← Building One with Tomer Cohen
Building One with Tomer Cohen
Building Hinge with Justin McLeod: Intentional Design for Meaningful Dating
00:50:11
Loading summary
Transcript99 lines
- [00:00]
Leah Smart
LinkedIn news innovation ultimately is solving problems. It came back to my own personal pain point.
- [00:11]
Jessi Hempel
It just takes a while to build that trust.
- [00:12]
Leah Smart
The problem that I was trying to solve, it was all I was thinking about.
- [00:16]
Jessi Hempel
You have to be obsessed with the human condition.
- [00:19]
Tomer Cohen
I'm Tomer Coyne, Chief product officer of.
- [00:21]
Jessi Hempel
LinkedIn and this is building one.
- [00:25]
Leah Smart
We lose 20% of the people plus during the onboarding process. You know if you were talking to a VC, you'd be like, Whoa, that's crazy.
- [00:32]
Tomer Cohen
That's Justin McLeod, the founder of Hinge. He's talking to me about how Hinge designs its onboarding for new users to make sure they're right for the app. We're going to get into that and so much more. So stick around.
- [00:49]
Leah Smart
From LinkedIn News I'm Leah Smart, host of Everyday Better, an award winning podcast dedicated to personal development. Join me every week for captivating stories and research to find more fulfillment in your work and personal life. Listen to Everyday better on the LinkedIn podcast network, Apple Podcasts or wherever you get your podcasts from. LinkedIn News. I'm Jessi Hempel, host of the hello Monday Podcast. Start your week with the hello Monday Podcast. We'll navigate career pivots. We'll learn where happiness fits in. Listen to hello Monday with me, Jesse Hempel on the LinkedIn podcast network or wherever you get your podcasts.
- [01:32]
Tomer Cohen
Starting a new company is hard. Really hard. But what might be even harder is restarting one. Imagine putting years of work into building something, seeing it grow, and then realizing it's not exactly right. That's the moment where most founders freeze. Do you push forward? Do you pivot? Or do you take the truly radical step of tearing it all down and starting all over again? My guest today faced the exact same moment with his own company. Today I have the pleasure of Speaking with Justin McLeod, the founder of Hinge, currently the fastest growing dating app in the US hinge initially launched in 2012, but hinge as we know it today didn't really start until 2015 when Justin decided to reboot the company and build it around a few core product principles as the dating app designed to be deleted, Hinge is built as a product for intentional daters and one that has the ultimate goal of getting them into relationships and off the app. Justin's focus and discipline in product building really stands out. In my conversation with him, we learned about things like how placing limits on user actions can create better interactions overall. Why Justin is proud to lose more than 20% of new users during the onboarding Process how Hinge combats feature and scope creep, and the role Justin sees for AI in improving our dating lives. That and a lot more. So let's just jump in.
- [03:05]
Jessi Hempel
Justin, it's a pleasure to have you on the show. Thank you so much for joining me today.
- [03:10]
Leah Smart
Yeah, thanks for having me.
- [03:11]
Jessi Hempel
So I thought we'll start kind of early with your journey. And like many of our guests, you built Hinge out of a deeply felt need. You wanted to find a girlfriend. As far as I understand now, most people, 99.9% of them would just try whatever is out there, but you decided to build something. So I'm curious why? What drew you to go and build a solution versus just use what's out there for you?
- [03:36]
Leah Smart
Yeah, it's a long story. And then there's the reboot, 2015, 2016, refounding of hinge. So we're talking about 2011 originally. Like, why did I get into this in the first place?
- [03:48]
Jessi Hempel
And this was the Facebook app. Hinge.
- [03:51]
Leah Smart
Yeah, Facebook app. And then. And then it was a mobile app, but it was a friends of friends kind of swiping experience. But this is 2011. So this is pre dating apps, right? There's dating websites like Match and eharmony and things like that. The day I graduated from college, I'd gotten sober. I worked for a few years, and I went to business school. And business school was like this very, like, social environment. But I was struggling, really to meet people, because I just didn't. It was like a totally foreign world to me, being able to go out to parties and, like, not drink, and everyone else is drinking. And I'd actually reached out to my college girlfriend, Kate, to try to win her back, and she told me in so many words to shove it. I was heartbroken and looked at the services that were out there and just felt like, gosh, this isn't really something that I would use. It's not something that my friends use. We all use at the time, Facebook. I think that there's an opportunity to create something that would be accessible and fun and kind of easy and approachable for people my age, people in their 20s, to use. So that really was the original thought and insight. I wanted it for me, but I also was like, gosh, almost none of us use Matri Harmony. If we do, we're not talking about it. And why is that? And started to really get underneath. What was the reason that people didn't use these services? Well, they were expensive. People used kind of screen names, which was very passe at the Time, because now we had Facebook. People were used to using their real names. You had to fill out very long, detailed, revealing, vulnerable profiles that were like essay questions. It just wasn't the way that people kind of interacted in my age. And so it felt like, what if you built a new service that really appealed to younger people that was, I would say, like, lower lift, more fun, more easy, got you in the door and just made it approachable. If it wasn't so expensive and it wasn't so hard, then I think people would feel more comfortable using it, saying they used it. And I also just always kind of been a romantic at heart. And I've always loved systems and math and algorithms. Always been kind of what I gravitated to is always my strong point in school. And I don't know, it all just kind of converged as a hit of. I just know I'm meant to do this. I had been working on in business school a number of different projects that I was going to enter into a business plan competition. But all of it kind of felt like homework. It all felt like I have to work on my thing now and then. As soon as this idea hit me, it was like. It was like, this thing is coming through me. It's like all I could think about. It's all I wanted to work on. I just felt like I just had to make it a reality.
- [06:35]
Jessi Hempel
Yeah, we hear this quite often from founders. That deeply felt need resonates with them on a level where they cannot not build it.
- [06:43]
Leah Smart
Yeah, that's how I felt. I felt like I couldn't not build it. I just. It was coming out of me one way or another.
- [06:50]
Jessi Hempel
You talked about the second incarnation of hedge in 2015, what you called a reboot. So it sounds like in many ways a starting point was Web 1.0 was kind of dominant. Then came really the social networking web with a lot more thinking around friends, authentic identities, also mobile. So then you're after the races. Externally, it looks like Hinge is doing really well. You decided to reboot. What triggered that?
- [07:21]
Leah Smart
It was a couple things. The real moment was there had just been an article written that was very popular in Vanity Fair called the dawn of the Dating Apocalypse. And it was about how all of these apps had just become a place for hookups and they were superficial and they were ruining romance and love and dating. And Hinge was heavily featured in that article. Also, a couple other dating apps had come out since then that were becoming even more popular than Hinge, specifically Tinder. They beat us at the Game of making it easy and approachable, swipe left and right. Like, really the market was ready for just like a let's make this really simple and really approachable. And it caught fire. We were trying to keep up with the competition and, you know, looking at what features they were releasing and we were releasing them and then we were just becoming, you know, if you're going to copy someone, don't copy the people that are already bigger than you. I don't think it's a winning strategy. We were just less and less differentiated, and so I had less and less of a place to be in the market. But more importantly than that, it just didn't resonate with me where it was going. Like, I really did, as we talked about, start this company, because I, like, wanted to find my person. I wasn't trying to create like a casual encounters type experience. And I actually went and even, like, met with the Nancy Jo Sales, the person who wrote that article, and sat down with her and, and wanted to get, like, what she'd heard. And I remember sitting down with my chief brand officer, Katie Hunt, and we were about to go home for Thanksgiving. It was 2015, and I'm like, gosh, this isn't the company that I want to build. I kind of just want to, like, tear the whole thing down and start over again. And she's like, well, what's stopping you? Like, you're the CEO. If that's what you want to do, then let's do it. And I went home over Thanksgiving break and I thought about it and I'm like, yeah, that's what we're going to do. And I came back and I got my executive team together and we decided to let go of half the company, reboot, throw out the old code base and start over from scratch and build something because a new opportunity had emerged. Like, originally the opportunity was make this thing simple, easy, approachable, and that was obviously like a winning strategy in many ways. And at the same time, for people who were really looking for their person, it was less effective, we'll say. And at that, because you didn't know a lot about these people, didn't know a lot about you, it's hard to make really smart, intelligent matches. And now that people are socialized to the idea of using dating apps, can we create something that really is about less winning because it's so simple and easy, but more because it really strikes that right balance of remaining approachable to the next generation, but also really asks more of them, put in more effort and get more out of it and really help you zero in on your person and be the dating app that's designed to be deleted.
- [10:14]
Jessi Hempel
So I love that. And it sounds like part of it was like the market pushing you. Part of it was like you internally felt like you wanted to build something different. I'm curious when you think right now on the target audience that, you know, you're building for your ideal kind of. You know when you look at two people and you're saying, hey, you should be unhinged, you should not. Like, Hinge is not for you. You're looking for casual connections. There's so many apps for that. You should go there. How do you think about bringing that to the product? So obviously there's branding. I love the design to be deleted thing. It goes really well. When you think about onboarding, do you think about how do you attract the right people on the end? And actually also, how do you tell the folks who should not be on the platform that there's other apps for them?
- [10:57]
Leah Smart
We have an extensive onboarding process and we lose 20% of the people plus during the onboarding process. You know, if you were talking to a vc, you'd be like, well, that's crazy. Like, you shouldn't lose 20% of people during onboarding. But for us, in a sense, it's a feature because it's a filtering mechanism for people who, like, really want to put in the right amount of effort and are intentional. And our target audience, we use the word intentional daters, people who want to put in thoughtful effort because they're looking for a real result. They're here because they want to get off dating apps. That's who we think about as we're building things. So we ask our daters to do more, to fill out three prompts, to add six photos, to give us lots of graphic information, to take thoughtful actions once you're in the product. It's not just the onboarding. Instead of using the swipe feature, you actually have to. If you like someone, like something about them, and you can add a comment, which makes people way more thoughtful and selective in their liking. You don't have to like someone to see if they liked you back. So there's no kind of gamifying the experience. If someone likes you, we just deliver it right to you. We tell you, this person liked you. Would you like to match with them? It does ask people to put themselves out a little bit more and put in a little bit more effort, but it allows us to learn their tastes much, much better. It creates a Pool of people who are all on the same page and have similar intentions and goals. And that creates for those people a really incredible experience, which is why I think we've grown so fast through word of mouth, why we're the fastest growing major dating app now.
- [12:29]
Jessi Hempel
And was there a tension there, Justin, in terms of the holy grail of consumers? Just let them in and walk them kind of progressively through the product, but don't add so much friction at the beginning. Was there tension for you there from the team or because you actually had the reboot? It was like, I actually know what I want to build right now, so let's move away from the quick Tinder like, interactions that we had in the past.
- [12:53]
Leah Smart
Yeah, this is a huge tension, right? We were trying to find the right balance between simple, fun and easy and super. Like, we didn't want to go as far as, like, asking people to fill out long essays and things like that. There's like, almost like an efficiency frontier of vulnerability. Right? It's like, what are people willing to do? But that also puts them out there enough and asks enough of them that we can really help them get the result that they want. You know, a good analogy would be food. You can either make fast food that's horrible for you but tastes great, but burns people out, or you could have food that's so nutritious it's kind of gross. Like, how do you make it sweet green, right? Like, how do you. How do you make it kind of fun and add enough little dressing and the right combination of flavors so it's nutritious, but you want to eat it. We're always trying to figure out what is the right balance of getting people to do what is good for them in the long run. Like some of the earlier explorations, we swung it too far to the other side. We were like, well, everyone's going to have to pay $7 a month to be on our platform, and if you like someone, you have to send a comment with your like. And we tested these things and it was too far. I was asking too much of people and people didn't want to engage and the thing collapsed. So it's definitely this, like, constant honing of trying to push people and push out that frontier of, like, how do you make people a little bit more authentic and put in a little bit more without turning them off and discouraging them?
- [14:18]
Jessi Hempel
I love that idea of efficiency frontier for vulnerability, what would be the X's? So one would be vulnerability and one.
- [14:23]
Leah Smart
Would be effort, or one is like, what people Are willing to do, and then what's effective? Right? Like, there's kind of a natural, linear relationship.
- [14:31]
Jessi Hempel
This.
- [14:31]
Leah Smart
This whole concept started when we started optimizing prompts. So prompts are those short little questions that we ask that people will.
- [14:40]
Jessi Hempel
Those are voice prompts.
- [14:41]
Leah Smart
Right?
- [14:41]
Jessi Hempel
You want to hear the voice as well.
- [14:43]
Leah Smart
Yeah, yeah. Voice is another actual manifestation of that video too far, like, reveals a lot about you, but most people aren't willing to do it. Simple text, too easy. People are definitely willing to do it, but it doesn't give you a lot of the. And voices. There's this amazing sweet spot with voice, and I think that voice is going to become a bigger and bigger part of dating. You get enough to, like, understand someone's personality, but people are willing to do it. But specifically, what I was talking about with prompts is when we started doing this, we would measure how effective is this at getting a conversation going and are people willing to answer this prompt? So there are prompts like my go to karaoke song, which is no longer a prompt because everyone was willing to answer my go to karaoke song, and no one cared what your go to karaoke song was. Right. Very few good conversations start because of, like, your named go to karaoke song, you know, and at the other end of the spectrum, very deep, vulnerable questions. If people were willing to answer these questions, they would lead to great conversations, but no one was willing to answer them. It was like pushing people too far to put that on their dating profile. So it's just like, again, that sweet spot of what I'm looking for in a plus one wedding date, which is something that people are willing to answer but reveals enough to get a good conversation going and signals to you what kind of values that you share. So there's that real art to it.
- [16:01]
Jessi Hempel
Love that when you do a reboot like that, when you think about true north, right? You have your vision, your mission. The app to be deleted talks volume about what you're trying to do. You're trying to help people actually find great relationships, not just go on dates. When you try to put a metric on that, that feels, like, really hard. I'm curious how you think about capturing that quantitatively so you can tell the team we're in the right direction versus a notion of, like, before. When you're concerned that, like, hey, the wrong metric can actually put us on the wrong path.
- [16:36]
Leah Smart
Our North Star metric is. I mean, ultimately, ultimately, it's getting people off the app and into relationships.
- [16:42]
Justin McLeod
But.
- [16:42]
Leah Smart
But it's hard to measure a feature based on that. Right. Because there's a long lead time between releasing a new prompt and saying, like, gosh, did this get more people off our app and into relationships? But what we do use is getting people out on dates. Our people are exchanging contact information and moving out on a date. And that really is our North Star. And so a product feature is good if it moves that up, and it's bad if it moves that down. So that really helps clarify things for us.
- [17:12]
Jessi Hempel
And you know that through the app or this is something you do after the fact.
- [17:16]
Leah Smart
We do have something called we met in the app. So if people do exchange contact information, the app we'll follow up to ask if they went on a date and if this is the type of person they want to see again. But feature by feature, if we're really determining, like, is this a winner or not? We're looking at, do people exchange phone numbers or contact information? And we want to move that number up. And it's hard to move. Right. I'm sure you all have your own kind of user funnels, right? You push in one place and it goes down there, but then it pops it up in another place. Right. So you can give people more matches, but then the match to conversation rate goes down and then it nets out. And when you're designing product teams and everything else, you're not having one team that's pushing on one part of the funnel, but only to watch it hurt another team's metric at another part of the funnel. How do you think, like, end to end and cohesively about that?
- [18:04]
Jessi Hempel
That's hard for us as well. So, you know, our true north is economic opportunity for people. And sometimes we see it in the app because people actually find a job. So you can see the coming in, or you can see them building a business, but that's also becoming a proxy. And when I think about hinge, obviously, relationships, a proxy for that is dating has to be there, and then the proxy for dating is exchanging information. But anyway, soon people are trying to find all the ways to, you know, put guardrails around the metric. And then what I found in the past, that, you know, sometimes it becomes very complex, and then you run the risk of the metric is no longer as simple as it was before. And then, you know, it's unclear if you're potentially, like, making a dent or not. You do something which is really interesting. You actually follow up with members after. You might not be sampling everybody, because that's very hard. But I'm assuming you have, you know, a strong qualitative measure of like, are we pacing in the right direction towards our vision?
- [19:00]
Leah Smart
Yeah, I would say that the we Met data and also the exit survey data. So if someone leaves Hinge, we say, you know, did you find someone? If you found them where this unhinged, you're leaving because you're just done with dating apps, you're done with Hinge. So we can use those to kind of make sure we're steering in the right direction generally. And when we make improvements in dates per user, it's actually good dates and good dates that are leading to relationships, which we generally do find. I think your metric of economic opportunity is probably quite hard to measure and comes in many different forms. For us, it actually is relatively simple. Like, we've got to get people out on good dates. Like, are these people chatting with people and exchanging information and planning dates, or are they not? And that is clarifying when it's that simple.
- [19:44]
Jessi Hempel
Curious. Do you also look at people who stay in the app for a long time and look at that as a sign of, okay, we're. There's a cohort or a need we're not delivering yet?
- [19:54]
Leah Smart
Yeah. So something that was kind of interesting that happened over the last like three years or four years is that Gen Z became our fastest growing segment. Like people 18 to 24, like started just growing really, really fast, which was surprising to us, honestly, because we think we're just like this super relationship intention. Like, find your person, get off the app. And most 21 year olds that you talk to probably aren't like trying to find their forever person. And yet what we found is that just the authenticity of the experience and the realness and the intentionality was more universal. Some people are in an earlier stage of their dating journey, but they still want to go on good dates with quality people. They don't want to just show up on a date and feel like this is a random stranger. And that's not bad for us in a sense. Right. We actually use these different ideas, like explorers and journeyers and destination people. Like, are you here because you're like, let's just do this. I'm ready to find my person. Get off. Is it like, I'm pretty ready to find my person, but I'm still kind of figuring out what I want? Or is it like, I'm just here to see what's out there? And I think as long as people are approaching it with authenticity and vulnerability and willing to put in effort, then hinge is definitely a place for them, which is why we've Added things like relationship intentions on Hinge, which we avoided for a long time because we were like, well, everyone here should be looking for, like, their person. So why would we add relationship intentions that'll dilute our brand. But we found is like, some people just want to have our same ethos when it comes to their dating lives. But even though they maybe are looking for a polyamorous relationship or they're just earlier in their dating journey. And that's okay, too, as long as people are able to align expectations and make sure that they're finding people who have the same expectations they do in many ways.
- [21:44]
Jessi Hempel
To your point, you maintain the core you care so much about right now.
- [21:48]
Leah Smart
Yeah. Our North Star is still getting people out on dates. Like, that doesn't change. Like, Even if you're 18, like, we still want to get you off our app and out on a good date with somebody.
- [21:56]
Tomer Cohen
We're going to take a quick break, but don't go anywhere. When we come back, Justin is going to explain why even really good features may not make the cut.
- [22:04]
Leah Smart
You have to consider how much complexity is this adding, how much maintenance, upkeep, interaction with other features is this going to have. It can really slow you down over time.
- [22:17]
Tomer Cohen
Race the rudders, Race the sails. Raise the sails.
- [22:21]
Jessi Hempel
Captain, an unidentified ship is approaching.
- [22:23]
Leah Smart
Over.
- [22:24]
Tomer Cohen
Roger, wait. Is that an enterprise sales solution?
- [22:28]
Leah Smart
Reach sales professionals, not professional sailors. With LinkedIn ads, you can target the right people by industry, job title, and more. We'll even give you a $100 credit on your next campaign. Get started today@LinkedIn.com marketer terms and conditions apply.
- [22:51]
Tomer Cohen
We're back, and I'm Speaking with Justin McLeod, the founder of Hinge, the dating app.
- [22:55]
Jessi Hempel
So one of the things I loved studying more about Hinge and how you think about building was kind of the strong principles you bring into place. And in fact, people can go on the Hinge website. It's all there. You called them out pretty clearly. And I love that a lot of the things we talk about in this show are principles people work by. I think it tells a lot about how you build and what kind of distinguishes your decision making from others. One of them has been love the problem. I assume that if you love the problem, the expectation from Justin would be that you understand it deeply. You can create a better solution for it because you understand it. However, from my understanding, this principle emerged from an understanding that the old Hinge took on too many goals, and those goals didn't necessarily advance Hinge's mission. Can you share more about the connection between that principle to that kind of scenario that you had in the old Hinge.
- [23:46]
Leah Smart
Yeah, there's sort of two evolutions of this principle. The first principle was like, in the early, early days of hinge, like hinge v1, we really got lost in terms of engagement, retention, monetization, all these different social media type metrics or business metrics. And we were chasing many goals and not getting really clear on what are we ultimately here to serve. And we lost sight of actually getting people out on good dates. We were too focused on did they stay until week two? Are they staying from month one to month two? But what we realized, like, if you're really solving the deep underlying user problem, then the scoreboard sort of takes care of itself. So that was one layer, I would call it, of love the problem. And that's kind of like just get clear on the problem that you really are ultimately here to solve and what's your like, number one priority problem. The second is getting a deeper understanding of a problem and also making sure you're always rooting in a problem. So what I often find is that someone will have an idea of a problem in their head and then they'll come up with a feature and then we get more wedded to the feature than the problem and there's kind of like this drift that happens and then pretty soon we forget why we're even building this feature in the first place. So people have different ideas about what it's supposed to be doing in their minds. This still happens.
- [25:02]
Jessi Hempel
It happens everywhere.
- [25:04]
Leah Smart
Yeah, it happens everywhere, but it's always like calling back. Sometimes it's like, well, actually, is it worth continuing to pursue this feature? Because originally it was supposed to be for this. It's like we re justify it. It's not doing what it was originally supposed to do, but it does this other thing. And. And we're like, well, is that what we're trying to solve here or not? Do we want to keep going with this or do we want to kill it? It's having that discipline to constantly be returning to like, what is the real big thing we're trying to solve here and what's the best way to solve that? And not getting too wedded to a particular feature idea. We say love the problem because it is spending that extra time to really get acquainted with the users and like, what their experience is and what the problem always leads to. A better solution versus just thinking, like, we need to get people on more dates, like, I have an idea. X, Y and Z. Like, it's a Very superficial understanding of the problem. But the more nuance and depth and humanity you put around a problem, the more creative and interesting and effective your solutions are going to be.
- [25:59]
Jessi Hempel
So I love a lot about what you said around, you know, obviously from pushing people to understand what they're solving for in a deeper way so they can clarify the nuances I'm hearing for you. Also, in a way, when something, even if it's successful, I want you to bring back the problem you were supposed to solve and see if you solved the problem versus hey, Justin, my metric is up and to the right. I'm doing really well. But your hypothesis was different about what we were trying to solve. I don't know of many that actually bring back the original hypothesis to connect it together. And I think that's actually a sign of a great builder.
- [26:31]
Leah Smart
Oh, we do. It's tough because sometimes people work really hard on building a feature and it did some other things, but it doesn't really do the thing that we need it to do. And it adds complexity. And this gets to our second principle of keep it simple. And they're very related. Complexity always has a cost. An extra feature always has a cost. And people are like, well, it's neutral. People get wedded to like, well, I built this feature and it didn't hurt the metrics and it feels good. So let's just put it in. And I'm like, well, we put it in and there's the maintenance cost, there's the bug cost, there's the interactions that it's going to have with future product features and it's not worth it. Like, like, in order to keep this product simple, which is a big piece of what's always rooted us, it's down to our black and white branding and everything else. Like, get out of the way, keep things simple, do few things and do them really, really well. It does require a level of discipline and hard conversations and stripping out features and killing features. Maybe they seem nice or people think they, they're cool, but if they're not moving our North Star metric, then they probably don't belong.
- [27:34]
Jessi Hempel
I think for myself, like, you know, the, the easiest kind of features to make a decision on are the ones that are not performing. It's an easy decision. Move out. The ones that are highly performing is a decision like keep investing. Then you have that like massive amount in the middle, which is features that somewhat perform. Yeah, they move the metric. Like you're seeing some qualitative good feedback about them, but it's not what you were Hoping for. Yeah, those are the hard ones to kill. Any examples that kind of come for you?
- [28:03]
Leah Smart
Yeah, there was a recent one that we did around little Edu bubbles when you subscribed to a premium part of Hinge. So you'd subscribe to Hinge plus or Hinge X and then we would sort of like walk you through and be like, you can do this and also you can do this over here. And it created a lot of complexity and interacted with other bubbles and it didn't hurt any metrics, but we didn't see subscriber retention go up and we saw like slight increases in usage of some of the premium features, but it didn't translate down downstream to actually like subscription reuptake. And so it's like, so are we going to leave this in here and maintain it? And every time we change a feature we're going to have to go back and, you know, change these edus. So we killed it. That's one that comes to mind. But there's little things like that that happen all the time.
- [28:53]
Jessi Hempel
It's a great example. Yeah.
- [28:55]
Leah Smart
And when you do that, it forces teams to think big because no one wants to work on something that is going to get thrown out at the end. Right. By pushing that discipline into the organization, people are going to think twice about doing things that feel kind of marginal in the future.
- [29:11]
Jessi Hempel
Great example. Because I think many ways you're almost having a threshold for. I don't want to say just positive impact. I want to see positive impact that justifies the investment in the feature in the long run. The complexity that we're potentially putting on the member, on the user, like there's a higher bar for what gets to stay.
- [29:29]
Leah Smart
Yes.
- [29:29]
Jessi Hempel
Now for many teams, and this is kind of the psychology behind it, they invested so much. It's positive. Let's just let it be and move on to the next feature. So it's a really harsh call to say no, maybe it's positive, but it's not positive enough. It hasn't crossed the bar.
- [29:45]
Leah Smart
Yeah, it has to really move that end metric and it has to move it meaningfully. Otherwise you're right. Like you have to consider how much complexity is this adding how much maintenance, upkeep, interaction with other features is this going to have. And it can really slow you down over time to keep stuff like that in. It just starts gunking up the system.
- [30:05]
Jessi Hempel
One principle you had, which I spent time thinking through and wanted to get your take was build with, not for now. When we build, we. A lot of talk about like, we're building for people. We're building for you. I'm trying to build something awesome for you. So the. For what? Sounds actually pretty good. Sounds like a good one to tell people about why you build and how you build.
- [30:24]
Tomer Cohen
Build with.
- [30:25]
Jessi Hempel
Sounds very opinionated. Like, I'm building with you. It sounds great. But in many ways, for many teams, maybe they think they're close to the customer or they're not close to the customer all the time to build with them. I'm curious, what does this mean for you? So this is one where, like, I have to have, you know, go deeper with Justin around how this comes into play in Hinge.
- [30:43]
Leah Smart
Yeah, I mean, it starts with creating an incredibly diverse team. If we're really building for a diverse, increasingly fluid group of people who are using Hinge, we need to have those perspectives with us in house. Otherwise, we're not really going to develop the empathy and understanding and really do the prioritization of the work that matters. We are building for a very diverse group of people who are coming to us with all kinds of very different needs and perspectives on love and dating. We want all those perspectives in house so that we can grow as an organization and cater to all users. And if we don't have it in house, then we partner with the right agencies or the right nonprofits so that we bring those perspectives in house. You know, it's one thing to just, like, do user interviews and things like that, but it's another to make sure that the work that we do and by the way, not just the product work, but the marketing campaigns or the social impact work that we do are bringing in the diverse perspectives from the beginning.
- [31:49]
Jessi Hempel
So it sounds like it starts from having the team that is building, reflecting the customers you're trying to serve.
- [32:02]
Leah Smart
Yes. Because there's this nuance that gets missed if you don't have that.
- [32:05]
Jessi Hempel
Yeah. And that manifests itself in, you know, experiences or features that otherwise you would not have thought about.
- [32:12]
Leah Smart
Yes. Or just implementing features or measuring features in a way that maybe you didn't think about before.
- [32:18]
Jessi Hempel
But there's one actually, I think was called out, which was hidden words.
- [32:22]
Leah Smart
Oh, yeah. Hidden words. Yep. Totally.
- [32:24]
Jessi Hempel
Where before the app decided there were specific hidden words, and then you let users decide what kind of words they wanted to filter by. In a way. Curious when it comes to the mechanics. In a way. Like, it's a marketplace mechanics. And when I was studying it, it was like, there's likes and roses and turn limits, and I can assume a lot of it comes from intuition A lot of those are scientific. How do you balance those themes internally? This is like a left brain, right brain team that come together and you're trying to kind of orchestrate the ultimate matching kind of experience and then you at the top decide how you do it, or you go and test and iterate. There's a sense of uniqueness within hinge about how you create those mechanics and how they come to life.
- [33:09]
Leah Smart
Yeah, well, again, we're really grounded in that North Star metric. What gets people out on more dates? Like the like limit is a great example of that. What we find is that most people don't even come near the like limit anyway. But you have a few people that can send a lot of likes and that can crowd out the experience of people who don't send a lot of likes. And so it's actually beneficial and gets more people out on dates to have some limit so that everyone's on a more even playing field across the app. So limits like that really do make sense. Or your turn limits more recently, which actually reduced the number of matches significantly, but even more significantly increased the match to exchange rate so that people got out at the end of the day on more dates.
- [33:54]
Jessi Hempel
So you can only like a certain amount of profiles, for example.
- [33:57]
Leah Smart
That's the like limit. But the your turn limit is you can only have so many conversations going where it's your turn. If someone sent you a message and you're not responding, we then stop you from sending more likes or matching with more people. And so that gets people either to say like, okay, I'm not interested in this person anymore, or wait, I was interested in this person, I should probably respond to them. And that's another limit. It did reduce matches because people ended up talking more and sending fewer likes and matching less. But because people were having more conversations and exchanging more phone numbers, it ended up getting people out on more dates. And so that ended up being like a great positive feature. Interestingly, like, even during the testing of that, we had some people being like, oh my God, this is amazing. What, are you going to charge me for this? Charge you for it? Like, no, no, no. This is, this is like a free experience because this is what benefits everybody. And we have a principle at hinge about we only charge for what we can't give away for free. Meaning if it benefits everyone and gets more people out on dates, then we give it away for free. But if it's something that the scarcity is the value like a rose or a boost, or if it actually hurts the ecosystem to Give it to everyone for free then. Those are the things that we put into a subscription tier.
- [35:17]
Jessi Hempel
Now, when you talked about the onboarding experience, when it comes to matching, I'm assuming that the more you know about the user, the better the match. You can create the higher likelihood of a date. How do you think about the data you need from them? I think one element is just the. I was hearing a lot of rave about the voice prompts. When you hear somebody in voice prompts, there's also an authentic layer to it. And I'm curious, is there like a set of ultimate data you need or hey, in this age of AI, tell me more about you and I'll be able to filter it and try to match the whole thing to the other side. I'm thinking about multiple use cases today that even a very different example than dating, when I think about learning. A lot of AI tutors today would basically say, hey, what are you interested in? And then I'll match the questions in a way that matches your interests. So you're no longer working on a boring math problem. You're actually working on taking something to the moon, which you're really interested in. So it's like, it's a way to kind of learn more about your personality so I can personalize the experience more towards you and move away from building something which is more generic.
- [36:24]
Leah Smart
Yeah.
- [36:25]
Jessi Hempel
I'm curious, when you think about data in this way, is it more about making the app more personal and not just about the match or even for the match? Like, there's fields that you would love to get from people, but they're not sharing today yet.
- [36:37]
Leah Smart
Certainly there's more information that we can be getting from people that they're not sharing today. And there's more things that we can be doing with the data using AI. So these are two really interesting and very important vectors that I think are going to really start to transform hinge over the next couple years, really, or even over the next 12 months. We're completely overhauling our recommendation system to be powered by AI. So part of that is, as you said, picking up on signals you weren't picking up on before, factoring in all the information, your prompts and your photos and all this to make really thoughtful, intelligent matches and make connections that we kind of weren't able to make before and get better and better at predicting who you're going to like and who's going to like you back. And then on top of that, we can add more information that is less structured than what we were used to. Being able to handle and actually factor that in in really interesting ways. So instead of just getting your. All the profile information that's mostly structured. Yeah. We have three prompts and then we have your six photos. You know, what's your religion, what's your like, all that kind of stuff. But now we're starting to test things like just talk to us a little bit about who you are and what really matters to you and what kind of person you're looking for and being able to take that information and shape your recommendations in a way that's much more nuanced and thoughtful than we were able to do before. LLMs and Generative AI. So both those vectors are ones that were moving down very rapidly in parallel.
- [38:09]
Jessi Hempel
So in many ways you'll have your profile which is structured. People will see what you decide to put out there, but there'll be a deeper set of knowledge about you that can be used for your match. Ultimately.
- [38:21]
Leah Smart
Yeah, I think that's a really exciting piece of where the future is going. Yeah. Right now you create a profile for yourself and what you put on that profile is what other people see. There's some fields that you can hide, you can choose to hide. But it's very kind of like here's what I'm putting out to the world. But now being able to sort of speak to us on the side to say like, here's my profile hinge and at the same time here are some things that I want you to think about as you start to match me up is yeah, a really cool and interesting world that we're going to be able to be in to I think get a deeper and more nuanced understanding of what people really want.
- [38:57]
Jessi Hempel
I'm assuming it's exciting at hint right now just thinking about all the changes you can make. I'm curious if you have to sometimes ground the team with like, hey, here's what's not going to change or here's where I'm not willing to bend the store AI or you know, thinking about this differently from an AI perspective, obviously the ultimate goal is humans interacting with humans and building long lasting relationships. But are there conversations when you're pushing on like, or the team is asking you like what do you believe will not change or how does that get created?
- [39:28]
Leah Smart
Yeah, definitely with AI specifically, you mean yes or beyond.
- [39:34]
Jessi Hempel
But curious from an AI perspective as well.
- [39:35]
Leah Smart
I think that's why it's important. Once again we have our book and we really root in some, you know, I think about your mission and your values as Things that change almost never. I mean, our mission and our values have been, like, very core to us for, like, a very, very long time. Then there's the principles level of things which change, I'd say, like, every few years. And then there's your company objectives, which change, like, every year or two. And so this kind of like concentric circles. And, like, the closer you get to the core, the less it should change or maybe not change almost at all. And some of those values are really shaping and showing up in ways, especially authenticity when it comes to AI, that we think, okay, how do we now come up with principles? Because we have, yes, those top four level principles he talked about, like, love the problem and keep it simple, tend to trust. But even when we build a new, like just a new feature or build a new initiative, we develop principles for it. And we've been developing principles around AI as well. And one of the biggest ones for us is that AI really should stand behind us and not between us. Meaning we want you primarily interacting with other people. And we certainly don't want AI speaking for you or generating images for you. Right. So one thing that we just released, for example, is prompt feedback. We don't write prompts for you. We don't suggest what to say, but we will give you really thoughtful feedback on your prompt and how you can make it better trained on what we have seen work for, what kinds of prompts actually generate conversations and which don't. So we won't let you copy and paste and put things in there that you didn't write. But we will say, hey, you just wrote brunch as an answer to my favorite Sunday. Where do you like to go? What do you like to do? Do you like to cook? And then they'll say, brunch, and I like to cook. And then we'll say, what do you like to cook? Like, sometimes you just have to pull it out of people and help them understand that by answering with more depth and vulnerability and authenticity, you can help coach them so that they get better matches. So that's like, one example of how we're using it to make sure that we really preserve authenticity in the experience. And you don't feel like you're Talking to an AI, or AIs are talking to AIs on your behalf or something like that. We, like, want to preserve the very human, messy experience of dating.
- [42:02]
Jessi Hempel
Going back to love the problem, I can think of new problem sets to go after. So, for example, go beyond the app, beyond the match to, I'm assuming for if you lack confidence when you go on a date, you could use some coaching. During.
- [42:20]
Leah Smart
Before, after coaching is a huge piece of where we're going. Two big vectors, basically, that I think AI is taking us down. The first is personalized matching. So this is what we were talking about before. Using the data that we do have. Better making personalized introductions, gathering more nuanced data that we can use to match you. That's all great, but no matter how well we can match you, some people just need help filling out their profile. So prompt feedback is one early, very simple example to much more nuanced and thoughtful coaching. And so we have a whole team called Hinge Labs, which is always doing research on what works, what doesn't. How do people succeed on Hinge and in dating generally. And we've taken that data and we've published date reports to the press. You know, we have a section in our app called what Works, where you can go see what kind of photos to pick. People don't really read that stuff. If they do, they don't really know how to apply it. And what's so exciting about where we're going is we can take this whole body of knowledge and now apply it to you in a very personalized way and deliver the right piece of feedback in the right moment. And that's like another huge vector that I think is going to help a lot of people because many people struggle to get that first match or have that first date. A lot of times it is because they just don't know to, like, fill out a more detailed prompt or choose better photos. Maybe they're not sending enough likes. There's a whole host of things that we could be giving them tips around that we just today haven't been able to. And so I think it's going to help a lot of people to be able to receive those just little nudges and moments in coaching along their journey from filling out their profile all the way to what do I talk about on this date tonight?
- [44:01]
Jessi Hempel
And that's where you see, in a way, Hinge kind of progressing where, like, the date is that I focus on getting the date. You know, that's like step one. Step two is like, actually transform that date into a relationship.
- [44:12]
Leah Smart
Yes, definitely.
- [44:14]
Jessi Hempel
Wonderful. Justin, I've learned so much, and I really appreciate the candor and also kind of the principled way of thinking, plus the nuances on how you bring this to life at Hinge. And I wanted to thank you for all the learnings. It's wonderful.
- [44:25]
Leah Smart
Yeah, yeah, it's been a really good conversation. Thanks for having me, and thanks for such great questions.
- [44:29]
Jessi Hempel
Thank you so much.
- [44:32]
Tomer Cohen
This wouldn't be an episode of Building One without some great takeaways, so let's just jump in. First, as one might expect, rebooting a company is brutal, but sometimes very necessary. Justin made the difficult decision to tear down Hinge and start all over again when he realized the company drifted from its original mission. The early version of Hinge was competing with Tinder in a race for engagement, and that wasn't why he built it. This reboot refocused Hinge on helping people build meaningful relationships, not just match endlessly. This approach did not guarantee the success Hinge is seeing today, but it was clear that without a reboot there would be no chance of success at all. Second, Intentionality builds better products and better matches Hinge's onboarding process intentionally filters out people who aren't serious about dating, losing 20% of signups in the process. That's a feature, not a bug. The platform also requires users to put in more effort, like commenting on a specific part of somebody's profile, or even sharing something about themselves via voice recording. That leads to a more thoughtful interaction and high quality matches. Third, find good proxies for your True north metrics. This is something we covered a bit on the show before, but it's especially relevant in Hinge's case. Hinge's North Star metric is people getting out on dates and then ultimately relationships. But all of it happens physically, not digitally. It's off the app. It can't be measured. Therefore, Hinge looks for all sorts of opportunities to infer or get feedback from users. Did they exchange contact information? Did they indicate that they actually met their match when they were surveyed? If someone decides to leave Hinge, they ask why. It might not be perfect information, but pieced together, it gives them visibility into how well the product is doing. Fourth, consider how AI can either help or hurt the human experience. For example, Hinge wants to help its users go out on good dates, but he doesn't believe it should be doing the work for them. Hinge believes that AI can coach people on how to build a better, more thoughtful profile. If someone's answer to a prompt is too simple, it's not going to generate a conversation. So instead AI should prompt the user to give a more thoughtful, deep response. This is not unlike Khan Academy's own approach to AI. The goal is to encourage people to do the meaningful part of the work, because if they just have the answers, they'll fail on their own later on. Lastly, I love how Hinge operates with a clear set of product principles. It's on their website and you can clearly hear it from Justin. These principles help the team make tough decisions quickly and ensure the product evolves without losing its core purpose. Strong product principles help create a stronger company.
- [47:23]
Justin McLeod
Want to hear more from Justin? Then check out his podcast interview on this is working with Dan Roth, LinkedIn's editor in chief. We'll be back in two weeks with Rohan Amin, the CPO of Chase. Building one is a production of LinkedIn News. Our host is Tomer Cohen, LinkedIn's chief product officer. This episode was produced by Max Miller. Our associate producer is Rachel Karp. This episode was mixed by John Partham and engineered by Assaf Gadram, and we get additional production support from Alicia Mann at LinkedIn News. Sarah Storm is senior producer. Dave Pond is head of productions and creative operations. Maya Pope Chapelle is director of content and audience development. Courtney Koop is head of original programming. Dan Roth is the editor in chief of LinkedIn. If you know a product leader we can all learn from, send us a line@pitchesinkedin.com.