0:00

In five years from now, nobody is

0:02

going to use Concur Dynamics. Nobody

0:04

is going to use Salesforce. These companies

0:06

have lost the way around innovation.

0:08

They don't know anymore how to innovate.

0:10

They cannot attract the people. They

0:12

know how to innovate. These companies

0:14

actually, I would argue, will just die.

0:17

That's what will happen because innovation

0:19

OpenAI. These companies lost their way

0:21

around innovation, and they will just die because

0:23

they are not allowed This is twenty VC,

0:25

the memo, where we saw the success story one

0:28

of the breakout companies of the last decade

0:30

and today. We have a special guest. Just

0:32

yesterday, they announced their rebrand from

0:34

trip actions to Navan. And wow, what

0:36

a show this was? I'm thrilled to welcome Ariel

0:38

Cohen, cofound CEO, Navan, formerly

0:41

tripassions, the number one travel management

0:44

super app used by over eight thousand

0:46

companies. Ariel's Ray over two billion

0:48

dollars for Navan from some of the best, including

0:50

Andresen, Ziv Ventures, LightSpeed,

0:53

Green Elon Deal, and private

0:55

trip Navan, Ariel Cove founded Stream

0:57

ones, which was successfully acquired by

0:59

Jive's software. But before we dive into the

1:01

episode's day, if you lead a team and have

1:03

ever had to buy software or services,

1:05

sell software or services, hire contractors

1:08

or make a deal. Basically, everyone

1:10

let's submit it. You've probably experienced how

1:12

slow and confusing contracting can be.

1:14

It doesn't have to be so hard, but it's twenty

1:17

twenty three. We need contrast to keep

1:19

pace with the speed of modern business,

1:21

and that's where IronPlanet comes in. IronPlanet

1:23

is a leading contract life cycle management

1:26

platform for innovative companies, unless

1:28

users create, understand, and control

1:30

contrast in a walk up usable browser

1:32

tab, making contracting painless and

1:35

fast. Check out their clients. L'Oreal,

1:37

Staples, Mastercard, and other leading

1:39

innovators using Ironclap to save

1:41

money and time for their business, and to

1:43

unlock contracts, data using the

1:45

most powerful AI and workflow capabilities

1:48

on the market. It's the only platform

1:50

flexible enough to handle every type of

1:52

contract workflow, whether a sales agreement,

1:54

an HR agreement, or even click wrap,

1:56

and the only platform that has processed over

1:59

billion contracts. To learn how Ironclad

2:01

can be the key to improving your bottom line

2:03

this year, visit ironclad dot com

2:05

and speak you top quality there with Ironclad,

2:08

Harvard management company is constantly seeking

2:10

out a next generation of great investors and

2:12

entrepreneurs. H m has managed

2:15

Harvard University's endowment for nearly

2:17

fifty years and was one of the first institutional

2:19

investors in venture capital. Their experience

2:21

in long term investment horizon made them

2:23

ideal partners to get world changing ideas

2:26

on a path to viability and success.

2:28

They work as a true partner, providing insightful

2:30

perspectives to how managers succeed. I

2:32

know I personally have had the pleasure of working

2:35

with the HMC team and can say that they are

2:37

truly exceptional partners and savvy

2:39

investors, whether you're launching your first fund

2:41

or your fifth, HMC welcomes

2:43

the opportunity to partner with both developing

2:45

and established managers. Have an idea you

2:47

wanna share with the team? Just send it to venture

2:49

at hmc dot harvard dot

2:52

EDU Clean. Clean. One.

2:56

What you got?

2:58

You have now arrived at your destination.

3:01

Mario, I am so excited for this. It's

3:03

been quite a few years now. So first, thank you

3:05

so much for joining me again. Thank you.

3:08

I'm so happy to beer again. And

3:10

I'm always listening to

3:11

you, interviewing really smart detail.

3:13

It's really interesting to hear you and also

3:15

listen to your Jeff Jordan one and it

3:17

was really fascinating. I was very, very

3:19

kind. I'll pay you later. But we were peeking

3:21

beforehand, and you did This is all

3:24

scheduled. I the first question being all

3:26

scheduled is, like, unusual, but fuck it. You

3:28

said to me, school and

3:29

university. You didn't really go to that

3:31

many latches or lessons. I

3:33

was just intrigued by why? The type

3:36

of person that I

3:36

name, I can learn stuff by myself,

3:39

but I can only learn stuff that really

3:41

interests me. The reality is that I think

3:44

that most of the things that you will learn at

3:46

school and universities, they'll just not be

3:48

interesting. They'll all of things that,

3:50

for example, some case studies that

3:52

you're gonna go through, and I could be

3:54

so much fascinated by that.

3:56

But then I could go all in. For

3:58

me, the way that I can learn, it's

4:01

only it interests me. If it's not, I

4:03

cannot really study it, then that was

4:05

my school life. It was actually a extremely

4:07

average students. If I could either

4:09

have, like, really, really, really good straight

4:12

day, but on the really good side of it,

4:14

or I would totally fail, and it's really

4:16

based on the way that kind of maybe my brain

4:18

works. If it gets my interest, I would geek

4:20

out and learn everything about

4:22

it. If it doesn't, you cannot get me when

4:24

starting to get the interest. Gosh, does

4:26

that impact how you advise your children

4:29

when it comes to how they pick up and

4:31

learn new

4:32

things? One hundred percent. So I have

4:34

three kids. Two of them that they are twins. They

4:36

are boy and a girl. They are fifteen. So they are

4:38

starting to think what would they do next?

4:40

And and so one. One of the things that they've told them,

4:42

hey, you do whatever you want to, but my

4:44

advice to you don't even at age eighteen

4:46

have this pressure to go to college because

4:49

I don't think that what you really want to do with your

4:51

life at age eighteen, and I know that a lot of

4:53

people are doing this keep here. And I like

4:55

traveling. A lot of my conclusions are happening

4:57

while I'm traveling. So my advice

4:59

to them, what you're gonna do with your

5:01

life in school and all of these things,

5:04

it will get figured out, but first

5:06

take the time to really learn to

5:08

really understand what to you. And I think

5:10

that in any case, we are in such interesting

5:13

time in our with the development, and

5:15

I know that you're gonna talk about it, but with the developments

5:18

of machine learning and AI.

5:20

So I've not even ensured that that

5:22

we are learning to not obsolete, and my advice

5:24

to them don't take these two

5:26

seriously. So I think secretly I've always

5:28

wanted to be psychologist, but we mentioned

5:30

children there, and psychologists always give everything

5:32

to your childhood and how it impacted

5:34

you. I do think everyone is a function of their histories.

5:37

What do you think you're running from Mario?

5:39

If you take a step back, I think the one that

5:41

comes in mind always for me is Salesforce

5:43

and SAP, which are not innovating.

5:46

I think the last time that SAP innovated

5:48

something was in the nineties and the last time

5:50

that that Salesforce innovated was in the beginning

5:52

of two thousand. And I'm looking at this and

5:54

besides the fact that I think it's embarrassing

5:57

that companies that are located in middle

5:59

of Silicon Valley have not been innovating

6:02

for years. It's also what I'm running

6:04

before. The one thing that scares

6:06

me is that one day, Navan

6:08

will be that kind of a company that the only

6:10

way that you innovated for M and A

6:12

and for using the OE market dynamics

6:15

and your power in the market to basically

6:17

sell. And this really scares me. Just jumping

6:19

on that I'm said, how do you retain

6:21

innovation, speed, agility

6:24

with scale? And we people are a trip of Navan

6:26

now. have three thousand and eight. How do you

6:28

retain that innovation and agility?

6:30

I think it's about decisions. And Elon

6:32

and I looked at this, Elon is my co founder

6:34

and this was, like, five years ago, and we said

6:37

if we would have started transactions again,

6:39

we wouldn't do it like that. Out of this

6:41

came everything that we did around

6:43

payments and expense management. Two

6:46

years ago, we looked at this again, and

6:48

we said today with machine learning,

6:50

you can achieve much better service

6:53

by servicing people fast. Where you cancel

6:55

your flight, are you going to get your tickets back

6:57

or not? Tell me how to automate

6:59

it. But we've managed to automate it with

7:01

a lot of usage of face smartness

7:04

inside. We are calling it, we have an internal body,

7:06

we call it Cati's IQ. And this

7:08

is distinct that keeps learning. So again,

7:10

we killed what we had. And recreated

7:13

it. So I think you keep this innovation

7:15

if you have the college, first of all, to look

7:17

at the mirror and come and say, okay, what we

7:19

had done up until now was relevant

7:21

back then with our understanding and the

7:24

technology back then, but maybe it's time to

7:26

kill it and change it. And we've been doing it times

7:28

and again here. Actually, my cofounder right

7:30

now is running two projects that

7:32

are making basically a lot of what we've done

7:34

to date to

7:35

obsolete. And I think that's how you maintain innovation.

7:38

I'm writing. When you get to my age, you forget

7:40

And so you write on your hand then you will

7:42

remember. How do you want you next? I think I'm all there

7:44

then you. I'm not sure, but I think I'm way older

7:47

than you, so maybe you

7:48

know. Roger and silver fox, my friend. My

7:50

first question was you mentioned some of the additional

7:52

products you create over time. How do you

7:54

common as a leader and strategist between

7:57

an additional product that you should chase

7:59

and should do versus a

8:01

distraction. It so easy to be distracted.

8:03

What's the difference seen a product you should do in

8:06

a distraction? I think people by

8:08

default,

8:08

oh, afraid. Of a change and

8:11

they will do whatever it takes to

8:13

not make a change. Even here in Silicon

8:15

Valley inside the tech companies And

8:17

even last night at eight PM, there

8:20

was a designer that actually showed me

8:22

yesterday a design of for the new unified

8:24

app. We are actually releasing this out of the

8:26

Navan branding. We are consolidating all

8:29

of our technologies to one super app.

8:31

We all 92BN using it here in a three actions

8:33

or inside Nirvana. Then, this designer

8:35

showed me a really interesting design.

8:37

Something much better and we are two

8:40

weeks before the release. So what

8:42

did you do? And I told the team, hey,

8:44

I cannot answer it. I saw this

8:46

design now. I cannot answer it.

8:48

That's should be our design. And this is

8:50

scary. Changing everything in the last minute

8:52

is scary. But then, here is the thing.

8:55

We've agreed it with the team that will do this change.

8:57

But then for quite some time will not do more

8:59

design changes because now

9:01

I want to focus on something else, other

9:03

innovation in a in a van. Are the two

9:06

big projects that we have here. So

9:08

this is scary because I've said, hey, it's

9:10

all about design. It's all about this thing, but now

9:12

I'm gonna deprioritize again three

9:14

weeks from now and do something else.

9:16

And I think having the ability to

9:19

make these decisions in bringing

9:21

the team with you and not having this fear

9:24

that the team will object to this. I think this

9:26

is in the heart of companies that are

9:28

innovating, but it's so in the heart of companies

9:30

that are not

9:31

innovating, you're sometimes too much afraid

9:33

as a company from the change. How

9:35

do you get over being afraid? Do you

9:37

bring people with you as well? Also, when you're deprioritizing

9:40

something, you don't want those that were

9:42

prioritized on it to feel like

9:44

the young love child. How do you get

9:46

over

9:46

that? First of all, you never give up. And

9:49

I'll give you an interesting anecdote. The COVID

9:51

started and we lost out of

9:53

our revenue, but this decided to continue

9:56

to be in the market in sale, do M

9:58

and A, raise money, but the most

10:00

important thing, innovate, continue

10:02

to innovate. Back then, I've decided that

10:04

the innovation would not necessarily be

10:06

just about Navan, but we're actually gonna

10:09

shift resources, we're gonna shift the designers,

10:11

product managers, evaluate from

10:14

the travel side to the expense

10:16

management side. That was

10:19

a decision that actually the employees

10:21

back then didn't like From a lot of various reasons,

10:24

again, we've talked about being afraid of change.

10:26

So in that process actually lost.

10:28

Most of my productive. Okay. And

10:31

two years afterwards, I started to look at the

10:33

data that it showed how much both are picking

10:35

up, travel and expense management. I

10:37

remember that I told the my co founder,

10:40

sometimes you need to do these decisions

10:42

even if it means that you're gonna lose the team

10:44

in the process because that was the right

10:47

decision. So sometimes you can break

10:49

people around with you. In most of the cases,

10:51

that's what I do and you can do that.

10:53

But you cannot be afraid of

10:55

the consequences of the implications of

10:58

doing what you believe intuitively

11:01

deep inside that is the right thing for

11:03

the

11:03

company. Speak of doing what you believe in the right

11:05

for the OpenAI, you also mentioned killing projects.

11:08

How do you determine which projects to

11:10

kill? First of all, you know, I think that

11:12

I will bring it back to maybe the

11:14

linear startups kind of concepts. I

11:16

think what I really like in this concept

11:18

is that you start something with the hypothesis

11:21

and you start to measure it. And your entire

11:23

Pampers is not to develop the product, but

11:25

to actually examine your hypothesis.

11:28

If you are strict about it, you cannot

11:30

lie to yourself. If you have an hypothesis, you

11:33

are saying, hey, this is what I

11:35

think that this product will do or this feature

11:37

we'll do or this project we'll do it. This is

11:39

how we think that the market will accept

11:42

it. Then you put the right metrics and

11:44

measure men's in place and you keep asking the question,

11:46

is it happening or not? Is it happening or not?

11:48

Then you can do some payload. So it doesn't mean

11:50

that your first implementation will be the right

11:52

one. But if you operate like that,

11:54

it's not that how to bring the team into

11:56

a room and say, we had these

11:57

hypotheses. It doesn't work. Let's move on.

12:00

Let's do something else. I think if you do

12:02

that very capable of. How long do

12:04

you give it? Because there's that kind of you can't

12:06

do it too

12:06

quickly, you need enough data, but then

12:08

too many founders are afraid to kill it and give it

12:10

way too long. Based on how important

12:13

it is for your strategy. If it is

12:15

fundamental for your strategy, you

12:17

define your strategy as a OpenAI. When you

12:19

are saying this thing. I'll give you an example.

12:22

We give rewards for employees

12:24

to save company money. So you come

12:26

to our product, you write a book and

12:28

hotel, it's in Paris see, then

12:30

we are telling you, but if you'll book

12:32

something that is lower than the policy,

12:35

we're gonna give you rewards, but part of the

12:37

Navan rewards program. We are rewarding you

12:39

to use us. We're rewarding you

12:42

to save company money. We are also getting

12:44

your Marriott's loyalty points

12:46

So you get it out, you I'm calling it.

12:48

You three pill deep. So with your credit card

12:50

if even you've bought you did four times.

12:52

So you're happy as an employee. But when we

12:54

started the OpenAI. We had the same

12:56

idea. Nobody would click on the stuff that

12:58

will give you the rewards, and nobody would

13:01

actually redeem these rewards. Now the

13:03

question is, are you pivoting or are

13:05

you doubling down? This feature

13:08

today feature back then it was the entire OpenAI.

13:10

Was so important for our strategy

13:12

that even though the hypothesis was

13:15

in collect, nobody was using it, we're measuring

13:17

it. We've decided to double that. We've

13:19

decided to continue to pivot, to change

13:21

the UI, to change the way that we are describing

13:24

it, to change the algorithm, and to continue

13:26

to double down. And eventually everybody

13:28

started to use it. So sometimes it is

13:30

so important to what you do to your set of

13:32

belief that you double down and eventually it will

13:34

be lot out of cash. Oh, you pivot

13:37

all your being successful. And sometimes, it's

13:39

more clear that you did it, you dried

13:41

it, but it's okay to give it up at home.

13:43

Ariel, you mentioned the time when COVID

13:45

happened your revenues went to zero.

13:47

Take me to that moment. How

13:50

did you respond? That is a oh

13:52

shit

13:52

moment. Yeah. Personally, and

13:55

sometimes your good precipiters may be called,

13:57

I can move on fast. I can come

13:59

and say to myself, the plan that we had

14:02

in that company that we had. Okay?

14:04

It's not relevant right now. And

14:06

so what does it mean for me?

14:08

And I'll get back to, I will mention this

14:10

magic again, and again, because it's really the

14:12

call of Nirvana. For me, March

14:14

twenty twenty, we lost all of our revenue,

14:16

obvious to me that we lost product

14:18

market fit overnight, nobody needs

14:20

the spot updating. So for

14:22

me, it got back to the basic question.

14:25

I do believe that there is this new

14:27

category of business software

14:29

that is designed for people, basically

14:31

creating people centric product.

14:33

That's something that I believe we started the

14:36

company with that belief. And I

14:38

do believe that travel and expense

14:40

is an important part of this.

14:42

And I also think that people will travel

14:44

in two years from now, again March twenty

14:46

twenty. So all of my beliefs are

14:48

still there. This new type of software people

14:51

will travel in the future. So now what do

14:53

you do? Do you kill the company? Do

14:55

you hibernate the OpenAI? Or do you die

14:57

build out? If in Cheslando, probably you

14:59

need to hibernate the company or kill it. But if

15:01

you still have this set of beliefs and

15:03

everybody around you, your investors, the

15:06

big part of your employees, still believing

15:08

it, you carry on, you continue. So you

15:10

make some tough decisions, cut costs.

15:12

See, we did like pretty big layoffs. We've

15:14

cut a lot of other costs, but on the

15:16

same

15:17

time, double down on innovation, double down

15:19

on our go to market. Can I be blunt? How

15:22

did you expand the business? And

15:24

how did you raise a double

15:26

digit valuation

15:27

for at the time, put only a travel

15:30

company. In the middle of COVID.

15:32

It's incredible, but how? It comes down

15:34

to this set of beliefs, so I remember in March

15:36

twenty twenty, the issue not only

15:39

that, obviously, we are going down very

15:41

fast, but at that point, people found that the

15:43

entire economy will go down. But that point,

15:45

there was so much pacemism and

15:47

by came to the team and to the board and

15:49

was saying, okay, we're gonna definitely stabilize.

15:52

No question about it. But we're gonna stay in

15:54

the market and sell travel. And in

15:56

fact, we're gonna expand to expense management

15:58

and sell to any T and E. When we're gonna

16:00

raise money? And of course, the reaction would be,

16:02

but how gonna do that. You're gonna call a CFO

16:04

right now and tell them you you need to use

16:06

the playback shares for travel, and they will tell

16:08

you, well, no traveling right now. What are you talking

16:10

about? But think about it, if you convince

16:13

them, it was not how to convince the travel

16:15

will come back. What's better than

16:17

to do change management from this

16:19

antiquated Comcast and

16:22

GBT, American Express GBT,

16:24

nobody likes it. The only reason that people

16:26

are staying with that is change management. It's

16:28

the best time to do it. Nobody is traveling,

16:30

nobody is expensing, changed out.

16:32

And that was our pitch. And in April,

16:35

it was very hard to do that because nobody

16:37

would even pick up the phone. I think people were

16:39

afraid that, you know, for their lives. But

16:41

May, June, July is the thing

16:43

started to open up. Navan was still not

16:45

happening, but people would sign up

16:47

with us. Our biggest customers would sign

16:50

up with us, things like Netflix, Heineken,

16:52

Adobe signed with us on these

16:54

times where people were not threatening

16:56

because, again, change management, the easy

16:58

thing to do at that time. So then you see change

17:00

management like now is the time. Logos

17:03

convert like amazing logos like you mentioned.

17:05

And then you go to the board and say, hey, look

17:07

at our sales, look at our logos, they're up into

17:09

the

17:09

right. We wanna raise a new round. How did

17:11

that catalyze a new round? Yes. So

17:13

I actually went to and back

17:16

then, could you know, and Neil actually looked

17:18

at this and because the first time I didn't

17:20

know him before, but he looked at this and I basically

17:23

told him that oh, hey, this and we don't

17:25

have revenue. We are burning a lot of

17:27

cash because the company think about a fairly

17:29

scaled OpenAI, but here is our plan.

17:31

If you believe in a unique ability to

17:34

execute, the only thing that

17:36

you will need to bet on here is

17:38

the Navan will come back. So just decide, do you

17:40

think the travel will come back or not? If

17:42

you think the travel will come back, I think we're

17:44

on a really good investment right now because

17:46

obviously it was in better terms than

17:48

what you would have got then if COVID was not

17:50

around us. And that's resonated now. I will

17:53

tell you you've talked probably with charity investors

17:55

twenty eight of them told us that

17:58

Sabah will never come back. I've talked to one of

18:00

the biggest VCs in Silicon

18:02

Valley and a very famous one in on very

18:04

good conversation, but they told me eventually, you

18:07

know what? We actually don't think the tavern will come

18:09

back. And I said, okay. So if tavern will not come

18:11

back, it will be very hard for me to make

18:13

a case for you to invest in

18:15

the OpenAI. But the ones that had the

18:17

college back then too, I think it was only

18:19

one bet. Will travel come back? But

18:21

I wouldn't come back before we will call Andy

18:23

or I don't know of

18:24

kids. I agree. But at nine

18:26

point two billion, as an investor as

18:29

I am, you have to see

18:31

a forty billion dollar company to

18:33

get at me

18:33

excited. I need to leave you in. For so the

18:35

nine point two billion, this is even more interesting

18:37

when he came recently So that came

18:40

actually in the middle of recession after

18:42

all of the collection. So as the

18:44

vaccine came, we actually did a five

18:46

billion dollar ounce in an Aplan to

18:48

pre cove it. But again, it was

18:50

Aplan, but it's only kind of back then,

18:52

really based on what happens if travel would come

18:54

back. Can definitely ask how would MGN0

18:57

current investors invested in nine point

18:59

two billion when the entire

19:02

market is collecting and can

19:04

it be forty billion dollar business. They

19:06

send PayPal spending one trillion

19:08

dollars on travel a year.

19:10

Most of this it is not managed.

19:12

Why is it not managed? Because nobody

19:14

like to use a concare and American

19:17

Express. So your employees are all

19:19

over the place. Half of the employees are booking

19:21

outside. You don't have control on it. You're always

19:23

fascinated by this. Everybody are fighting

19:25

all day long. We think and

19:27

our investors think that we can

19:29

capture the entire market for business

19:31

turbine and the entire market for expense.

19:34

If you take one trillion dollars

19:36

of OpenAI and you think about how much money

19:38

is out there, nine billion dollars

19:40

is lawsuit about the execution of the

19:42

company to date. But if you look forward,

19:44

three years, four years, five years, in five

19:47

years from now, nobody is going to

19:49

use Concur Dynamics. Nobody is going to use

19:51

Salesforce. Nobody is going to use

19:53

software that makes you put

19:55

something inside a ticket, and then

19:57

it goes for a workflow Navan then

19:59

I don't know if something happens and the

20:01

really interesting part of this you need to

20:04

make to force it your employees, your salespeople

20:06

need to use Salesforce, your travelers, need

20:08

to use Comcare. Nobody is going to do this

20:10

thing. People today are asking for something else

20:12

and the technology especially with the

20:15

mobile and AI allows for it to

20:16

happen. I love this as a path to a

20:18

trillion. My question is the other bet may

20:20

be and I'm just like ideating with you. The other bet

20:23

may be is Will Ariel and

20:25

Navan or Trebastions at the time

20:27

be able to make the transition from point

20:29

solution or more of point solution to,

20:31

as you described earlier, a super app.

20:33

What do you think are the biggest challenges in

20:36

moving from point solution to super

20:38

app? Yeah. Anything you should probably ask, we'll do

20:40

your job for a second of joking. You

20:42

should partly ask me why even changing the name

20:44

because pretax is the bread is good and people

20:46

actually know the OpenAI, so

20:49

why even do that? And Yeah. I think

20:51

yeah. I'm This is Marita. Why

20:53

are you changing the name, Mariel? It's funny.

20:55

I was saying it because I think it is related

20:57

to your question, but we wanted

21:00

to actually create this new

21:02

category, to show the way, to lead the

21:04

way and show all of the companies around,

21:06

you have to develop a new kind

21:08

of software, software that is more inviting

21:10

Navan. Just the name, it's about navigation,

21:13

navigating the world, and about AVAT.

21:15

This is really where the name is coming

21:17

there from. It's about us inviting

21:19

all of the employees all over the world to

21:22

Navan, inviting the company

21:24

to achieve all of the company's goals while

21:26

their the employees are participating in

21:28

it. So it's magical. And if you think about

21:30

it in these terms, this also shows the

21:32

way To companies that clause that way, to

21:34

companies like Salesforce, like SAP,

21:37

it shows the way for all of the companies

21:40

around the world that today you

21:42

can create really good software for

21:44

the company and the employees. What are the big

21:46

challenges in moving from a more point

21:48

solution to that Supra First of

21:50

all, it's to bring it all together. Right?

21:52

To think about it from a design perspective,

21:55

you are booking a trip. And today, you're gonna

21:57

book a hotel. Whatever, you're

21:59

gonna do it again and again. So I travel a lot

22:01

to London. Our biggest team actually is

22:03

in of nobody is in London. I'm travelling

22:05

there a lot. And here is the thing. I'm always taking

22:07

the same flight. I'm always staying in the same hotel

22:10

and guess what? I'm always going to the same coffee

22:12

in the morning. I'm visiting the same kind of

22:14

restaurants. So why do I need to book it again

22:16

Navan again and again. It doesn't make any

22:19

sense. I have all of these data. We know that

22:21

our machine learning algorithm, which we've

22:23

implemented six years ago, knows

22:25

that. So why not just asking you,

22:27

rebook the trip? The trip that you are always doing

22:29

just book it, and I think that we are the best,

22:32

and I mean it, the best all in one,

22:34

travel expense and payment solution.

22:37

So we are the best by far. Bring it together,

22:39

L21 app, and just to one app to

22:41

every one

22:41

app. Everybody can slap some things that are not connected

22:44

into one app, but bring it to one experience.

22:46

Experience that create value to both sides

22:48

when you think about Concur and Amex,

22:51

do you think they're just gonna sit back and go, oh,

22:53

fine. Let Nirvana go do

22:55

this. They're eating all lunch, how did they

22:57

respond? In the nineties, SAP

22:59

and Concierge innovated something. It

23:01

was the first acting generation of IT

23:03

software this process and how do

23:05

you get efficiency by running workflow in

23:07

some approvals? And basically, how do you treat

23:10

people like packages? And it's not surprising

23:12

that Kompeo found himself as part of Vessey

23:14

Pete, it's like you're outrating people like packages.

23:16

And probably back then, that was innovation.

23:18

But the world has moved on and these

23:21

companies have lost the way

23:23

around innovation. They don't know anymore

23:25

how to innovate. They cannot attract people

23:28

that know how to innovate. Nobody

23:30

would expect OpenAI to be

23:32

a research project inside SAP. These

23:34

companies actually, I would argue, will

23:36

just die. That's what will happen because

23:38

innovation OpenAI. These companies lost

23:40

their way around innovation, and they will

23:42

just die because are not relevant, we

23:45

joke that we are old until we know a set

23:47

of companies from the '90s. Take Siebel,

23:49

for example, to form the '90s that are not

23:51

gonna lock Why aren't there? Because

23:53

the Benelux came in the beginning of

23:55

twenty twenty two years ago. And by the real life,

23:58

they were saying no software Right? He was lying

24:00

to people. There is software, the audio developers

24:02

inside their Salesforce, and they just installed

24:04

it in a computer that is in a bigger

24:06

that the center, the new one frame data center.

24:08

So he lied, but he got his attention. And

24:11

how do I know that he's died? Because I

24:13

know that we are investing so much in software,

24:15

so much in innovation. So there is software.

24:17

And this software is driving the organization

24:20

today. So I think these companies that

24:22

innovated something in the sebel,

24:25

SAP, Concierge, and these companies

24:27

that innovated something ten years later,

24:29

Salesforce, service now

24:31

in a walk day innovated back then.

24:34

Now, people want software like OpenAI

24:36

AI, like Slack, like Nirvana.

24:39

And by the way, look at stuck totally inside Salesforce,

24:41

it doesn't really survive there. You can

24:43

see by the minute how the software

24:45

is becoming worse and and war with under

24:48

kind of a new type of

24:49

company. I actually really think that they

24:51

ask will they respond or not. I don't think that

24:53

they'll be armed. Today, not just become sales

24:55

machines invest so heavily in

24:57

sales, and then change management and stickiness

25:00

is strong. And that's why they keep

25:02

growing. Look at the

25:03

numbers, they keep growing. Yeah, first

25:05

of all, if I look at the numbers, I think, both Salesforce

25:07

and SAP, actually, specifically now are

25:09

not growing and, in fact, declining. So that's

25:12

specifically now. I think they do

25:14

growth for M and A. That's the only

25:16

power that they have. So it's basically M and A and

25:18

the channel. I'm gonna use my connect

25:20

channels into IT, into enterprise, I

25:22

to continue and push whatever I

25:25

purchased yesterday. For this channel,

25:27

I used the word for and that was

25:29

the play. I did the HP software. That

25:31

was the play. Let's keep buying companies and

25:33

then push them for the channel. Over

25:35

time, they are my massive changes.

25:37

For HP, it was the entire cloud. There are

25:39

massive changes and you are not going to survive

25:42

it. I think the massive change right now again

25:44

powered by AI is that companies

25:47

want to have software that

25:49

solves this problem without they

25:51

need to enforce it, without they need to fight

25:53

with their employees all day long. I think, yeah, I

25:55

respect actually their channel,

25:57

the sales organizations that they've created,

26:00

the ability to do m and a, but eventually,

26:02

innovation, technologies,

26:04

breakthroughs, always waiting. Can they not buy

26:06

their way to survival? If they put a twenty percent

26:08

you can't if they

26:09

put a twenty billion dollars on the table for Nirvana,

26:11

Yeah. We are not full sale, and there is a reason

26:13

for that. And by the way, we've passed these tests

26:16

several times. Other companies could have sent

26:18

me to retirement few times. By just

26:20

the by gas and we didn't sell because

26:22

we think that we have something very special

26:24

that can't change, right, the way that software

26:27

works in in organizations. Navan

26:29

that's why we will not sell. Now can they buy

26:31

their way? Maybe yes, but you actually

26:33

see that the answer is no. Think about it, the

26:35

novel to a number three in Salesforce. One

26:38

of the restored battlefield, one of the best

26:40

innovators in Silicon Valley, couldn't

26:42

stay in Salesforce ever and he thought that he would

26:44

be the CEO of Salesforce, and he couldn't data.

26:46

And again, by the way, I am a Slack customer,

26:49

and the team right now is looking for alternatives.

26:51

We love Slack. But you can see that over

26:53

time, the cell tracks a little bit towards

26:56

the buttons are moving to the wrong places. I

26:58

care a lot about product and design, and I

27:00

think it deteriorates over the time. So can

27:02

they buy their way into it, yes,

27:04

but to a point, can they retain

27:06

the team that made this innovation, developing

27:09

a DSLA is or we cannot retain this

27:11

team. Therefore, they're gonna lose the entire product.

27:13

So what pitfalls do you have to

27:15

avoid to prevent you from becoming the

27:17

next skincare or Amax. You're

27:20

only asking about my unite, ma'am. I think

27:22

there are probably a handful of companies

27:24

that are capable of continuing

27:27

to innovate in their twentieth, thirtieth,

27:30

fortieth anniversary. I think, by the way, look

27:32

at Microsoft, it's so amazing how

27:34

they came out. Now they are like Vivel dot

27:36

about them right now is Google. It's unbelievable.

27:39

I think, but again, there are a handful of companies

27:41

and by the way, leaders. I've mentioned that innovation

27:43

comes with leaders that can pull this off. And

27:46

I hope that we will be able to pull this

27:48

off to continue to innovate even

27:50

in about ten years anniversary and an or

27:52

twenty years anniversary and so on. And

27:54

it's not innovation for the sake of innovation.

27:56

It's innovation that creates value,

27:59

magic, or customers. In for

28:01

the users are using one. And if

28:03

we do that, I'll be very happy. That's actually

28:05

my nightmare that we will not be like that

28:07

and will become an SAP exactly, you

28:09

know, the same way that Salesforce became SAP.

28:12

Salesforce were the opposite of SAP. They

28:14

were like, I think that Benav cannot stand

28:16

them. But over time, they became SAP,

28:18

and I hope it will never in

28:20

American Express. You spoke about kind of

28:22

product innovation driving value that you chose

28:24

to integrate OpenAI into your

28:26

chatbot. Told me about that decision.

28:28

What does it mean for the business? Why did you

28:30

make that decision? And I guess, broad,

28:33

how do you see the future of increasing to

28:35

your company. A year ago, we launched

28:37

our first bot. That

28:39

bot today handels thirty

28:42

percent of our support interactions.

28:44

Basically, behaves like a driving

28:46

agent that will say change your

28:49

flight, will handle a flight installations

28:51

will go and get a refund for Minotel. In

28:53

fact, Navan users, when

28:56

they check out for Minotel, they don't need

28:58

to go to the reception and get the receipt.

29:00

We all for a bot that we have are actually

29:02

calling the hotel, getting the receipt for

29:05

them, automatically itemizing it

29:07

and automatically expensing it. That's

29:09

how much today our technology is

29:11

effectively around AI and machine learning.

29:13

And all of this was not powered by OpenAI.

29:16

OpenAI did a huge leapfgged,

29:18

so they basically leapfgged above

29:20

all of the AI technology and machine technology

29:23

including ours. Right? So getting

29:25

back to the innovation, would you kill what

29:27

you have? Because there was a major

29:29

development in compute in software.

29:32

So we decided to do it. We decided

29:34

to kill what we have to use the infrastructure. The

29:36

bot is the bot in the UI is the UI the

29:38

ability to change a flight comes from our API

29:40

to change a flight, but they're to use their technology.

29:43

And if the results internally, I

29:45

think that this Alva. I think we'll probably

29:48

by the end of the year, we'll replace fifty percent

29:50

of function interactions in the contact

29:52

center. And I think by the end of next

29:55

year, it will place ninety percent of

29:57

it, but here is the cool path. Then there is the

29:59

ten percent. The ten percent that solves the

30:01

really hard problems you all stack somewhere

30:03

and somebody really need to help you out.

30:05

And I think this is where we're gonna actually

30:07

dedicate all of our agents to do this

30:09

stuff that really makes sense. Where we're

30:11

listening right now in February, first version

30:14

of Eva powered by OpenAI

30:16

and it would be

30:17

amazing. Carl, just I'm naive. What makes

30:19

OpenAI so much more effective

30:21

than even your internal tools that you built

30:23

for years and everyone else's

30:25

stack. What makes it so much better? Yeah.

30:27

I if people have a lot of opinions about

30:29

it, I think it comes down to the dataset. There

30:31

was a lot of time including with us that was

30:34

used to train the only

30:36

AI to be smart. Right? We used

30:38

to call this thing a cube. So what is

30:40

Airbus a cube? It's this kind of it's the blame.

30:42

We support somebody and that support was

30:44

successful or no. You can measure it so

30:46

this ad for are they going to repeat

30:48

and calling the call center again, stuff

30:50

like that. So it basically would need to give

30:52

the system to signal while it's

30:55

successful or not was the answer. So this

30:57

training, I think, was the old

30:59

way of doing AI. That's how we operate

31:01

it, and I think that's by the way, how ninety eight percent

31:03

of the companies did AI did it. Think one

31:05

OpenAI AI did is they prove

31:07

that if you put it a huge data

31:10

that. Inside this thing, magic

31:12

will happen. I'm not even sure. I actually don't know

31:14

how this magic happens by the way. But

31:16

I think that magic will happen. And we see it.

31:18

I'll give you an example of a magic. You could have

31:20

gotten to the bot that we had before

31:23

OpenAI a I and asked to book

31:25

the Westin in New York. In whatever

31:27

we'll do at that point, she would ask you,

31:29

what do you want to book? But I just told you the

31:31

West. If you'll go like now to check Ept,

31:34

and ask to book the Westin in New York,

31:36

you will ask you in what dates. Okay?

31:38

So Chegg EPT realized that

31:40

the Westin is a hotel. On both

31:43

an entire IQ that was planned

31:45

for years, Three weeks ago, with

31:47

the old technology, we'd not be able to figure

31:49

out that the Westin is a hotel. It's

31:52

day infused with so much data

31:54

and death by doing GPT three think about

31:56

GPT4 and how much more data is there,

31:58

so they infuse it with so much data

32:00

that now magic OpenAI. And that's

32:02

what we see. In real

32:04

life, right, in the or that that I'm using

32:06

right now, it very soon will be available for all

32:08

of our customers. My question is, when we

32:10

look at public markets, drive to

32:12

cash flow, drive to profitability, economics.

32:15

There's some uncomfortable realities that

32:18

you as a business leader and me as an ambassador

32:20

know. And that cut out lot of costs.

32:22

And I know that you can absolutely deploy those

32:25

costs to ten

32:25

percent, but you don't need half of them.

32:27

Yeah. I think that what we've shown

32:30

actually in the travel industry that

32:32

you can run this business today.

32:34

I'll give you one thing we usually don't share our financials,

32:36

but I will give you one number. I'll go with

32:38

margins today at seventy five

32:41

percent. By the end of the year, probably it will get

32:43

to eighty percent. So I'm talking about the year. This is

32:45

high. Exactly. And this is high, generally.

32:47

I think there are some tech companies that are doing better,

32:50

but it's not a thing in services companies.

32:52

So how do you get these

32:55

gross margins a service oriented

32:57

business. And usually, you don't. By the way, you can

32:59

look at some of our competitors are public, and you

33:01

can look at their reports, you're talking about fifty,

33:03

fifty five percent gross margins compared to

33:05

seventy five today, going to eighty,

33:07

maybe even eighty five, eighty five is not

33:09

model deal, but maybe even eighty five in

33:11

the next two to three years. This is

33:13

highly profitable based if you

33:16

think about it from a unit economics perspective.

33:18

What does it mean? It means that businesses

33:20

like always with innovation will become

33:22

more efficient and the businesses that

33:25

will adopt it will get to see another

33:27

day. And the businesses that will not, will

33:29

vanish, either is this famous American

33:31

Basin, I think blog about software is

33:33

eating the

33:33

world. I think software kind of ate the

33:35

world and the companies that didn't

33:37

embrace it died. And I think I say

33:39

maybe AI is eating the wall then and

33:42

eating software and companies that will

33:44

not join the party will not be around. Do

33:46

you worry about having such a core part

33:48

of business and margin structure, I'm

33:50

sure, over time, to worry about having

33:52

that provider by an external provider. I

33:54

think that eventually the bot

33:56

can do one thing. It can answer

33:58

questions based on all of the data data that

34:00

you have. Eventually, you need to operationalize

34:03

it. You do need to change that flight. You

34:05

do need to book that flight. You do need

34:07

to have the supplier's relationships.

34:10

You do need to have all of these things. I

34:12

think the mode that we've created

34:14

is actually not around that technology. The mode

34:16

is all of the infrastructure here that

34:18

allows us to go globally, to

34:20

service our customers globally, to provide

34:23

one T and E report for all of the

34:25

companies that are using us. That's the mode.

34:27

Now, the alarm mode is one of the cultural

34:30

mode. Do you have the courage to keep

34:32

killing everything that you did because the new technology

34:34

showed up? And connect it to our infrastructure.

34:37

And I think if you do that, whether it's external

34:39

or internal

34:40

level, it's not important. So I look at businesses

34:43

like machines with levers. And when you think

34:45

about eighty percent of gross margins on yours

34:47

today. You have a lot of leverage to

34:49

go down to sixty percent and put twenty

34:52

percent more in marketing brand

34:54

growth flight fucking 92BN the

34:56

boats

34:57

go. Why didn't you do that? So actually,

34:59

it's interesting in a part of Navan

35:01

and the rebranding. We are opening

35:04

our product for everybody. So everybody

35:06

can go and sign up to the product. As

35:08

part of this, we are introducing new

35:10

cost structure for different

35:12

markets. So for example, on the lower

35:14

end of the market, we are free. We are

35:16

completely free. You go in. So I think

35:18

if you're basically talking should I use my

35:21

gross margins to reach more

35:23

people. The answer is yes. That's actually

35:25

what we are doing. We're doing another thing. We

35:27

are actually introducing our own loyal club,

35:30

which means costs to make you more

35:32

loyal to us. So again, how

35:34

can we do all of these things? Because

35:36

we've gained over the last eight years

35:38

because of our backs on technology,

35:41

on machine learning. On AI, we've

35:43

gained a lot of efficiency. When I'm

35:45

walking in an airport, I was at I have to tell

35:47

you though. People at Nirvana hearing

35:49

this story because I cannot stop mentioning this

35:51

story recently. I was stuck on an

35:53

airplane with my family, the

35:55

trip kept getting delayed on the OpenAI

35:58

say it four hours there eventually, and

36:00

we have a tight connection in Houston and

36:02

we have already missed the connection. So what we

36:04

gotta do? And I'm using obviously

36:06

the Navan product, and I'm keep getting notifications

36:09

of exactly what is going at you. I'm going

36:11

to miss your connection, but don't worry,

36:13

we took care of you. By the way, that was infused

36:16

both by our product technology, but

36:18

also by these ten percent that I'm talking

36:20

about, the VIP Robert agents that

36:22

we have. This is the Reed and Mackay acquisition

36:24

that we had. And eventually, I was in

36:26

Houston Stebbings in the immigration line.

36:29

I had everybody from the flight with me. All

36:31

of them are booked on the four thirty flight that

36:33

just got canceled, by the way, I am booked on

36:35

the nine thirty flight. Do you know why? Because

36:37

I used the best solution on

36:39

the planet to manage my travel and this is not

36:41

bad. And I mean it. So now I'm waking up in

36:43

the morning and I'm walking down this immigration

36:45

line and it painful and I mean it and it's

36:48

really not just money for me. It's painful

36:50

to think that if all of these people would have

36:52

used Navan, they would have actually need

36:54

to wait during the Houston airport until four

36:56

thirty to the flight that actually got cancelled. First

36:58

of all, they will know that the flight got cancelled. And

37:00

two, the system will automatically

37:03

book them on the right flight. So it's painful

37:05

for me that people are not using us and

37:07

still using something that is learning

37:09

dense software can can and something that

37:11

is really bad service like

37:13

AmEx, GPT. And to your point, I

37:15

want to go to market, fasten, open it

37:17

to everybody and remove any type of burial

37:19

for entry, and we are going to use our economics

37:21

for that. You should give every employee

37:24

a hoodie with a QR code

37:25

they should wear to airpool so you can download

37:27

it. So when you're walking past lines, you're like, hey,

37:30

download it here. And this is amazing idea

37:32

and maybe I should have you joining our marketing

37:34

team. Like, with this

37:35

idea, I'm actually going to talk about this with

37:37

a team later today. It's a really good idea.

37:40

You're not the first person still about marketing.

37:43

If my question to you is, I'm fascinated.

37:45

You set up marketing the loyalty scheme

37:47

that is your

37:47

own. How do you define a

37:49

retained user? We have very

37:51

loyal customers and install

37:54

base, but I want them to feel great

37:56

that they didn't compromise if they

37:58

booked through their corporate, recommended

38:01

solution or enforced solution. In fact, I

38:03

don't want this thing to be enforced. So

38:05

we are making sure that they will always get

38:07

their loyal loyalty from the hotels

38:09

chains from the airlines, but also will

38:11

get their benefits if they save company

38:13

money, but also from us and that getting

38:15

more of their melting into us into

38:17

November. But when you look at Facebook, they said

38:19

when you get five friends, then you're

38:21

like a retained customer. I think it's three

38:23

orders on Go

38:24

Puffs. Your retained customer. What's

38:26

your understanding of what a journey

38:29

main customer is? Because of the expense

38:31

side, we actually know if

38:33

you booked outside. So we know if you booked

38:36

outside. If you know that you're

38:38

getting one hundred percent, what you are

38:40

calling inside a payment of the

38:42

company's budget, we know that we

38:44

are retaining the employees and the customers

38:46

time and again, not just as a big customer

38:49

that maybe use us fifty percent of the time,

38:51

but retaining them all the time. So we are

38:53

comparing this data all the time and

38:55

we have a really good idea of who is a

38:57

user and also who decided to not use

38:59

us. If you're not using us, we're getting

39:01

curious about it. Why? Then we'll contact

39:03

you. We'll call you. This is something that we started when

39:05

we started the company and all the way

39:07

till now. And based on the feedback we're

39:09

introducing stuff the loyalty club that they've

39:11

mentioned. What percent of customers have

39:13

a hundred percent attainment? It's I

39:15

think that it goes by segment

39:18

more of the higher end of the market, what we

39:20

are calling named enterprise, and

39:22

even bid market will get one hundred percent

39:24

attainment. As you start to go down market.

39:27

You are getting more towards the eighty percent

39:29

attainment. Even that number,

39:31

if you go and compare it to the industry

39:33

standards, that's compared to fifty to

39:35

forty to fifty percent

39:36

attainment. So this is significantly more

39:38

than what everybody else are getting out there. Yeah.

39:40

I've had so much fun with this. I will do a

39:42

quick fire round. So I say a short statement

39:44

arrow, then you give me your immediate

39:46

thoughts. Does that sound okay? Let's do it.

39:48

Okay, my friend. What does six Sasmeen tea.

39:50

I would say making my kids proud that that's

39:52

the most important thing for

39:54

me. The day will be part of me. I love that. What has

39:56

been a great father, the mean I think it's

39:58

everything. I'm spending a lot of time

40:00

with my kids although I'm running

40:01

a company that think that I like to

40:03

do the most is to spend time with my kids. It

40:05

means everything to me.

40:06

What opportunity have you not taken

40:08

with refactions now, Navan, that

40:10

you wish you had have taken? I think that had

40:13

I known how things aren't available

40:15

will happen after COVID, I would have

40:17

invested even more and be more aggressive

40:20

on gaining more share during COVID side

40:22

of

40:22

it. We more than tripled the business over Kobi that

40:24

would have gotten much more aggressive if I had

40:26

this side side of when Kobi do land and

40:28

in what kind of skill? What's the most pain

40:30

full lesson that you've been

40:31

through. But you're also pleased to have been through it

40:34

for the learnings. think it's really

40:36

scaling up the company be

40:38

fall COVID. We are going fast

40:40

by let's say year over year, very hyped.

40:43

And I think we allowed some people

40:45

to get into the company that are usually

40:47

joining the hype, and I've learned it the hard

40:49

way. Will never happen again. I'm happy that

40:51

they learned it, and I'm happy that also I did

40:54

the opportunity to reset it during

40:55

COVID. What's a single biggest tip you give

40:57

me on hiring. Hiring, first of

40:59

all, embrace the fact that you're gonna make mistakes

41:01

and just do the collection. Fast.

41:04

Where is AmEx in ten

41:05

years? Is this win and take school?

41:07

First of all, it's AmEx Ghibity, so that's the travel

41:09

side, and I don't think that they're gonna be around. I

41:11

think it is a winner takes all. Even if

41:13

it is not, the other one will definitely not

41:16

be AmEx. It would be a tech company

41:18

that is hypo servicing the

41:19

company. What

41:20

would you most like to change about the world of adventure?

41:23

Think there is something powerful here which is the echo

41:25

kind of chamber, but it's also what kind

41:27

of creates a lot of frustration. I think

41:29

that sometimes people need to open their minds

41:31

outside of silicon

41:32

valley. Final one, my friend. Twenty

41:35

twenty eight. Okay. We're gonna sit down

41:37

again. I'll be a silver fox

41:38

then. But twenty twenty eight. Well,

41:41

we'll be then. Pain that picture for me. I

41:43

think everything around us will be more

41:45

efficient because of AI and

41:48

maybe the opposite of what we think

41:50

all of us will be smarter because of AI.

41:52

Remember that we had this calculus computer

41:54

when we were kids and they're doing mathematics, I

41:56

think this is the new type of that

41:58

thing. It will make all of us smarter. Aero, my

42:00

friend. I've loved doing this. Thank you so much

42:02

for joining me, and this has been so much fun.

42:04

Thank you so

42:05

much. I really enjoyed it.

42:06

The time really, you know, flew by ear. It was

42:08

amazing. Thank you. I

42:11

mean, my word, Ariel did not hold back their

42:14

own predictions for Salesforce, 92BN or

42:16

Kinko. I absolutely love that. If you wanna

42:18

see more from us, of course, you can. On twenty

42:20

v c dot com or searching for twenty

42:22

v c. In YouTube, love to

42:24

see you there. But before we leave you today, if

42:27

you leave a team and have ever had to buy

42:29

software or services, sell software

42:31

or services, hire contractors or

42:33

make a deal. Basically, everyone, let's

42:35

submit it. You've probably experienced how slow

42:37

and confusing contracting can it

42:39

doesn't have to be so hard, though. It's twenty

42:41

twenty three. We need contracts to keep

42:43

pace with the speed of modern business, and

42:45

that's where IronPlanet comes in. IronPlanet

42:48

is a leading contract life cycle management

42:50

platform for innovative companies, it lets

42:52

users create, understanding, control

42:54

contract in a walk up usable browser

42:57

tab, making contracting painless and

42:59

fast. Check out their clients. L'Oreal,

43:02

Staples, Mastercard and other leading

43:04

innovators using Ironclap to save

43:06

money and time for their business and to

43:08

unlock contracts, data using the

43:10

most powerful AI and workflow k birties

43:12

on the market, it's the only platform

43:14

flexible enough to handle every type of

43:17

contract workflow, whether a sales agreement,

43:19

an HR agreement, or even click

43:21

and the only platform that has processed over

43:23

a billion contracts. To learn how Ironclad

43:26

can be the key to improving your bottom line this

43:28

year, visit ironclad dot dot com.

43:30

And speaking of top quality there with Ironclad,

43:33

Harvard management company is constantly seeing

43:35

taking out a next generation of great investors

43:37

and entrepreneurs. HMC has managed

43:39

Harvard University's