0:00

LinkedIn News. innovation ultimately

0:02

is solving problems. It

0:04

came back to my own personal

0:07

pinpoint. It just takes a while

0:09

to build that trust. The problem

0:11

that I was trying to solve,

0:13

it was all I was thinking

0:15

about. You have to be obsessed

0:17

with the human condition. I'm Tomer-Coin,

0:20

Chief Parak officer of LinkedIn,

0:22

and this is building one. We

0:24

lose 20% of the people plus during the

0:26

onboarding process. You know, if you were talking

0:29

to a VC, you'd be like, well, that's

0:31

crazy. That's Justin McCloud, the founder of

0:33

Hinch. He's talking to me about how

0:35

Hinge designs its onboarding for new users

0:37

to make sure they're right for the

0:39

app. We're going to get into that and so much

0:41

more, so stick around. From

0:49

LinkedIn News, I'm Leah Smart, host

0:51

of Everyday Better, an award-winning podcast

0:53

dedicated to personal development. Join me

0:55

every week for captivating stories and

0:58

research to find more fulfillment in

1:00

your work and personal life. Listen

1:02

to Everyday Better on the LinkedIn

1:04

podcast network, Apple Podcasts, or wherever

1:06

you get your podcasts. Did you

1:09

know that small and medium businesses

1:11

make up 98% of the global

1:13

economy, but most B2B marketers still

1:15

treat them as a one-size-fits-fits-all? Linkton's

1:18

Meet the SMB

1:21

report reveals why

1:23

that's a missed

1:26

opportunity and how

1:28

you can reach

1:31

these fast-moving decision-makers

1:36

effectively. Learn

1:38

more at LinkedIn.com,

1:40

backslash, Meet-the-SMB. Imagine

1:42

putting years of

1:44

work into building

1:46

something, seeing it

1:48

grow, and then

1:51

realizing it's not

1:53

exactly right. That's the moment

1:55

where most founders freeze. Do you

1:57

push forward? Do you pivot? Or

2:00

do you take the truly radical step

2:02

of tearing it all down and starting

2:04

all over again? My guest today faced

2:06

the exact same moment with his own

2:08

company. Today, I have the pleasure of

2:10

speaking with Justin McCloud, the founder of

2:13

Hinge, currently the fastest growing dating up

2:15

in the US. Hinge initially launched in

2:17

2012, but Hinge as we know it

2:19

today, didn't really start until 2015, when

2:21

Justin decided to reboot the company. and

2:23

build it around a few core product

2:26

principles. As the dating app, designed to

2:28

be deleted, hinges built as a product

2:30

for intentional daters, and one that has

2:32

the ultimate goal of getting them into

2:34

relationships and off the app. Justin's focus

2:36

and discipline in product building really stands

2:38

out. In my conversation with him, we

2:41

learn about things like how placing limits

2:43

on user actions can create better interactions

2:45

overall. Why Justin is proud to lose

2:47

more than 20% of new users during

2:49

the onboarding process? How hinge combats feature

2:51

in scope creep? And the role Justin

2:53

sees for AI in improving our dating

2:56

lives. That and a lot more. So

2:58

let's just jump in. Justin, it's a

3:00

pleasure to have you on the show.

3:02

Thank you so much for joining me

3:04

today. Yeah, thanks for having me. So

3:06

I thought we'll start kind of early

3:09

with your journey and... Like many of

3:11

our guests, you built Hinge out of

3:13

a deeply felt need. You wanted to

3:15

find a girlfriend as far as I

3:17

understand. Now, most people, 99.9% of them,

3:19

would just try whatever is out there,

3:21

but you decided to build something. So

3:24

I'm curious why? What drew you to

3:26

go and build a solution versus just

3:28

use what's out there for you? Yeah,

3:30

it's a long story. And then there's

3:32

the reboot 2015, 2016 Refounding of Hinge.

3:34

So we're talking about 2011, originally, like

3:37

why did I get into this in

3:39

the first place? And this was the

3:41

Facebook app, uh, hinge. Yeah, Facebook app,

3:43

and then it was a mobile app.

3:45

but it was a friends-of-friends kind of

3:47

swiping experience. But this is 2011, so

3:49

this is predating apps, right? There is

3:52

dating websites like Match and E Harmony

3:54

and things like that. The day I

3:56

graduated from college, I got sober, I

3:58

got sober, I worked for a few

4:00

years, and I went to business school,

4:02

and business school is like this very

4:04

social environment, but I was struggling really

4:07

to meet people because I just didn't,

4:09

it was like a totally foreign world

4:11

to me being able to go out

4:13

to partieses and like not drink and

4:15

everyone else is drinking. and I'd actually

4:17

reached out to my college girlfriend, Kate,

4:20

to try to win her back and

4:22

she told me in so many words

4:24

to shove it, I was heartbroken and

4:26

looked at the services that were out

4:28

there and just felt like, gosh, this

4:30

isn't really something that I would use,

4:32

it's not something that my friends use.

4:35

We all use, at the time, Facebook.

4:37

I think that there's an opportunity to

4:39

create something that would be... accessible and

4:41

fun and kind of easy and approachable

4:43

for people of my age, people in

4:45

their 20s to use. So that really

4:48

was the original thought and insight. I

4:50

wanted it for me, but I also

4:52

was like, gosh, almost none of us

4:54

use matry harmony if we do, we're

4:56

not talking about it. And why is

4:58

that? And started to really get underneath,

5:00

like, what was the reason that people

5:03

didn't use these services? Well, they were

5:05

expensive. People used kind of screen names,

5:07

which was very passé at the time

5:09

of now. We had Facebook. People were

5:11

used to using their real names. You

5:13

had to fill out, like, very long

5:15

detailed, revealing, vulnerable profiles that were, like,

5:18

essay questions. It just wasn't the way

5:20

that people kind of interacted in my

5:22

age. And so felt like, what if

5:24

you built a new service that really

5:26

appealed to younger people? that was, I

5:28

would say, like lower lift, more fun,

5:31

more easy, got you in the door,

5:33

and just made it approachable. If it

5:35

wasn't so expensive and it wasn't so

5:37

hard, then I think people would feel

5:39

more comfortable using it, saying they used

5:41

it. it. And I also just always

5:43

kind of been a romantic at heart

5:46

and I've always loved like systems and

5:48

math and algorithms like always been kind

5:50

of what I gravitated to as always

5:52

my strong point in school. And I

5:54

don't know all just kind of converged

5:56

as like a hit of I just

5:59

know I'm meant to do this. I

6:01

had been working on in business school

6:03

a number of different projects that I

6:05

was going to enter into a business

6:07

plan competition. but all that kind of

6:09

felt like homework. It all felt like

6:11

I have to work on my thing

6:14

now. And then as soon as this

6:16

idea hit me, it was like, it

6:18

was like this thing is coming through

6:20

me. It's like all I could think

6:22

about, it's all I wanted to work

6:24

on. I just felt like I just

6:26

had to make it a reality. Yeah,

6:29

we hear this quite often from founders

6:31

that deeply felt need resonates with them

6:33

on the level where they cannot not

6:35

build it. Yeah, that's how I felt

6:37

like I felt like I couldn't not.

6:39

build it. I just, it was coming

6:42

out of me one way or another.

6:44

You talked about the second incarnation of

6:46

Hage in 2015, where you call the

6:48

reboot. So it sounds like in many

6:50

ways a starting point was, you know,

6:52

Web 100 was kind of dominant and

6:54

came really the social networking web with

6:57

a lot more thinking around friends, authentic

6:59

identities, also mobile. So then, you know,

7:01

you're after the racist. Extremely it looks

7:03

like hinges doing really well, you decided

7:05

to reboot. What triggered that? It was

7:07

a couple things. The real moment was

7:10

there had just been an article written

7:12

that was very popular in Vanity Fair

7:14

called the Dawn of the Dating Apocalypse,

7:16

and it was about how all of

7:18

these apps had just become a place

7:20

for hookups, and they were superficial, and

7:22

they were ruining romance and love and

7:25

dating, and dating. Hinge was heavily featured

7:27

in that article. Also, a couple other

7:29

dating apps had come out since then

7:31

that were becoming even more popular than

7:33

Hinge. specifically tender. They beat us at

7:35

the game of making it easy and

7:37

approachable. A swipe left and right. Like

7:40

really the market was ready for just

7:42

like a let's make this really simple

7:44

and really approachable and it caught fire.

7:46

We were trying to keep up with

7:48

the competition and you know looking at

7:50

what features they were releasing and we

7:53

were releasing them and then we were

7:55

just becoming you know if you're going

7:57

to copy someone don't copy the people

7:59

that are already bigger than you. I

8:01

don't think it's a winning strategy. we

8:03

were just less and less differentiated and

8:05

so I had less and less of

8:08

a place to be in the market.

8:10

But more importantly than that, it just

8:12

didn't resonate with me where it was

8:14

going. Like I really did as we

8:16

talked about start this company because I

8:18

like wanted to find my my person.

8:21

I wasn't trying to create like a

8:23

casual encounters type experience. And actually went

8:25

and even like met with the Nancy

8:27

Joe sales, the person who wrote that

8:29

article and sat down with her and

8:31

sat down with her. Chief Brand Officer

8:33

Katie Hunt, and we were about to

8:36

go home for Thanksgiving, it was 2015,

8:38

and I'm like, gosh, this isn't the

8:40

company that I want to build. I

8:42

kind of just want to like tear

8:44

the whole thing down and start over

8:46

again. And she's like, well, what's stopping

8:48

you? Like, you're the CEO, if that's

8:51

what you want to do, then let's

8:53

do it. And I went home over

8:55

Thanksgiving break, and I thought about it.

8:57

And I'm like, yeah, that's what we're

8:59

what we're going to what we're going

9:01

to do. let go of half the

9:04

company, reboot, throw out the old code

9:06

base, and start over from scratch, and

9:08

build something because a new opportunity had

9:10

emerged. Like originally the opportunity was make

9:12

this thing simple, easy, approachable, and that

9:14

was obviously like a winning strategy in

9:16

many ways. And at the same time,

9:19

for people who were like really looking

9:21

for their person, it was less effective,

9:23

we'll say, at that. because you didn't

9:25

know a lot about these people. It

9:27

didn't know a lot about you. It's

9:29

hard to make really smart, intelligent matches.

9:31

And now that people are socialized to

9:34

the idea of using dating apps. and

9:36

create something that really is about less

9:38

winning because it's so simple and easy,

9:40

but more because it really strikes that

9:42

right balance of remaining approachable to the

9:44

next generation, but also really asks more

9:47

of them, put in more effort and

9:49

get more out of it, and really

9:51

help you zero in on your person

9:53

and be the dating app that's designed

9:55

to be deleted. So I love that.

9:57

And it sounds like part of it

9:59

was like the market pushing you, part

10:02

of it was like you intently felt

10:04

like you wanted to build something different.

10:06

I'm curious when you think right now

10:08

on the target audience that you know

10:10

you're building for your ideal kind of

10:12

you know when you look at two

10:15

people you're saying hey you should be

10:17

on hinge you should not like hinge

10:19

is not for you're looking for casual

10:21

connections there's so many apps for that

10:23

you should go there how do you

10:25

think about bringing that to the product

10:27

so obviously there's branding I love the

10:30

designed to be the deleted thing it

10:32

goes really well when you think about

10:34

onboarding Do you think about how do

10:36

you attract the right people around in?

10:38

Actually, also how do you tell the

10:40

folks who should not be on the

10:42

platform that there's other apps for them?

10:45

We have an extensive onboarding process and

10:47

we lose 20% of the people plus

10:49

during the onboarding process. if you were

10:51

talking to a VC, you'd be like,

10:53

well, that's crazy. You shouldn't lose 20%

10:55

of people during onboarding. But for us,

10:58

in a sense, it's a feature because

11:00

it's a filtering mechanism for people who

11:02

really want to put in the right

11:04

amount of effort and are intentional. And

11:06

our target audience, we use the word

11:08

intentional daters, people who want to put

11:10

in thoughtful effort because they're looking for

11:13

a real result. They're here because they

11:15

want to get off dating apps. That's

11:17

who we think about as we're building

11:19

things. So we ask our dators to

11:21

do more, to fill out three prompts,

11:23

to add six photos, to give us

11:26

lots of demographic information, to take thoughtful

11:28

actions. Once you're in the product, it's

11:30

not just the onboarding. Instead of using

11:32

the swipe feature, you actually have to,

11:34

if you like someone, like something about

11:36

them, and you can add a comment,

11:38

which makes people way more thoughtful and

11:41

selective in their liking. to like someone

11:43

to see if they liked you back,

11:45

so there's no kind of game of

11:47

finding the experience. If someone likes you,

11:49

we just deliver it right to you,

11:51

we tell this person liked you, would

11:53

you like to match with them? It

11:56

does ask people to put themselves out

11:58

a little bit more and put in

12:00

a little bit more effort, but it

12:02

allows us to learn their tastes much,

12:04

much better. It creates for those people

12:06

really. incredible experience, which is why I

12:09

think we've grown so fast through word

12:11

of mouth. Why were the fastest growing

12:13

major dating app now? And was there

12:15

tension there just in in terms of,

12:17

you know, the holy grail of consumers,

12:19

just like let them in and like

12:21

walk them kind of progressively for the

12:24

product but don't add so much friction

12:26

at the beginning? Was there tension for

12:28

you there from the team or because

12:30

you actually had the reboot? It was

12:32

like, I actually know what I want

12:34

to build right now, so let's move

12:37

away from... the quick, you know, tinder

12:39

like interactions that we had in the

12:41

past. Yeah, this is a huge tension,

12:43

right? We were trying to find the

12:45

right balance between simple fun and easy

12:47

and super, like we didn't want to

12:49

go as far as like asking people

12:52

to fill out long essays and things

12:54

like that. There's like almost like an

12:56

efficiency frontier of vulnerability, right? It's like,

12:58

what are people willing to do, but

13:00

that also puts them out there enough

13:02

and ask enough of them that we

13:04

can really help them get the result

13:07

that they want? a good analogy would

13:09

be food. You can either make fast

13:11

food that's horrible for you but tastes

13:13

great but burns people out. Or you

13:15

could have food that's so nutritious, it's

13:17

kind of gross. Like how do you

13:20

make it sweet green, right? Like how

13:22

do you make it kind of fun

13:24

and add enough little dressing and the

13:26

right combination of flavor so it's nutritious,

13:28

but you want to eat it. We're

13:30

always trying to figure out what is

13:32

the right balance of getting people to

13:35

do what is good for them in

13:37

the long run. Like some of the

13:39

earlier explorations we swung it too far

13:41

to the other side we were like

13:43

well everyone's gonna have to pay seven

13:45

dollars a month to be on our

13:48

platform if you like someone you have

13:50

to send a comment with your like.

13:52

And we tested these things and it

13:54

was too far. It was asking too

13:56

much of people and people didn't want

13:58

to engage in the thing collapsed. So

14:00

it's definitely this like constant honing of

14:03

trying to push people and push out

14:05

that frontier of like how do you

14:07

make people a little bit more authentic

14:09

and put in a little bit more

14:11

without. turning them off and discouraging them.

14:13

I love that idea of efficiency frontier

14:15

for vulnerability. What would be the excess?

14:18

So one would be vulnerability and one

14:20

would be effort or? One is like

14:22

what people are willing to do and

14:24

then what's effective. Right? Like there's kind

14:26

of a natural linear relationship. This whole

14:28

concept started when we started optimizing prompts.

14:31

So prompts are those short little questions

14:33

that we ask that. People will... Those

14:35

are voice problems, right? Do you want

14:37

to hear the voice as well? Yeah,

14:39

yeah. The voice is another actual manifestation

14:41

of that. Video, too far. Like, reveals

14:43

a lot about you, but most people

14:46

aren't willing to do it. Simple text,

14:48

too easy. People are definitely willing to

14:50

do it, but it doesn't give you

14:52

a lot of the... And voices, there's

14:54

this amazing sweet spot with voice, and

14:56

I think that voices can become a

14:59

bigger and bigger part of dating. You

15:01

get enough... to like understand someone's personality,

15:03

but people are willing to do it.

15:05

But specifically what I was talking about

15:07

with prompts is when we started doing

15:09

this, we would measure how effective is

15:11

this at getting a conversation going. and

15:14

are people willing to answer this prompt?

15:16

So there are prompts like my go

15:18

to karaoke song, which is no longer

15:20

a prompt, because everyone was willing to

15:22

answer my go to karaoke song and

15:24

no one cared what your go to

15:26

karaoke song was, right? Very few good

15:29

conversations start because of like your named

15:31

go to karaoke song. And at the

15:33

other end of the spectrum, very deep

15:35

vulnerable questions. If people were willing to

15:37

answer these questions. they would lead to

15:39

great conversations, but no one was willing

15:42

to answer them. It was like pushing

15:44

people too far to put that on

15:46

their dating profile. So it's just like,

15:48

again, that sweet spot of like, what

15:50

I'm looking for in a plus one

15:52

wedding date, which is like something that

15:54

people are willing to answer, but reveals

15:57

enough to get it. good conversation going

15:59

and signals to you what kind of

16:01

values that you share. So there's that

16:03

real art to it. Love that. When

16:05

you do a reboot like that, when

16:07

you think about true north, right, you

16:10

have your vision, your mission, the app

16:12

to be deleted, talks of volume about

16:14

what you're trying to help people actually

16:16

find great relationships, not just go on

16:18

dates. When you try to put a

16:20

metric on that, that feels like really

16:22

hard. I'm curious how you think about

16:25

capturing that quantitatively. So you can tell

16:27

the team we're in the right direction

16:29

versus a notion of like before, when

16:31

you're concerned at like, hey, the wrong

16:33

metric can actually put us on the

16:35

wrong path. Our North Star metric is,

16:37

I mean, ultimately, ultimately it's getting people

16:40

off the app and into relationships, but

16:42

it's hard to. measure a feature based

16:44

on that, right? Because there's a long

16:46

lead time between releasing a new prompt

16:48

and saying like, gosh, did this get

16:50

more people off our app and into

16:53

relationships? But what we do use is

16:55

getting people out on dates. Are people

16:57

exchanging contact information and moving out on

16:59

a date? And that really is our

17:01

North Star. And so a product feature

17:03

is good if it moves that up

17:05

and it's bad if it moves that

17:08

down. So that really helps. clarify things

17:10

for us. And you know that through

17:12

the app or this is something you

17:14

do after the fact? We do have

17:16

something called we met in the app

17:18

so if people do exchange contact information

17:21

the app will follow up to ask

17:23

if they went on a date and

17:25

if this is the type of person

17:27

they want to see again but feature

17:29

by feature if we're really determining like

17:31

is this a winner or not we're

17:33

looking at do people exchange phone numbers

17:36

or contact information and we want to

17:38

move that number. And it's hard to

17:40

move, right? I'm sure you all have

17:42

your own kind of user funnels, right?

17:44

You push in one place and it

17:46

goes down there, but then it pops

17:48

it up in another place, right? So

17:51

you can get people more matches, but

17:53

then the match to conversation rate goes

17:55

down and then it nets out. And

17:57

when you're designing... product teams and everything

17:59

else, you're not having one team that's

18:01

pushing on one part of the funnel,

18:04

but only to watch it hurt another

18:06

team's metric at another part of the

18:08

funnel. How do you think click in-to-end

18:10

and cohesively about that? That's hard for

18:12

us as well. So, you know, our

18:14

true Norfolk is economic opportunity for people.

18:16

And sometimes we see it in the

18:19

app because people actually find a job

18:21

so you can see the coming in

18:23

or you can see them building a

18:25

business, but that's also becoming a dating.

18:27

has to be there and then the

18:29

proxy for dating is exchanging information. But

18:32

anyway, people are trying to find all

18:34

the ways to, you know, put guard

18:36

bills around the metric. And then what

18:38

I found in the past that, you

18:40

know, sometimes it becomes very complex and

18:42

then you run the risk of the

18:44

metric is no longer as simple as

18:47

it was before. And then, you know,

18:49

it's hung clear if you're potentially like

18:51

making a dent or not. You do

18:53

something which is really interesting. You actually

18:55

follow up with members after you might

18:57

not be sampling everybody because that's very

18:59

hard. But I'm assuming you have a

19:02

strong qualitative measure of like are we

19:04

pacing in the right direction towards our

19:06

vision? Yeah, I would say that we

19:08

met data and also the exit survey

19:10

data. So if someone leaves Hinge, we

19:12

say, you know, did you find someone?

19:15

If you found them, we're this on

19:17

Hinge, you're leaving because you're just done

19:19

with dating apps, you're done with Hinge.

19:21

So we can use those to kind

19:23

of make sure we're steering in the

19:25

right direction generally. And when we make

19:27

improvements in dates per user, it's actually

19:30

good dates and good dates that are

19:32

leading into relationships, which we generally do

19:34

find. I think your metric of economic

19:36

opportunity is probably quite hard to measure

19:38

and comes in many different forms. For

19:40

us, it actually is relatively simple. Like,

19:43

we've got to get people out on

19:45

good dates. Like, are these people chatting

19:47

with people and exchanging information and planning

19:49

dates, or are they not? And that

19:51

is clarifying when it's that simple. Curious,

19:53

do you also look at people who

19:55

stay in the up for a long

19:58

time and look at that as a

20:00

sign of, okay, there's a cohort or

20:02

a need, we're not delivered? Yeah. So

20:04

something that was kind of interesting that

20:06

happened over the last like three years

20:08

or four years is that Gen Z

20:10

became our fastest growing segment like people

20:13

18 to 24 like started just growing

20:15

really really fast which was surprising to

20:17

us honestly because we think we're just

20:19

like this super relationship intention like find

20:21

your person get off the app and

20:23

most 21 year olds that you talk

20:26

to probably aren't like trying to find

20:28

their forever person. And yet what we

20:30

found is that just the authenticity of

20:32

the experience and the realness and the

20:34

intentionality was more universal. Some people are

20:36

in an earlier stage of their dating

20:38

journey, but they still want to go

20:41

on good dates with quality people. They

20:43

don't want to just show up on

20:45

a date and feel like this is

20:47

a random stranger. And that's not bad

20:49

for us in a sense, right? We

20:51

actually use these different ideas like explorers

20:54

and journeyers and destination people. Like are

20:56

you here because you're like, let's just

20:58

do this, I'm ready to find my

21:00

person, get off? Is it like, I'm

21:02

pretty ready to find my person, but

21:04

I'm still kind of figuring out what

21:06

I want? Or is it like, I'm

21:09

just here to see what's out there?

21:11

And I think as long as people

21:13

are approaching it with authenticity and vulnerability

21:15

and willing to put in effort, then

21:17

hinge is definitely a place for them,

21:19

which is why we've added things like

21:21

relationship and tensions. on Hinge, which we

21:24

avoid it for a long time because

21:26

we're like, well, everyone here should be

21:28

looking for like their person. So why

21:30

would we add relationship intentions? That'll dilute

21:32

our brand. But what we found is

21:34

like some people just want to have

21:37

our same ethos when it comes to

21:39

their dating lives, but even though they

21:41

maybe are looking for a polyamorous relationship

21:43

or they're just earlier in their dating

21:45

journey, and that's okay too. As long

21:47

as people are able to align expectations,

21:49

they do. In many ways to your

21:52

point you maintain the core you care

21:54

so much about right now. Yeah, our

21:56

North Star is still getting people out

21:58

on days like that doesn't change like

22:00

even if you're 18 like we still

22:02

want to get you. off our app

22:05

and out on a good date with

22:07

somebody. We're going to take a quick

22:09

break, but don't go anywhere. When we

22:11

come back, Justin is going to explain

22:13

why even really good features may not

22:15

make the cut. You have to consider

22:17

how much complexity is this adding, how

22:20

much maintenance, upkeep, interaction with other features

22:22

is going to have.

22:24

It can really slow

22:26

you down over time.

22:28

Small and marketers trying

22:30

to reach them. On

22:32

LinkedIn, more SMB decision-makers

22:35

are actively looking for

22:37

new solutions to help

22:39

them grow, whether it's

22:41

software or financial services.

22:43

Our Meet the SMB

22:45

report breaks down how

22:47

these businesses buy and

22:49

what really influences their

22:52

choices. Learn more at

22:54

LinkedIn.com, Meet-the-SMB. We're

23:00

back and I'm speaking with Justin McCloud

23:02

the founder of Hinch the dating app

23:04

so One of the things I loved

23:07

studying more about Hinge and how you

23:09

think about building was Kind of the

23:11

strong principles you bring into place and

23:13

in fact people can go on the

23:15

Haines website It's all there You called

23:17

them out pretty clearly and I love

23:19

that a lot of the things we

23:21

talk about in the show are principles

23:24

people work by I think it tells

23:26

a lot about how you build and

23:28

what kind of distinguishes your decision-making from

23:30

others One of them has been love

23:32

the problem. I see that if you

23:34

love the problem, the expectation from Justin

23:36

would be that you understand it deeply,

23:38

you can create a better solution for

23:41

it because you understand it. However, from

23:43

my understanding, this principle emerged from an

23:45

understanding of the old hinge to count

23:47

too many goals, and those goals didn't

23:49

necessarily advance hinges mission. Can you share

23:51

more about the connection between that principle

23:53

to that kind of scenario that you

23:56

had in the old hinge? two evolutions

23:58

of this principle. The first principle was

24:00

like in the early early days of

24:02

hinge, like Hinge v1, we really got

24:04

lost in terms of like engagement, retention,

24:06

monetization, like all these different social media

24:08

type metrics or business metrics, and we

24:10

were chasing many goals and not getting

24:13

really clear on like what are we

24:15

ultimately here to serve? And we lost

24:17

sight of actually getting people out on

24:19

good dates. We were too focused on

24:21

like, did they stay until week two?

24:23

Are they, are they staying from month

24:25

one to month two? But what we

24:27

realize, like, if you're really solving the

24:30

deep underlying user problem, then the scoreboard

24:32

sort of takes care of itself. So

24:34

that was one layer, I would call

24:36

it, of love the problem. And that's

24:38

kind of like, just get clear on

24:40

the problem that you really are ultimately

24:42

here to solve and what's your, like,

24:45

number one priority problem. The second is

24:47

getting a deeper understanding of a problem

24:49

and also making sure you're always rooting

24:51

in a problem. So what I often

24:53

find is that someone will have a

24:55

problem in their problem and then... they'll

24:57

come up with a feature, and then

24:59

we get more wedded to the feature

25:02

than the problem, and there's kind of

25:04

like this drift that happens. And then

25:06

pretty soon we forget why we're even

25:08

building this feature in the first place.

25:10

So people have different ideas about what

25:12

it's supposed to be doing in their

25:14

minds. This is, people have different ideas

25:16

about what it's supposed to be doing

25:19

in their minds. This still happens, it

25:21

happens, it's like, well, actually, is it

25:23

worth continuing to do, but it's this

25:25

other thing. And we're like, well, is

25:27

that what we're trying to solve here

25:29

or not? Do we want to keep

25:31

going with this or do we want

25:33

to kill it? It's having that discipline

25:36

to constantly be returning to, like, what

25:38

is the real big thing we're trying

25:40

to solve here and what's the best

25:42

way to solve that? And not getting

25:44

too wedded to a particular feature idea.

25:46

We say love the problem because it

25:48

is spending that extra time to really

25:51

get acquainted with the users and like

25:53

what their experience is always leads to

25:55

a better solution. We need to get

25:57

people on more dates, like I have

25:59

an idea, X, Y, and Z. Like,

26:01

it's a very superficial understanding of the

26:03

problem. but the more nuanced and depth

26:05

and humanity you put around a problem,

26:08

the more creative and interesting and effective

26:10

your solutions are going to be. So

26:12

I love a lot about what you

26:14

said around, you know, obviously from pushing

26:16

people to understand what they're solving for

26:18

in a deeper way so they can

26:20

clarify the nuances. I'm hearing for you

26:22

also. in a way when something, even

26:25

if it's successful, I want you to

26:27

bring back the problem you were supposed

26:29

to solve and see if you solve

26:31

the problem versus, hey, Justin, my metric

26:33

is up to the right, doing really

26:35

well, but your hypothesis was different than

26:37

what we were trying to solve. I

26:40

don't know if many actually bring back

26:42

the original hypothesis to connect it together,

26:44

and I think that's really a sign

26:46

of a great builder. Oh, we do.

26:48

It's tough because sometimes people work really

26:50

hard on building a feature and it

26:52

did some other things, but it doesn't

26:54

really do the thing that we need

26:57

it to do. And it adds complexity.

26:59

And this gets to our second principle

27:01

of keep it simple and they're very

27:03

related. Complexity always has a cost. An

27:05

extra feature always has a cost. And

27:07

people get wedded to like, well, I

27:09

built this feature and it didn't hurt

27:11

the metrics and it feels good, so

27:14

let's just put it in. And I'm

27:16

like, well, we put it in and

27:18

there's the maintenance costs, there's the bug

27:20

costs, there's the interactions that it's going

27:22

to have with future product features, and

27:24

it's not worth it. Like, in order

27:26

to keep this product simple, which is

27:29

a big piece of what's always rooted

27:31

us, it's down to our black and

27:33

white branding and hard conversations, and... stripping

27:35

out features and killing features. Maybe they

27:37

seem nice or people think they're cool,

27:39

but if they're not moving our North

27:41

Star metric, then they probably don't belong.

27:43

I think for myself, like, you know,

27:46

the easiest kind of features to make

27:48

a decision on the ones that are

27:50

not performing, it's an easy decision, move

27:52

out. The ones that are highly performing

27:54

is a decision, like keep investing. Then

27:56

you have that like massive amount in

27:58

the middle. Which is features that somewhat

28:00

perform, they move the metric, like you're

28:03

seeing some qualitative good feedback about them,

28:05

but it's not what you were hoping

28:07

for. Yeah. Those are the hard ones

28:09

to kill. Any examples that kind of

28:11

come for you? Yeah, there was a

28:13

recent one that we did around little

28:15

EDU bubbles when you subscribed to a

28:17

premium part of hinge. So. you'd subscribe

28:20

to Hinch Plus or Hinch X, and

28:22

then we would sort of like walk

28:24

you through and be like, you can

28:26

do this, and also you can do

28:28

this over here. And it created a

28:30

lot of complexity and interacted with other

28:32

bubbles, and it didn't hurt any metrics,

28:35

but we didn't see subscriber retention go

28:37

up. And we saw like slight increases

28:39

in usage of some of the premium

28:41

features, but it didn't translate downstream to

28:43

actually like subscription re-up take. And so

28:45

it's like. So are we going to

28:47

leave this in here and maintain it?

28:49

And every time we change a feature,

28:52

we're going to have to go back

28:54

and change these EDUs. So we killed

28:56

it. That's one that comes to mind.

28:58

But there's little things like that happen

29:00

all the time. It's a great example.

29:02

And when you do that, it forces

29:04

teams to think big, because no one

29:06

wants to work on something that is

29:09

going to get thrown out at the

29:11

end, right? by pushing that discipline into

29:13

the organization, people are going to think

29:15

twice about doing things that feel kind

29:17

of marginal in the future. Great example,

29:19

because I think many ways, like you're

29:21

almost like having a threshold for, I

29:24

don't want to see just positive impact,

29:26

I want to see positive impact that

29:28

justifies the investment in the future in

29:30

the long run, the complexity that we're

29:32

potentially putting on the member, on the

29:34

user, like there's a higher bar for

29:36

what gets to stay. Yes. Now for

29:38

many teams, and this is kind of

29:41

the psychology behind it. They invested so

29:43

much, it's positive, let's just let it

29:45

be and move on to the next

29:47

feature. So it's a really harsh call

29:49

to say no, maybe it's positive, but

29:51

it's not positive enough. Like it hasn't

29:53

crossed the bar. Yeah, it has to

29:55

really move that end metric and it

29:58

has to move it. meaningfully.

30:00

Otherwise, you're right, like, you have to

30:02

consider how much complexity is this adding,

30:04

how much maintenance, upkeep, interaction with other

30:07

features is this going to have. And

30:09

it can really slow you down over

30:11

time to keep stuff like that in.

30:13

It just starts gunking out the system.

30:16

One principle you had, which I spent

30:18

time thinking through and wanted to get

30:20

your take, was build with not four.

30:22

No. When we build, we allow to

30:25

talk about like we're building for people,

30:27

we're building for you, I'm trying to

30:29

build something awesome for you, so the

30:31

for what sounds actually pretty good. Sounds

30:34

like a good one to tell people

30:36

about why you build and how you

30:38

build. Build with sounds very opinionated. Like

30:40

I'm building with you. It sounds great,

30:43

but in many ways from many teams,

30:45

maybe they think they're close to the

30:47

customer all the time to build with

30:49

them. I'm curious, what does this mean

30:52

for you? So this is one where

30:54

I have to have, you know, go

30:56

deeper with Justin around how this is

30:58

coming to play in Hinge. Yeah. I

31:01

mean, it starts with creating an incredibly

31:03

diverse team. If we're really building for

31:05

a diverse, increasingly fluid group of people

31:07

who are using Hinge, we need to

31:10

have those perspectives with us in-house. Otherwise,

31:12

we're not really gonna... develop the empathy

31:14

and understanding and really do the prioritization

31:16

of the work that matters. We are

31:19

building for a very diverse group of

31:21

people who are coming to us with

31:23

all kinds of very different needs and

31:26

perspectives on love and dating. We want

31:28

all those perspectives in-house so that we

31:30

can grow as an organization and cater

31:32

to all users. And if we don't

31:35

have it in-house, then... we partner with

31:37

the right agencies or the right non-profits

31:39

so that we bring those perspectives in

31:41

house. You know, it's one thing to

31:44

just like do user interviews and things

31:46

like that, but it's another to make

31:48

sure that the work that we do,

31:50

and by the way, not just the

31:53

product work, but the marketing campaigns or

31:55

the social impact work that we do,

31:57

are bringing in the diverse perspectives from

31:59

the beginning. It starts from having the

32:02

team that is building or reflecting the

32:04

customers you're trying to serve. Yes. Because

32:06

there's just nuance that gets missed if

32:08

you don't have that. Yeah. And that

32:11

manifests itself in experiences or features that

32:13

otherwise you would not have thought about.

32:15

Yes. Or just implementing features or measuring

32:17

features in a way that you didn't

32:20

think about before. But there's one actually

32:22

I think was called out which was

32:24

hidden words. Oh yeah, hidden words. Yep,

32:26

totally. Where before the app decided there

32:29

was specific hidden words and then you

32:31

let users decide what kind of words

32:33

they wanted to filter by in a

32:35

way. Curious when it comes to the

32:38

mechanics in a way that it's a

32:40

marketplace mechanics. And when I was studying

32:42

it, it was like there's likes and

32:44

roses and turn limits and... I can

32:47

assume a lot of it comes from

32:49

intuition, a lot of those are scientific.

32:51

How do you balance those teams internally?

32:53

This is like a left brain, right

32:56

brain teams that come together and you're

32:58

trying to kind of orchestrate the ultimate

33:00

matching kind of experience and then you

33:02

at the top decide how you do

33:05

it or you go and test an

33:07

iterate. There's a sense of uniqueness within

33:09

hinge about how you create those mechanics

33:12

and how they come to life. Yeah,

33:14

well again, we're really grounded in that

33:16

North Star metric. what gets people out

33:18

on dates. Like the like limit is

33:21

a great example of that. What we

33:23

find is that most people don't even

33:25

come near the like limit anyway, but

33:27

you have a few people that can

33:30

send a lot of likes, and that

33:32

can crowd out the experience of people

33:34

who don't send a lot of likes.

33:36

And so it's actually beneficial and gets

33:39

more people out on dates to have

33:41

some limit so that everyone's on a

33:43

more even playing field across the app.

33:45

So limits like that really do make

33:48

sense or your turn limits more recently,

33:50

which actually reduce the number of matches

33:52

significantly, but even more significantly, increase the

33:54

match to exchange. rate so that people

33:57

got out at the end of the

33:59

day on more dates. So you can

34:01

only like a certain amount of profiles

34:03

for example? That's the like limit but

34:06

the your turn limit is you can

34:08

only have so many conversations going where

34:10

it's your turn. If someone sends you

34:12

a message and you're not responding we

34:15

then stop you from sending more likes

34:17

or matching with more people and so

34:19

that gets people either to say like

34:21

okay I'm not interested in this person

34:24

I should probably respond to them. And

34:26

that's another limit. It did reduce matches

34:28

because people ended up talking more and

34:30

sending fewer likes and matching less. But

34:33

because people are having more conversations and

34:35

exchanging more phone numbers, it ended up

34:37

getting people out on more dates. And

34:39

so that ended up being like a

34:42

great positive feature. Interestingly, like even during

34:44

the testing of that, we had some

34:46

people being like, oh my God, this

34:48

is amazing. What are you going to

34:51

charge me for this? experience because this

34:53

is what benefits everybody. And we have

34:55

a principle of hinge about we only

34:57

charge for what we can't give away

35:00

for free. Meaning, if it benefits everyone

35:02

and gets more people out on dates,

35:04

then we give it away for free.

35:07

But if it's something that the scarcity

35:09

is the value, like a rose or

35:11

a boost, or if it actually hurts

35:13

the ecosystem to give it to everyone

35:16

for free, then those are the things

35:18

that we put into a subscription tier.

35:20

Now, when you talk about the onboarding

35:22

experience, when it comes to matching, I'm

35:25

assuming that the more you know about

35:27

the user, the better the match you

35:29

can create, the higher likelihood of a

35:31

date. How do you think about the

35:34

data you need from them? I think

35:36

one element is just the, I was

35:38

hearing a lot of rave about the

35:40

voice prompts when you hear somebody in

35:43

voice prompts, there's also an authentic layer

35:45

to it. And I'm curious, is there

35:47

like a set of ultimate data you

35:49

need or... Hey, you know, in this

35:52

age of AI, tell me more about

35:54

you and I'll... be able to filter

35:56

it and try to match the whole

35:58

thing to get aside. I'm thinking about

36:01

multiple use cases today that even a

36:03

very different example than dating, when I

36:05

think about learning, a lot of AI

36:07

tutors today would basically say, hey, what

36:10

are you interested in? And then I'll

36:12

match the questions in a way that

36:14

matches your interests. So you're not working

36:16

on a boring math problem. You're actually

36:19

working on taking something to the moon,

36:21

which you're really interested in. It's a

36:23

way to kind of learn more about

36:25

your personality. personalize the experience more towards

36:28

you and move away from building something

36:30

which is more generic. Yeah. I'm curious

36:32

when you think about data in this

36:34

way is it more about making the

36:37

app more personal and not just about

36:39

the match or even for the match

36:41

like there's fields that you would love

36:43

to get from people but they're not

36:46

sharing today yet. Certainly there's more information

36:48

that we can be getting from people

36:50

that they're not sharing today and there's

36:52

more things that we can be doing

36:55

with the data using AI. So these

36:57

are two really... interesting and very important

36:59

vectors that I think are going to

37:02

really start to transform hinge over the

37:04

next couple years, really, or even over

37:06

the next 12 months. We're completely overhauling

37:08

our recommendation system to be powered by

37:11

AI. So part of that is, as

37:13

you said, picking up on signals we

37:15

weren't picking up on before, factoring in

37:17

all the information your prompts and your

37:20

photos and all this stuff to make

37:22

really thoughtful intelligent matches and make connections

37:24

that we kind of weren't able to

37:26

make you back. And then on top

37:29

of that we can add more information

37:31

that is less structured than what we

37:33

were used to being able to handle

37:35

and actually factored that in in really

37:38

interesting ways. So instead of just getting

37:40

your all the profile information that's mostly

37:42

structured, yet we have three prompts and

37:44

then we have your six photos, you

37:47

know, what's your, like, all that kind

37:49

of stuff. But now we're starting to

37:51

test things like, just talk to us

37:53

a little bit about... who you are

37:56

and like what really matters to you

37:58

and what kind of person you're looking

38:00

for. and being able to take that

38:02

information and shape your recommendations in a

38:05

way that's much more nuanced and thoughtful

38:07

than we were able to do before

38:09

LLLMs and generative AI. So both those

38:11

vectors are ones that were moving down

38:14

very rapidly in parallel. So in many

38:16

ways you'll have your profile which is

38:18

structured, people will see what you decide

38:20

to put out there, but there'll be

38:23

a deeper set of knowledge. about you

38:25

that can be used for your match

38:27

ultimately? Yeah, I think that's a really

38:29

exciting piece of where the future is

38:32

going. Yeah, right now you're, you create

38:34

a profile for yourself and what you

38:36

put on that profile is what other

38:38

people see. There's what other people see.

38:41

There's some fields that you can hide,

38:43

you can choose to hide, but it's

38:45

very kind of like, here's what I'm

38:47

putting out to the world. But now

38:50

being able to sort of speak to

38:52

me up. is a really cool and

38:54

interesting world that we're going to be

38:57

able to be in, to I think

38:59

get a deeper and more nuanced understanding

39:01

of what people really want. I'm assuming

39:03

it's exciting at hit right now, just

39:06

thinking about all the changes you can

39:08

make. I'm curious if you have to

39:10

sometimes ground the team with, hey, here's

39:12

what's not going to change, or here's

39:15

what I'm not willing to bend this

39:17

story eye, or thinking about this different

39:19

different AI perspective. Obviously, the ultimate goal

39:21

is humans interacting with humans. building long

39:24

lasting relationships, but are there conversations when

39:26

you're pushing on like, or the team

39:28

is asking you, like, what do you

39:30

believe will not change? Or how does

39:33

that credit gets created? Yeah, definitely. With

39:35

AI specifically, you mean? Yes. Yeah. Or

39:37

beyond, but curious from an AI perspective

39:39

as well. I think that's why it's

39:42

important. Once again, we have our book

39:44

and we really root in some, you

39:46

know, I think about your mission and

39:48

your values as things that change. Almost

39:51

never. I mean, our mission and our

39:53

values have been like very... to us

39:55

for like a very, very long time.

39:57

Then there's the principles level of things

40:00

which change, I'd say like every few

40:02

years. And then there's your company objectives,

40:04

which change like every year or two.

40:06

And so this kind of like concentric

40:09

circles and like the closer you get

40:11

to the core, the less it should

40:13

change or maybe not change almost at

40:15

all. And some of those values are

40:18

really shaping and showing up in ways,

40:20

especially authenticity when it comes to AI

40:22

that we think. Okay, how do we

40:24

now come up with principles? Because we

40:27

have, yes, those top four level principles

40:29

we talked about, like love the problem

40:31

and keep it simple, tend to trust.

40:33

But even when we build a new,

40:36

like, just a new feature, or build

40:38

a new initiative, we develop principles for

40:40

it. And we've been developing principles around

40:43

AI as well. And one of the

40:45

biggest ones for us is that AI

40:47

really should stand behind us and not

40:49

between us. Meaning, we want you primarily

40:52

interacting with other people. for you or

40:54

generating images for you, right? So one

40:56

thing that we just released, for example,

40:58

is prompt feedback. We don't write prompts

41:01

for you, we don't suggest what to

41:03

say, but we will give you really

41:05

thoughtful feedback on your prompt and how

41:07

you can make it better, trained on

41:10

what we have seen work, for like

41:12

what kinds of prompts actually generate conversations

41:14

and which don't. So we won't let

41:16

you copy and paste and put things

41:19

in there that you didn't write, but

41:21

we will say, hey, you just wrote

41:23

brunch as an answer to my favorite

41:25

Sunday. Where do you like to go?

41:28

What do you like to do? Do

41:30

you like to do? Do you like

41:32

to cook? And then they'll say, what

41:34

do you like to cook? Sometimes you

41:37

just have to pull it out of

41:39

people and help them understand that by

41:41

answering with more depth and vulnerability and

41:43

authenticity. you can help coach them so

41:46

that they get better matches. So that's

41:48

like one example of how we're using

41:50

it to make sure that we really

41:52

preserve authenticity in the experience. and you

41:55

don't feel like you're talking to an

41:57

AI or AIs or talking to AIs

41:59

on your behalf or something like that.

42:01

We like want to preserve the very

42:04

human, you know, messy experience of dating.

42:06

Going back to like love the problem,

42:08

I can think of new problem sets

42:10

to go after. So for example, go

42:13

beyond the app, beyond the match to

42:15

I'm assuming for if you lack confidence

42:17

when you go on a date, you

42:19

could use some coaching. During, before, after,

42:22

coaching is a huge piece of where

42:24

we're going. Two big vectors, basically, that

42:26

I think AI is taking us down.

42:28

The first is personalized matching. So that's

42:31

what we were just talking about before.

42:33

Using the data that we do have

42:35

better, making personalized introductions, gathering more nuanced

42:38

data that we can use to match

42:40

you. That's all great. But no matter

42:42

how well we can match you, some

42:44

people just need. help filling out their

42:47

profile. So prompt feedback is one early,

42:49

very simple example, to much more nuanced

42:51

and thoughtful coaching. And so we have

42:53

a whole team called Hinge Labs, which

42:56

is always doing research on what works,

42:58

what doesn't, how do people succeed on

43:00

Hinge, and in dating generally. And we've

43:02

taken that data and we've published date

43:05

reports to the press. We have a

43:07

section in our app called What Works,

43:09

where you can go see what kind

43:11

of photos to pick. People don't really

43:14

read that stuff. If they do, they

43:16

don't really know how to apply it.

43:18

And what's so exciting about what we're

43:20

going is we can take this whole

43:23

body of knowledge and now apply it

43:25

to you in a very personalized way

43:27

and deliver the right piece of feedback

43:29

in the right moment. And that's like

43:32

another huge vector that I think is

43:34

going to help a lot of people,

43:36

because many people struggle to get that

43:38

first match or have that first date.

43:41

A lot of times it is because

43:43

they just don't know to like fill

43:45

out a more detailed prompt or choose

43:47

better photos. Maybe they're not sending enough

43:50

likes. There's a whole host of things

43:52

that we could be giving them tips

43:54

around that we just today haven't been

43:56

able to. And so I think it's

43:59

going to help a lot. of people

44:01

to be able to receive those just

44:03

little nudges and moments and coaching along

44:05

their journey from filling out their profile

44:08

all the way to what do I

44:10

talk about on this date tonight? And

44:12

that's where you see in a way

44:14

hinge kind of progressing. We're like the

44:17

data is I'll focus on getting the

44:19

date. You know that's like step one

44:21

step two is like actually transform that

44:23

date into a relationship. Yes, definitely. Justin,

44:26

I've learned so much. I really appreciate

44:28

the candor and also kind of the...

44:30

principle way of thinking plus the new

44:33

answers and how you bring this to

44:35

life at hinge and I wanted to

44:37

thank you for all the learning. It's

44:39

wonderful. Yeah, yeah, it's been a really

44:42

good conversation. Thanks for having me and

44:44

thanks for such great questions. Thank you

44:46

so much. This wouldn't be an episode

44:48

of building one without some great takeaways,

44:51

so let's just jump in. First, as

44:53

one might expect, rebuilding a company is

44:55

brutal, but sometimes very necessary. Justin made

44:57

the difficult decision to tear down Hinge

45:00

and start all over again when he

45:02

realized the company drifted from its original

45:04

mission. The early version of Hinge was

45:06

competing with Tinder in a race for

45:09

engagement and that wasn't why he built

45:11

it. This reboot refocused Hinge on helping

45:13

people build meaningful relationships, not just match

45:15

endlessly. This approach did not guarantee the

45:18

success Hinge is seeing today, but it

45:20

was clear that without a reboot. there

45:22

would be no chance of success at

45:24

all. Second, intentionality builds better products and

45:27

better matches. Hinch's onboarding process intentionally filters

45:29

out people who aren't serious about dating,

45:31

losing 20% of signups in the process.

45:33

That's a feature, not a bug. The

45:36

platform also requires users to put in

45:38

more effort, like commenting on a specific

45:40

part of somebody's profile, while even sharing

45:42

something about themselves via voice recording. That

45:45

leads to a more thoughtful interaction and

45:47

high quality matches. Third, find good proxies

45:49

for your true North metrics. This is

45:51

something we covered a bit on the

45:54

show before. It's especially relevant in... Hinge's

45:56

case. Hinge's North Star metric is people

45:58

getting out on dates and then ultimately

46:00

relationships. But all of it happens physically,

46:03

not digitally. It's off the app. It

46:05

can't be measured. Therefore, Hinge looks for

46:07

all sorts of opportunities to infer or

46:09

get feedback from users. Did they exchange

46:12

contact information? They indicate that they actually

46:14

met their match when they were surveyed.

46:16

If someone decides to leave Hinge, they

46:18

ask why? It might not be perfect

46:21

information, but piece together it gives the

46:23

visibility into how well the product is

46:25

doing. Fourth, consider how AI can either

46:28

help or hurt the human experience. For

46:30

example, Hinge wants to help its users

46:32

go out on good dates, but it

46:34

doesn't believe it should be doing the

46:37

work for them. Hinge believes that the

46:39

AI can coach people on how to

46:41

build a better, more thoughtful profile. If

46:43

someone's answer to a prompt is too

46:46

simple, it's not going to generate a

46:48

conversation a conversation. So, instead, AI should

46:50

prompt the user to give a more

46:52

thoughtful, deep response. This is not unlike

46:55

Khan Academy's own approach to AI. The

46:57

goal is to encourage people to do

46:59

the meaningful part of the work, because

47:01

if they just have the answers, they'll

47:04

fail on their own later on. Lastly,

47:06

I love how Hinge operates with a

47:08

clear set of product principles. It's on

47:10

their website, and you can clearly hear

47:13

it from Justin. This principles helped the

47:15

team make tough decisions quickly. and ensure

47:17

the product evolves without losing its core

47:19

purpose. Strong product principles help create a

47:22

stronger company. Want to hear more from

47:24

Justin? Then check out his podcast interview

47:26

on This is Working with Dan Roth,

47:28

LinkedIn's editor and chief. We'll be back

47:31

in two weeks with Rohan Amin, the

47:33

CPO of Chase. Building one is a

47:35

production of LinkedIn News. Our host is

47:37

Homer Cohen, LinkedIn's chief product officer. This

47:40

episode was produced by Max Miller. Our

47:42

associate producer is Rachel Cartm. This episode

47:44

was mixed by John Partham and engineered

47:46

by Asaf Gedrom. and we we

47:49

get additional production

47:51

support support from At LinkedIn

47:53

At LinkedIn News, Sarah senior

47:55

producer. Dave Dave head

47:58

of head of creative operations.

48:00

creative Pope Chappelle is

48:02

director of content

48:04

and audience development. and

48:07

Courtney Coop is head

48:09

of original programming. head

48:11

of Dan Roth is

48:14

the editor -in -chief

48:16

of LinkedIn. If you

48:18

know a product leader we can all learn from,

48:20

leader send us a line at send us a line at .com.