doesn't quite matter. So spending a lot of time on retention upfront, if you can, really helps make sure that every time, every, every second, and every dollar you spend on acquisition goes a lot further.

AI assessment note: “The north star metric for Duolingo at all times was daily active users.”

Q Ok, tell me, what's the biggest mistake founders make when hiring growth teams?

A Definitely hiring before product market fit. People are so excited by A-B testing that they want to jump right in. And I feel guilty about that because I do a talk for early stage founders about how Duolingo A-B tested and how we grew from three to two hundred million users. And then everyone wants to jump into A-B testing. But what you don't realize is that it's really hard to do that if you are not able to reach statistical significance. And then founders start getting into the trap of directionally statistical significance, which is no longer math or science. And so then that doesn't make any sense. So when you're, so I think that's number one, the number two, I would say, especially for early stage is, and it's something I will disagree with some, some other guests you've had on your podcast is hiring a leader too soon, because when you hire a leader, they might be a little bit too removed to actually want to roll up their sleeves and try the crazy stuff that like might not make a lot of sense, but could work because you're trying to get like creative. Um, so, you know, as an example, Duolingo actually hired someone amazing who had a great experience in marketing. Who then hired me and they ended up keeping me and promoting me to head of growth because this person was a great, they, they were great at coming up with ideas and then asking other people to do those ideas. I wa…

AI assessment note: “Definitely hiring before product market fit.”

Q Let's do a final one. Let's do, when you think about companies at least recently, they've been really successful. Which one has stuck with you and really impressed you?

A Oh, well, I guess I, this is a really good opportunity to plug my friend, Mike Krieger, co-founder and CTO of Instagram. I'm really interested by, um, their new app. Mike and Kevin are working on a new app called Artifact, and it's disrupting the way that, um, you know, you consume news, which I'm really interested by, actually, because I worked at Quartz Media, so this is something that really, you know, is, is Quartz to me, but also just their lessons in scaling Instagram will be applied to this, and I'm excited about it because I think That we don't know how to discuss news as a society, and that's a really big problem because we never evolve. If we only talk to people who believe what we believe, it's very hard to change your mind and to understand why there are other points of view in the world, so that's a huge problem that we can solve to make humanity better.

AI assessment note: “Mike and Kevin are working on a new app called Artifact”

Q How much control do you want over the content? Journalists can be challenging, bluntly, in terms of what they produce. I've had journalists write snippets of, like, what I've said and turn it into the main piece, and it's been not what I wanted. How much control do you give? How do you think about ensuring the piece that you get is the piece that you wanted?

A You don't get control. Like if you're working with real journalists, they think, and they write, and that's the best thing about journalism, but it can also be the worst thing about journalism. Um, so there is a couple of layers here. First of all, when you write your release, because you know exactly what you want to say, write it exactly as you would like it to be published with all the hyperlinks there. That really matters. It really matters if you have hyperlinks. I cannot emphasize this enough. Um, Then, you know, you can send it out. I would say that most journalists, not at the top publications, but you would be surprised with most journalists. Tend to be fairly lazy. And I don't fault journalists. They have way too many things to do. They're trying to click, you know, get clicks. They are under immense amount of pressure. And so a lot of them will copy paste. And so that helps a lot. If you're able to just have exactly what you wanted in there in the first place, and then it's like copy paste it. And then you get the story. That's not you having control. That's you gaming the system. Um, but then, you know, you can't have control and tell a journalist, I want you to write this or that, because I actually will rub them in an incredibly the wrong way, much like When you're working with an engineer and you tell them like, here's the solution I want you to build versus here…

AI assessment note: “You don't get control. Like if you're working with real journalists, they think”

Q So I have two questions for you. One is, okay, I, you're on my team, and you're doing this amazing work, and I'm super happy. Gina, I need to measure some form of success from your work. You know, we've got links to TechCrunch, and to Forbes, and to whatever publication. How do we measure success of PR, and actually align it back to the core business?

A It's really hard. And that's, and that's not a cop out answer. And you have to, especially if you're early stage. So for Duolingo, it was DAUs. Like if we were actually able to get new users that became active users, that was our success metric, because then we knew that we were reaching the right audience and that, you know, the story that was told was the story that we needed to tell in order to get people to download and try Duolingo and that, you know, they were likely users of Duolingo. So that's what really mattered. But it depends, right? So if you're trying to raise funding, then it's like, if you're able to raise funding, or if you got people who are referencing the stories, um, on a regular basis, like I heard about you through this, or I heard about you through that, like that's, you could measure that if you had like a landing page or something like, how did you hear about this? Like that could be helpful, but some of that is going to be intangible because you grow your brand equity over time, and that can have a huge impact on your business. But when you're a startup, you don't have the luxury to just be like, well, I'm just building this brand equity and we'll see where it goes.

AI assessment note: “for Duolingo, it was DAUs. Like if we were actually able to get new users”

Q Ok, tell me, what's the biggest mistake founders make when hiring growth teams?

A Definitely hiring before product market fit. People are so excited by A-B testing that they want to jump right in. And I feel guilty about that because I do a talk for early stage founders about how Duolingo A-B tested and how we grew from three to two hundred million users. And then everyone wants to jump into A-B testing. But what you don't realize is that it's really hard to do that if you are not able to reach statistical significance. And then founders start getting into the trap of directionally statistical significance, which is no longer math or science. And so then that doesn't make any sense. So when you're, so I think that's number one, the number two, I would say, especially for early stage is, and it's something I will disagree with some, some other guests you've had on your podcast is hiring a leader too soon, because when you hire a leader, they might be a little bit too removed to actually want to roll up their sleeves and try the crazy stuff that like might not make a lot of sense, but could work because you're trying to get like creative. Um, so, you know, as an example, Duolingo actually hired someone amazing who had a great experience in marketing. Who then hired me and they ended up keeping me and promoting me to head of growth because this person was a great, they, they were great at coming up with ideas and then asking other people to do those ideas. I wa…

AI assessment note: “Definitely hiring before product market fit.”

Q send them the press release? I'm going through this now with, with something. Do you send them the press release? Do you say, hey Gina, would love to hop on a call, and then before the call, or after the call you send it, do you, do you have to do it to get their attention? How do you think about using docs or a press release to get attention?

A I would say there's no, there's no recipe, and people do it differently, and especially, you know, agencies do it differently. The way that work, that has worked out for me in the past, is generating FOMO. And so I don't send all of the information up front, which is annoying to journalists. I'm sorry for the journalists who are listening to this. I'm sure it's annoying, but it piques their interest and human beings react to things that pique their interest. Um, and he wants the person to react. And so you, I would write to these journalists and I would, you know, because on, on LinkedIn, you only get a really small message that you can send and on Twitter, there's like a 140 characters. So I got really good at just sending that short message. But the short message has to be interesting enough that the person would reply with one word. Like, would you like to hear more? Sure. I got a lot of that. You know, now this is my, the way I read into this, like psychology of it. If the person has responded to you once, now they have a relationship with you, even if it's a very marginal relationship, but before they had zero relationship to you. So it's very easy to ignore an email that you've never responded to before, but if you've responded to someone once, you're much more compelled to respond a second time. So now you already have a foot in the door and in a much better place, and t…

AI assessment note: “I don't send all of the information up front, which is annoying to journalists.”

Q you're going for new customers, well, shit, maybe, I don't know, um, uh, Wall Street Journal, or Shipping Weekly, or I don't know whatever random vertical you're going for, but TechCrunch is not going to get you more customers if you're in shipping, or heavy logistics, right? Like, who's the, what's the objective, and who are we going for? And let that trickle down into your publication list, I think.

A No, that's absolutely true. Um, and one thing to take into consideration is that if you're able to get something on TechCrunch, even if it wasn't your, you know, core consumer possibility, then very quickly tell other publications that are more, that feel more insecure than TechCrunch, like, like editors and journalists, and the link to TechCrunch just covered it, they're much more likely to cover it as well, so you might be able to have your cake and eat it too. But one thing that I see a lot is exactly that. It's not understanding the motivation behind what you're doing and behind what the people who are reading that publication. So I see a lot of startups say, oh, we got a funding round and we did this big announcement and we didn't get that many new users from it. Well, yeah, because people who are reading about funding rounds and who are reading about like the monetization of a startup are not necessarily the people whose problems you're solving with that, with that startup. And so that's the people you want to reach, which might be like, You know, uh, Elle magazine. Because, like, you know, women are going to be interested in learning a language, too. And, you know, so you don't think about that.

AI assessment note: “No, that's absolutely true. Um, and one thing to take into consideration”

Q How much control do you want over the content? Journalists can be challenging, bluntly, in terms of what they produce. I've had journalists write snippets of, like, what I've said and turn it into the main piece, and it's been not what I wanted. How much control do you give? How do you think about ensuring the piece that you get is the piece that you wanted?

A You don't get control. Like if you're working with real journalists, they think, and they write, and that's the best thing about journalism, but it can also be the worst thing about journalism. Um, so there is a couple of layers here. First of all, when you write your release, because you know exactly what you want to say, write it exactly as you would like it to be published with all the hyperlinks there. That really matters. It really matters if you have hyperlinks. I cannot emphasize this enough. Um, Then, you know, you can send it out. I would say that most journalists, not at the top publications, but you would be surprised with most journalists. Tend to be fairly lazy. And I don't fault journalists. They have way too many things to do. They're trying to click, you know, get clicks. They are under immense amount of pressure. And so a lot of them will copy paste. And so that helps a lot. If you're able to just have exactly what you wanted in there in the first place, and then it's like copy paste it. And then you get the story. That's not you having control. That's you gaming the system. Um, but then, you know, you can't have control and tell a journalist, I want you to write this or that, because I actually will rub them in an incredibly the wrong way, much like When you're working with an engineer and you tell them like, here's the solution I want you to build versus here…

AI assessment note: “You don't get control. Like if you're working with real journalists”

Q So I have two questions for you. One is, okay, I, you're on my team, and you're doing this amazing work, and I'm super happy. Gina, I need to measure some form of success from your work. You know, we've got links to TechCrunch, and to Forbes, and to whatever publication. How do we measure success of PR, and actually align it back to the core business?

A It's really hard. And that's, and that's not a cop out answer. And you have to, especially if you're early stage. So for Duolingo, it was DAUs. Like if we were actually able to get new users that became active users, that was our success metric, because then we knew that we were reaching the right audience and that, you know, the story that was told was the story that we needed to tell in order to get people to download and try Duolingo and that, you know, they were likely users of Duolingo. So that's what really mattered. But it depends, right? So if you're trying to raise funding, then it's like, if you're able to raise funding, or if you got people who are referencing the stories, um, on a regular basis, like I heard about you through this, or I heard about you through that, like that's, you could measure that if you had like a landing page or something like, how did you hear about this? Like that could be helpful, but some of that is going to be intangible because you grow your brand equity over time, and that can have a huge impact on your business. But when you're a startup, you don't have the luxury to just be like, well, I'm just building this brand equity and we'll see where it goes.

AI assessment note: “So for Duolingo, it was DAUs. Like if we were actually able to get”

Q I love that. Is there any other massive mistakes you make with found, you see founders made with PR that we should touch on before we do a quick fire?

A I think the biggest mistake is not understanding that what matters to you doesn't matter to the other person. Again, this goes back to everything we've been talking about, like in growth, PR, relationships, like that's some, just so core to human nature, you know, you write something that like matters to you. So if I'm a founder and I'll write something that like Is almost, like, so obvious to me because I've been entrenched in it for so long that I understand the problem so deeply, and I'm already, like, talking about the solution, and I think that it's really impressive that I hired Harry Stemmings or whatever, and I put it in there, and people don't know None of this matters to the person who I'm writing to. So you need to understand like how to communicate things in a way that really matters to people. And that is not obvious, even though it should be. And you need to understand who you're speaking to specifically. So if you're writing to someone from like a science background or a tech background, it's very different. If you're writing with someone from like a beauty or fashion background, then you need to understand what's the angle that's going to pique their personal interest. And if you take the time and you're like, actually make a list of the top Applications you care about, and you understand what that journalist might care about, they're much more likely to respond…

AI assessment note: “I think the biggest mistake is not understanding that what matters to you doesn't matter”

Q Okay, so we have that story. Why did PR care? Everyone has a story today. Everyone thinks their story is unique. How do we take it one step further?

A So the first thing I would say is not just write your story, but you need to condense that story down into one liner. And that's of course, like what you would do for a startup in general, but one line really matters because when you're messaging someone for the first time, they really don't care who you are. They don't care about what you're writing about. So their attention needs to be peaked immediately. So what I advise founders to do is basically come up with every name or number that might be impressive about their startup or their, or their background and condense it into one line. So for example, You know, I'll, I'll just go with mine, right? Like, like Gina Gotthilf has, um, uh, if I, if I introduce myself to someone, I'll say, hi, Harry, I led growth at Duolingo from three to two hundred million users before working for the Mike Bloomberg presidential campaign, dah, dah, dah, dah, and, you know, coauthored a neuroscience study. Like, I don't know if that's what I would do, but like, these are three moments and three signifiers that Harry would say, I don't know who Gina is and what she did, but I, she just said three things That make me think this is important and I should pay attention. So you need to find those signifiers. And if you don't have them yet, it's okay. You can find proxies. Like for example, I went to this university or so-and-so has used my product, or…

AI assessment note: “condense that story down into one liner”

Q What was your North Star metric at Duolingo when you think back to it? Is it the number of sessions per week? Is it the number of levels completed? What was, how did you define success as a user?

A The north star metric for Duolingo at all times was daily active users. That's it. DAU. And so of course we had some metrics that helped us understand whether, you know, like how we were impacting daily active users, but we cared about daily active users from the beginning, even before we were doing growth experiments with my team and we were just looking at PR. So I was told like PR only matters if we're able to increase the number of daily active users. And if you talk to a PR agency, they would say that's impossible. I don't want to do that. But it is possible because if you're able to get this product in front of a lot of people who are your target users, then they will actually use what, you know, what you're, you're, you're showing them. Um, whereas a lot of people focus on metrics closer to acquisition and growth gets correlated with acquisition a lot of the time, but it doesn't matter how many people you get. If it's not, if your retention isn't there, Then it doesn't quite matter. So spending a lot of time on retention upfront, if you can, really helps make sure that every time, every, every second, and every dollar you spend on acquisition goes a lot further.

AI assessment note: “The north star metric for Duolingo at all times was daily active users.”

Q you're going for new customers, well, shit, maybe, I don't know, um, uh, Wall Street Journal, or Shipping Weekly, or I don't know whatever random vertical you're going for, but TechCrunch is not going to get you more customers if you're in shipping, or heavy logistics, right? Like, who's the, what's the objective, and who are we going for? And let that trickle down into your publication list, I think.

A No, that's absolutely true. Um, and one thing to take into consideration is that if you're able to get something on TechCrunch, even if it wasn't your, you know, core consumer possibility, then very quickly tell other publications that are more, that feel more insecure than TechCrunch, like, like editors and journalists, and the link to TechCrunch just covered it, they're much more likely to cover it as well, so you might be able to have your cake and eat it too. But one thing that I see a lot is exactly that. It's not understanding the motivation behind what you're doing and behind what the people who are reading that publication. So I see a lot of startups say, oh, we got a funding round and we did this big announcement and we didn't get that many new users from it. Well, yeah, because people who are reading about funding rounds and who are reading about like the monetization of a startup are not necessarily the people whose problems you're solving with that, with that startup. And so that's the people you want to reach, which might be like, You know, uh, Elle magazine. Because, like, you know, women are going to be interested in learning a language, too. And, you know, so you don't think about that.

AI assessment note: “It's not understanding the motivation behind what you're doing”

Q we were chatting, that it's actually kind of heavily underutilized and underrated, I think, by founders. Actually, I think it's often denigrated, if we're being totally honest. Fluffy PR marketing, whatever stuff. Um, and it's handed to like the intern, I find, and it's really just not the right way to do it. So my question to you is, how should founders think about PR at the most tactical level?

A PR is so not sexy. It's considered really old school. It's like, ugh, no, I want to do product-led growth. I'm going to be doing A-B tests. I ended up doing PR before I knew what PR was. I just thought, you know what, if I get these big newspapers to write about Tumblr, then I think that like, we're going to get a lot of users. And then I was like, how do I get these big papers to write about Tumblr? Maybe I should just find who the journalists are and then I should figure out like what interests them. And that I guess was PR. Um, and it honestly was the most important lever that helped Tumblr grow in Latin America and that helped Duolingo grow. In the first two years, I would say, which is crazy. And of course, we were talking about retention. I only had the luxury of doing PR because those were retentive products. Those were products that already had a baseline of retention such that, you know, we knew that it had product market fit, and people actually wanted to use it. So from a very tactical perspective, here's how I think about PR. Looking back, and I have a playbook, but at the time, just to really highlight my ignorance, I didn't have a playbook. I was just really trying things out, and they worked. So here's what you need to do. First, you need to have a story to tell. And if you don't have a story to tell, it doesn't mean that it's not there. It just may, it might mea…

AI assessment note: “So from a very tactical perspective, here's how I think about PR.”

Q This is so relevant, because I'm literally doing this as we speak with some companies, but it's like, how early before a release do you go? Is it two weeks, a week, two days? How long do journalists need to not feel like it's too long, but not too short?

A You want to give the journalist enough time to work on this story because most journalists have a lot of work and they're overwhelmed and they also have their own schedules. And so if you, if you, if you're too close to the date, you really risk the journalist just not, not covering it, even if they were interested, which is a really big miss. So I've had to move artificial dates before because the journalist was like, great, but I don't have time for this. And I go, okay, well, what if, what if it was like a week later, you know, like, and like trying to make that work. Um, of course you don't want to do it too early that it's like irrelevant. So I think that starting to email, so here's the thing. I would say sending the release or sending the information with journalists two weeks before probably works, but you need to think, and now Harry's writing on his arm, so I think I said something relevant.

AI assessment note: “sending the release or sending the information with journalists two weeks before probably works”

Q I love that. Is there any other massive mistakes you make with found, you see founders made with PR that we should touch on before we do a quick fire?

A I think the biggest mistake is not understanding that what matters to you doesn't matter to the other person. Again, this goes back to everything we've been talking about, like in growth, PR, relationships, like that's some, just so core to human nature, you know, you write something that like matters to you. So if I'm a founder and I'll write something that like Is almost, like, so obvious to me because I've been entrenched in it for so long that I understand the problem so deeply, and I'm already, like, talking about the solution, and I think that it's really impressive that I hired Harry Stemmings or whatever, and I put it in there, and people don't know None of this matters to the person who I'm writing to. So you need to understand like how to communicate things in a way that really matters to people. And that is not obvious, even though it should be. And you need to understand who you're speaking to specifically. So if you're writing to someone from like a science background or a tech background, it's very different. If you're writing with someone from like a beauty or fashion background, then you need to understand what's the angle that's going to pique their personal interest. And if you take the time and you're like, actually make a list of the top Applications you care about, and you understand what that journalist might care about, they're much more likely to respond…

AI assessment note: “I think the biggest mistake is not understanding that what matters to you doesn't”

Q about retention there being the focus. I had Alex Schultz, the CM of Meta, on the show, and he said basically, retention's king. Everything else doesn't really matter. How do you think about effective North Star metric setting? Because I think most people actually just set the wrong North Star and then optimize for the wrong thing. How do you think about that, and how did retention play into that?

A I agree. Retention is everything. To me, retention is almost like product market fit. If you have retention, it means that people actually like the thing that you made, and it's actually useful for them. And that determines whether something's going to succeed or is not going to succeed. So that's like a very basic way of looking at it. You can, you can, you know, create like, you know what, Harry, I just made this box. I think it's an incredible box. And if I invest a lot of money in marketing this box, and if I'm a marketing genius, It's going to sell a lot, and then the minute I stop investing, it will not sell a lot because it's actually not that great, and it's just like every other box out there. But if you make something that's amazing, you know, then every dollar you pour, every minute you pour into making people aware that that exists and that that's a possibility to solve their problem, then that money and that time goes far, goes a lot further because it actually is useful to people.

AI assessment note: “Retention is everything. To me, retention is almost like product market fit.”

Q We're gonna do three more, okay? Now, first one of the three, why is now the best time to invest in LATAM?

A Now is the best time to invest in LATAM for a number, because of a number of reasons that are coalescing at this moment. First, there's just the sheer opportunity because development in tech in LATAM is quote unquote behind that of markets like Silicon Valley. There's just less huge solutions that come from LATAM. And the reason why that matters is because Latin America is huge. It's like, it's actually occupies the largest latitude on the map, which is why my country, my company is called Latitude. But we don't behave like it. And we're actually very alike. We like to focus on how we're all very different, but we're actually very alike. We like very similar music. We like similar food and our governments are just as corrupt and you're just as likely to get robbed or kidnapped in most of our countries. Um, and so there's just, it's important to focus on the similarities. That's something I learned while, while growing globally at Duolingo. So you have a very, very large number of people with huge problems that have not been solved and that can be solved at scale. Through tech. And then you have people in developed markets who absolutely don't know what it's like to, to grow up in those markets and don't understand those pain points and don't understand what are the, the barriers that they're going to face while we're there, because they're just don't know how complicated it can…

AI assessment note: “Now is the best time to invest in LATAM for a number, because of a number of reasons”

Q This is so relevant, because I'm literally doing this as we speak with some companies, but it's like, how early before a release do you go? Is it two weeks, a week, two days? How long do journalists need to not feel like it's too long, but not too short?

A You want to give the journalist enough time to work on this story because most journalists have a lot of work and they're overwhelmed and they also have their own schedules. And so if you, if you, if you're too close to the date, you really risk the journalist just not, not covering it, even if they were interested, which is a really big miss. So I've had to move artificial dates before because the journalist was like, great, but I don't have time for this. And I go, okay, well, what if, what if it was like a week later, you know, like, and like trying to make that work. Um, of course you don't want to do it too early that it's like irrelevant. So I think that starting to email, so here's the thing. I would say sending the release or sending the information with journalists two weeks before probably works, but you need to think, and now Harry's writing on his arm, so I think I said something relevant.

AI assessment note: “sending the information with journalists two weeks before probably works”

Q Can I be really boring, just really tactical? Do we write that story down on a Google Doc? Do we, how do we codify it? Who, who often is involved? I just want to get as tactical as possible, because It will save me doing this work, and then I can just send them this episode instead.

A Yes, of course, Harry. So what I would recommend is you just tell this, like tell the story of what you, what you think is meaningful about your app or what you're building over and over and over and over. And you watch very carefully for how people react to things. And then you learn what actually matters to people and what doesn't matter to people. And so if you have a start, a startup or a product, you know, and I go, Harry, let me tell you about my, the thing I'm building. And I like tell you, and at first you're going to pretend that you're interested because you're like really nice and I'm your friend or whatever. And I'll say a lot of things during the course of like, well, I'm telling you about my product and you're going to ask a question about something in particular. Are you going to react to something in particular? That's when you know you have a hook about something that matters to people. And you can actually write that down if that's going to make it easier to you. And I would actually write all of those moments down and understand how that sort of backtracks into a story.

AI assessment note: “You can actually write that down if that's going to make it easier to you.”

Q Can I be really boring, just really tactical? Do we write that story down on a Google Doc? Do we, how do we codify it? Who, who often is involved? I just want to get as tactical as possible, because It will save me doing this work, and then I can just send them this episode instead.

A Yes, of course, Harry. So what I would recommend is you just tell this, like tell the story of what you, what you think is meaningful about your app or what you're building over and over and over and over. And you watch very carefully for how people react to things. And then you learn what actually matters to people and what doesn't matter to people. And so if you have a start, a startup or a product, you know, and I go, Harry, let me tell you about my, the thing I'm building. And I like tell you, and at first you're going to pretend that you're interested because you're like really nice and I'm your friend or whatever. And I'll say a lot of things during the course of like, well, I'm telling you about my product and you're going to ask a question about something in particular. Are you going to react to something in particular? That's when you know you have a hook about something that matters to people. And you can actually write that down if that's going to make it easier to you. And I would actually write all of those moments down and understand how that sort of backtracks into a story.

AI assessment note: “I would actually write all of those moments down and understand how that sort of backtracks”