0:00

We rarely get a peek into what it's

0:02

like to work at TikTok. What are some

0:04

core principles or values or just how TikTok

0:06

operates? The number one thing is context no

0:08

control. That's why we're always encouraging people to

0:10

sink themselves as a business owner. You give

0:12

them all the information they need and then

0:14

let them just do things without specific instructions.

0:16

How do you actually solve the puzzle by

0:18

connecting all the dots together? Just like how

0:20

I see some of my friends, their kids

0:22

playing Legos. If you don't really see the

0:24

full picture, you won't be able to make

0:26

the Lego as one thing at the end

0:28

of the day. You have to see the

0:30

other pieces. What are also important cultural values

0:32

of TikTok, of how TikTok operates that everyone

0:34

always has in mind when they're building? We

0:36

always have this mentality. We are a startup.

0:38

We are a young company. We're always hungry

0:40

for growth. And a very reactive way is

0:42

like, how can I rent my second half

0:44

of my marathon faster than the first half?

0:49

Today, my guest is Ray Tsao.

0:51

Ray is the global head of

0:54

monetization product strategy and operations at

0:56

TikTok, where he's been for over

0:58

four years. Prior to TikTok, Ray

1:00

spent six years at Google, helping

1:02

scale Google Shopping globally. TikTok

1:05

is interesting for two big reasons. One,

1:08

it's one of the most successful businesses

1:10

in history, last valued at over $80

1:12

billion. And its parent

1:14

company is the most valuable private company in the

1:16

world, last valued at over $200 billion. Two,

1:20

TikTok is quickly becoming one of

1:22

the biggest advertising platforms alongside Meta

1:24

and Google, and generated nearly $10

1:26

billion in advertising revenue just a

1:28

couple years ago. So for

1:30

both these reasons, TikTok is a really

1:32

interesting business and team to learn from.

1:35

And I've seen very few podcasts and even

1:37

media get a peek inside how TikTok operates.

1:39

In our conversation, we discuss TikTok's culture,

1:41

their core principles and values, how they

1:43

hire, how they move so fast, their

1:46

emphasis on working hard, how they do

1:48

OKRs and planning. We also get into

1:50

how to succeed on TikTok's ad network.

1:52

Why you want to be testing at

1:54

least 10 videos a week, how

1:56

it's different from running ads on Instagram, how

1:58

to make content that does well. well on

2:00

TikTok, and so much more. This

2:02

episode has a lot of interesting lessons and

2:04

insights. Obviously, TikTok is at the center of

2:06

a lot of debate globally. Some people love

2:08

it, some people hate it, but no matter

2:10

your opinion of TikTok, there's a lot that

2:12

we can learn from their success. If you

2:14

enjoy this podcast, don't forget to subscribe and

2:17

follow the podcast on your favorite podcasting app

2:19

or YouTube. It's the best way

2:21

to avoid missing future episodes and it helps

2:23

the podcast tremendously. With that, I bring you

2:25

Ray Tsao after a short word from our

2:28

sponsors. This episode

2:30

is brought to you by WorkOS. If

2:33

you're building a SaaS app, at

2:35

some point, your customers will start

2:37

asking for enterprise features like SAML

2:39

authentication and SCIM provisioning. That's where

2:41

WorkOS comes in, making it fast

2:43

and painless to add enterprise features

2:45

to your app. Their APIs

2:47

are easy to understand so that you

2:49

can ship quickly and get back to

2:52

building other features. And hundreds of other

2:54

companies are already powered by WorkOS, including

2:56

ones you probably know, like Vercel, Webflow,

2:58

and Loom. WorkOS also

3:00

recently launched Authkit, a complete

3:03

authentication and user management service.

3:05

It's essentially a modern alternative to Auth0, but

3:08

with better pricing and more flexible APIs. Authkit's

3:12

design is stunning out of the box

3:14

and you can also fully customize it

3:16

to fit your app's brand. It's an

3:18

effortless experience from your first user all

3:20

the way to your largest enterprise consumer.

3:22

Best of all, Authkit is free for

3:24

any developer up to 1 million users.

3:27

Check it out at workos.com

3:29

slash Lenny to learn

3:32

more. That's workos.com/Lenny. This

3:35

episode is brought to you by EPPO. EPPO

3:38

is a next generation A-B testing

3:40

and feature management platform built by

3:42

alums of Airbnb and Snowflake for

3:44

modern growth teams. Companies like

3:47

Twitch, Miro, ClickUp, and DraftKings rely

3:49

on EPPO to power their experiments.

3:52

Experimentation is increasingly essential for driving

3:54

growth and for understanding the performance

3:56

of new features. And EPPO helps

3:58

you increase experience. experimentation velocity while

4:01

unlocking rigorous deep analysis in a way

4:03

that no other commercial tool does. When

4:06

I was at Airbnb, one of the

4:08

things that I loved most was our

4:10

experimentation platform, where I could set up

4:12

experiments easily, troubleshoot issues, and analyze performance

4:15

all on my own. EPPO does all

4:17

that and more with advanced statistical methods

4:19

that can help you shape weak top

4:21

experiment times, an accessible UI for diving

4:23

deeper into performance, and out-of-the-box reporting that

4:26

helps you avoid annoying prolonged analytic cycles.

4:29

EPPO also makes it easy for you to

4:31

share experiment insights with your team,

4:33

starting new ideas for the A-B

4:35

testing flywheel. EPPO powers experimentation across

4:38

every use case, including product, growth,

4:40

machine learning, monetization, and email marketing.

4:43

Check out EPPO at

4:45

getepo.com/Lenny and 10x your

4:48

experiment velocity. That's geteppo.com

4:51

slash Lenny. Hey,

4:56

thank you so much for being here and

4:58

welcome to the podcast. Thank

5:00

you Lenny for having me. It's a pleasure. It's

5:02

my pleasure. I am really excited to have you here

5:04

because it feels like we rarely get a

5:06

peek into what it's like to

5:09

market TikTok, Hey, TikTok builds product

5:11

and operates. Also how to

5:13

be successful in TikTok as a business, as an

5:15

advertiser. So I have all these kinds of questions

5:17

for you and so I'm really happy to be

5:19

chatting. I want to just start with

5:22

a little bit about your time before TikTok, which

5:24

was at Google and kind of comparing that to

5:26

TikTok. So you're at Google for six years, I

5:28

believe. Now you're at TikTok. I'm

5:30

curious what stood out to you

5:32

about the cultural differences between how

5:34

Google operates and TikTok operates. Three

5:37

major things I would say. Number one is

5:40

really how these two companies are

5:43

thinking about innovation. Right,

5:46

so I think Google has a very

5:49

strong philosophy of we're

5:51

an engineering lab and there's a lot

5:53

of technology driven and a

5:55

lot of pieces. They are not necessarily always

5:58

trying to I would

6:01

say cope with the market even. However,

6:03

I think at TikTok, I think besides

6:06

the technology part, we do have a

6:08

very keen, I would

6:10

say, appetite to really understand what

6:12

the market really wants and

6:15

also how can we really service our clients

6:17

in a better way. And the clients here

6:19

is not necessarily only for advertisers, including

6:21

our user and also creator altogether.

6:25

So that's one of the things I

6:27

think is very different in terms of TikTok

6:29

way of work is very customer centric in

6:31

a way. And again, the customer here is not

6:34

necessarily only for the business partner, but also for

6:36

our regular user and the creators on the platform.

6:39

And the second one is really

6:41

thinking about how we

6:43

take approach on product development.

6:47

So a lot of times that we

6:50

take a very rigid approach in terms

6:52

of product development. And oftentimes

6:54

you see us that experimenting a lot

6:56

of different things at all the

6:59

same And also we do

7:01

have a lot of engineering and

7:03

ladder project in the back end

7:05

to really understand how can we

7:07

optimize better for the platform. So a

7:09

lot of times these are the things that

7:11

I think TikTok is doing really, really well.

7:14

The last piece I have to say is

7:16

approach for global prioritization. A

7:20

lot of times that you see a

7:22

US-born company go global and

7:24

oftentimes still they are really

7:26

rooted with the US market. There's

7:29

nothing wrong with it to be honest because this is the

7:31

biggest market for them. As

7:33

a far east-born company, I

7:36

think a lot of times that we can take

7:38

approach with truly how do we think about globalization.

7:42

And for example, we launched a lot

7:44

of product not necessarily first in

7:46

North America. We launched in

7:48

Southeast Asia, for example, for our

7:50

shopping really like very big initiative

7:53

internally for shopping. And

7:55

we launched our Creator Fund

7:58

here in North America. We

8:01

launched our gaming approach, really like

8:03

service our EUI gaming advertisers, really

8:06

really strong over there. So there

8:08

are a lot of different approach in

8:10

terms of how do we prioritize our

8:13

go-to-market and also product development. So

8:15

that's the part I feel like we're very unique

8:17

in the market or unique

8:19

to some of the tech company

8:22

born in the US. It

8:24

reminds me there's this piece by this smart

8:27

guy Eugene Way who wrote a few things about

8:29

TikTok over the years and just like why it's

8:31

been so successful. And one of his

8:33

really big point is that TikTok can

8:35

work really well in other markets because it's

8:37

basically you don't need to know a ton

8:39

about the market because it's this algorithm that

8:41

figures out what people in each market want.

8:43

Is there anything along those lines you've seen that

8:45

just has been really fundamental to it working

8:49

so well in many different markets? The

8:51

algorithm is definitely helping right because it is

8:53

basically the machine is doing a lot of

8:55

heavy lifting. That's actually I

8:58

think across the board on the technology

9:00

company today. The difference is

9:02

actually how much you are willing to take the

9:04

heavy lifting over there in the market. By

9:07

that I mean really like sending your

9:09

troops into the market, hiring

9:11

your local talent, understanding

9:13

the culture and really understanding

9:15

the behavior from those users. I understand like

9:17

the machine can do things but also

9:20

at the same time that you actually get

9:22

local talent to fine-tune the machine. So

9:25

there are a lot of conversation about how I would

9:28

say technology is able to change our life but

9:30

I do think that at the end

9:32

of the day I do believe technology is a tool. So

9:35

if we do have a ambition to

9:37

go global you have to do one more

9:39

thing is actually take your step

9:41

into global rather than having the machine

9:43

to have a lifting you have to really understand

9:45

in local culture. I had a fun

9:47

background for my first job

9:50

is to really doing like go-to-market research

9:52

in the Southeast Asia area. I think

9:55

there was only one thing open

9:58

my eyes after a year and a half in these

10:00

career paths is different markets

10:02

have totally different I would

10:04

say culture and these

10:06

market behaviors are actually coming out of

10:09

this culture. One

10:11

of the fun example I always been

10:13

using was I was

10:15

doing market research for one of the suppliers

10:17

for Turner and also these Inca cartridges for

10:21

talent as a like a

10:23

go-to market research. One

10:25

of the things is always concerned to

10:27

my at that point the

10:30

client was they cannot

10:32

figure out why their premium

10:35

product cannot sell

10:37

in Thailand and then we

10:40

just figure out like you know because the

10:42

quality of their printing

10:44

machine and also their Inca cartridges are

10:46

premium and the quality of

10:48

the paper and everything is very good. But

10:51

when you actually do talk to those consumers

10:53

in those market the answer

10:55

is very eye-opening. They literally told me at

10:57

that time is I don't care. I

11:00

don't care if your Inca cartridges or

11:02

your printer is at the premium quality.

11:04

Maybe the printer I can use but

11:06

I can use compatible Inca cartridges or

11:08

Turner for that because my

11:11

consumer won't care about

11:13

your printing quality or

11:16

the majority of my consumer won't care. So

11:18

in that case you should

11:20

not necessarily worry about if you are

11:23

a premium product. It's actually more about

11:25

how long, how durable, how reliable you're

11:27

able to print things and people can

11:29

read. So I think these

11:31

are the insights I think a lot of times you

11:33

will be neglected from

11:36

some of the clients or you know the

11:38

manufacturers or even the owner of the business

11:41

because they think that you know we want

11:43

to serve this segmentation. However,

11:45

this segmentation is that big in

11:47

this area. So that's why

11:50

the culture is really the key part

11:52

from the market. If you don't understand the culture

11:55

you won't be able to understand the behavior over

11:58

there. It's more about. that I

12:00

think when we say about

12:02

globalization or take the product go-to-market

12:04

in a global scheme or

12:07

even build it apart, you have to get

12:10

your hands dirty and to really understand the

12:12

local culture so that you can understand local

12:14

behavior. I love that

12:16

advice. The way you described it, which I

12:18

love also, is that you kind of have to

12:20

fine-tune the algorithm and the product to work in

12:22

different cultures. Is there an example of how that

12:24

was done with TikTok, like a tweak that

12:26

had to be made or some kind of fine-tuning

12:29

that happened for it to work in a

12:31

different market? Yeah, I think we

12:33

did a lot of fine-tuning on our

12:36

user product side to really think about content.

12:39

That's the number one thing that's going to

12:42

be super different coming from each of the market

12:44

and also from each of the culture. For

12:46

example, in Japan, how

12:48

do you actually get more content that relevant

12:51

for the culture? A lot of

12:53

people may think, okay, are you guys only doing

12:55

dancing or doing singing for

12:58

Japan? The answer is not. It

13:00

is actually more food on

13:03

the TikTok side. Like how

13:05

do you actually introducing new

13:07

food restaurant or new recipes?

13:09

Also, sometimes that you're introducing

13:11

a new technology, I would

13:13

say 3C product, like a

13:15

consumer electronics product over there. These

13:18

other content get really popular sometimes in

13:21

Southeast Asia or even Japan area versus

13:24

in the US, as everybody knows that we're

13:26

starting from really lip syncing

13:29

at a very early stage. But

13:31

now really we're expanding to shopping

13:33

behaviors and also a lot of

13:35

people using us as a main

13:37

platform to acquire new discovery for

13:40

the product. These are the things

13:43

I think different market definitely

13:45

deserves and demand

13:47

different kind of treatment. And

13:49

if you are able to do this a lot, you're

13:51

able to find success over there. That's

13:54

really interesting because you could think if there's this algorithm

13:56

that figures everything out for you, but I think what

13:58

you're pointing out is you have to seed it with the right use

14:01

cases that that culture is most excited

14:03

about. Another good example will be

14:05

creative right so it's a very good example

14:07

how human can work with technology together. We

14:10

have a ton of creatives and we have a ton

14:12

of a content right so of course

14:14

we use machine to label those content using

14:17

metadata to analyzing those content. However

14:20

a lot of times you can find that you know

14:22

when we really thinking about how creative

14:24

can help advertisers humans

14:26

actually make a more interesting or

14:28

more I would say influencing

14:30

decisions over there. For some of

14:32

the verticals we can say that oh you

14:35

know what maybe we can try a

14:38

coupon image with a

14:41

new product like a sticker

14:43

on the top. This maybe actually work

14:46

better compared to some of the

14:48

price promotion even. So a lot

14:50

of things it is really depends on how

14:53

do you actually interpreting the numbers and bring

14:55

in the data points but also at the

14:57

same time your business cocoon is going to

14:59

be very important here to make a judgmental

15:01

call for some of the situation like that.

15:03

I think we still rely a lot on

15:05

both machine and also our own experts

15:08

to analyzing those trends and give

15:10

the recommendations. Awesome okay so

15:12

there's a few threads I'm gonna follow later you talked about

15:14

the product development process so I'm gonna I want to spend

15:16

time there. Also about how to be

15:18

successful in tik-tok both as a creator also as

15:20

a business I'm excited to hear your advice there

15:23

but I want to spend a little more time first on

15:25

just what it's like to work within tik-tok and the culture

15:27

of tik-tok. What are some core

15:30

principles or values or just like

15:32

how tik-tok operates if you

15:34

have to identify like here's the ways that we all

15:36

think about what we want to do right and the

15:39

most important to your day-to-day work what

15:41

words and concepts come to mind. The

15:44

number one thing reasoning with you really really well

15:46

with me is context no control.

15:49

Well fun times when we looking around companies

15:52

different sizes. When

15:54

you we're looking at how to collaborating

15:57

often times we see the behavior that a

15:59

lot of people just working on a smaller piece

16:02

based off their job description. So

16:04

hey, you're working on Go to Market, and

16:07

you're working on Data Analytics, and you're working

16:09

on this book of business, e-commerce,

16:11

and you're working on auto industry,

16:13

for example. And a

16:15

lot of times that this human-made

16:18

silos is actually

16:20

Sloan's day style. Because

16:23

humans are not, or our

16:25

talent, they're not supposed to be

16:27

categorized into different baskets. They may have their

16:29

own majority responsibility, for

16:31

sure. But we don't want to cap

16:34

them into this kind of a box

16:36

we created. That's the reason why we're

16:39

always encouraging people to think out of the box and

16:41

think more, and think themselves

16:43

as a business owner, rather than

16:46

a piece of machine, like keep

16:48

the machine running. Oftentimes, I

16:50

would say, contact no control. That means you

16:53

actually can go above and beyond to really

16:55

think about your whole business

16:58

problem as your own problem.

17:00

And your piece is maybe one part of it to solve

17:02

the puzzle. But how do you actually solve the

17:05

puzzle by connecting it to all the dots we get? We're

17:07

encouraging all the people to think alike that way. And

17:10

by that, I think we kind of

17:12

mentally break out those walls. So encouraging

17:14

our team members to do a little

17:16

bit more thinking is very important. It's

17:18

a little bit more thinking, because the

17:20

think part is very important. And

17:23

then, now, in terms of getting

17:26

things into behavior or changes or getting

17:28

to action, then you need

17:30

to really collaborate with other teams, because we

17:32

don't want to necessarily create, hey,

17:35

you're on other people's working group

17:37

now. You're actually stepping on other people's toes

17:39

now. It is not the situation we're trying

17:41

to encourage. But what's encouraging more is

17:43

contact no control. Think more about how you

17:45

can change. And then we

17:48

do really actually take some actions,

17:50

reactive. You reach out to who's supposed

17:52

to be the owner of that, and then have

17:54

a discussion. So then

17:58

you're able to connect all the dots together. So

18:00

that's one thing I think is very unique to

18:02

our culture. I think it's very, very

18:04

important for us to continue to grow at

18:07

this speed because everybody have a full

18:09

visibility towards our full ownership to

18:12

their mindset, how they can contribute. And

18:16

the key there is context implying you give them

18:18

all the information they need and then let

18:20

them just do things

18:22

without giving them specific instructions. Hey,

18:25

I need you to hit this goal, work on this project,

18:27

launch this thing. Here's what we know. Yeah.

18:30

Things you think are best roughly. I know it's not just

18:32

like anyone does anything, but I imagine that's

18:34

kind of the implication there. Yeah,

18:36

I think it's context no control

18:38

plus proactive thinking and reactive doing.

18:41

Right? So you have to do more proactive

18:43

thinking with these contexts. Now, reactive doing

18:45

means that you need to collaborate. But

18:48

when everybody has this kind of mindset, the

18:50

collaboration should be very smooth

18:52

because people have the context altogether.

18:55

I think the part that I see

18:57

other, maybe some of the other

19:00

companies facing challenges is actually there's too many

19:02

IOs in between. And you

19:04

have people that just protecting their own thing and

19:06

working their own thing and then undelivering. But

19:09

just like how I see some of my

19:11

friend, their kids playing Legos, if

19:13

you don't really see the full picture, you won't be

19:15

able to make the Lego as one thing at the end of

19:17

the day. You have to see the other pieces. So

19:20

I think that's the part I think is

19:22

really powerful and reasoning really, really well. When

19:25

we're really thinking about product development and also product go

19:28

to market. Right? So it's a

19:30

pretty full cycle. People have to see

19:33

this and then they have the context. I

19:35

love this. And this has come up actually a few times recently

19:37

when I was talking to the

19:39

CTO of Netflix and also OpenAI. They're

19:42

very similar in culture where it's give

19:45

people a lot of autonomy and freedom,

19:47

and not a lot of do this, do this, do this.

19:49

The key there is to hire very high

19:52

quality people and very high caliber people because

19:55

if not, then things won't work out too great.

19:58

Is there anything along those lines that you can share? just like,

20:00

yeah, the kinds of people you end up hiring and how you

20:03

hire people that can work well in that environment.

20:05

I agree with you, right? So the caliber of

20:09

these people is actually pretty important

20:11

to support the structure I just

20:13

talked about. And oftentimes

20:15

I can see some people that with

20:17

the quality of, you know, always

20:19

curious. Curiosity is

20:22

a very important quality when

20:24

I'm actually talking to my interviewers because

20:26

I wanna see that they are naturally

20:29

curious to new things. They

20:31

wanna learn more about the new things

20:33

and don't really get stuck with their

20:35

own things. That's one thing. And

20:38

the other thing is the discipline because

20:41

like I said, it is actually a double

20:43

sword, a double-edged sword in this case, right?

20:46

So it could potentially introducing

20:49

some of the chaotic situation in

20:51

a company because everybody's thinking everything. The

20:54

discipline here is actually how you are

20:56

really following the guidance on reactive doing,

20:59

be always thinking about how to collaborate. And

21:02

the discipline here and also the rigorous approach

21:04

here is also gonna be very important. One

21:07

of the good example that is the ability to prioritize

21:10

because I don't believe one thing is everybody can

21:12

do everything. You have to do

21:15

a prioritize properly so that you're able to

21:17

push the right agenda. So I

21:19

think that's more of a kind of

21:22

a quality of the people

21:24

we're looking for is it is hard.

21:26

Don't get me wrong. It's

21:28

really hard to say that we can find everyone

21:30

like that, but we would love to believe that

21:33

we can train our employees like that so that

21:35

they're able to even do better in their longer

21:37

term career. Essentially what you look

21:39

for when you're hiring people is making

21:42

sure they're always curious, they have high

21:44

discipline and that they prioritize well. Coming

21:47

back to the cultural pieces of

21:50

TikTok. So the main one

21:52

you've shared so far is this idea of context, not

21:54

control. What else

21:56

are important cultural values of TikTok

21:58

and how TikTok operates at every... and always has

22:00

in mind when they're building and meetings,

22:02

making decisions. Yeah, another internal

22:05

thing that we always say is always

22:07

day one. We want to

22:09

make sure that we always have this mentality. We

22:11

are a startup. We are a young

22:14

company. We're always hungry for growth. We

22:16

don't want to fall into the trap that people may think, oh,

22:19

you guys are very successful in the market. And then

22:21

you are not necessarily need to

22:24

worry about your existence anymore. I

22:26

think it is actually something we

22:29

try to avoid. We always want to

22:31

make sure that in our team members, always think like,

22:33

OK, if this is actually a new day for you,

22:35

I know what other things that you always want to keep

22:37

in your mind you want to do. And

22:40

also, to keep that spirit is very important.

22:43

A lot of times that I can see some

22:45

of the mature company, they're not

22:47

necessarily losing the edge of, I

22:49

would say, this competition or losing the edge

22:51

of being innovative. I think it's more

22:53

about some of the culture has

22:57

been shifted because you have a

22:59

lot of new employees diluting your culture. So

23:01

not necessarily it's not going to be like

23:03

the old days that the co-founder is sitting

23:05

among of you. But I

23:07

do think this company has a very interesting behavior.

23:10

I see there is I can

23:12

talk to anyone at any time

23:14

via our internal communication system. I

23:17

can ping show right now. I can

23:19

ping the co-founder if I want to

23:22

tomorrow. We always keep

23:24

this kind of mentality internal is that we're

23:27

still a young company. We want to grow. And

23:29

you can feel free to talk to anyone. We

23:31

don't have a limitation for that. As long as you have a good

23:34

opinion, I would love to hear from you. Is

23:36

that creating some of the, I would

23:38

say, that chaotic situation? It might

23:40

be. But I do think that this

23:42

keeps the company very energetic. People

23:45

are willing to share, people are willing to

23:47

engage. That's very important. I

23:50

want to add one more thing we just talked about. You

23:52

asked me what is actually the

23:54

uniqueness of TikTok and versus the other company.

23:57

It's very tied up to that is I have I've

24:00

never seen a company, the engineering team,

24:03

and the product team, and the sales team are

24:05

so close. That's definitely

24:07

one of the ah-ha moments I had.

24:10

Because if

24:12

you're thinking about, if your engineer does

24:14

not really know what the market

24:17

wants, and if your

24:19

PM doesn't really know what

24:21

is actually the client's feedback, they

24:24

won't be able to get a right

24:26

product in the market. They

24:28

just won't be. They won't even tell a

24:30

good go-to-market story to advertisers

24:32

or even to our users because

24:34

they just don't know what

24:37

are the end users are thinking. So

24:39

I think it's a very

24:41

secret source for us is that our sellers and

24:44

our engineering team and our product

24:47

team and also data scientist team, we're

24:50

all collaborating really, really closely. And

24:53

that's very much a, I would say, a

24:55

such big advantage for us compared

24:57

to when a company becomes too big and nobody talks

24:59

to each other. So I do

25:02

hope that it is the thing that we're gonna continue reinforce

25:05

along the years where we are continuing to grow the company. What

25:09

does that actually look like? I imagine people

25:11

hearing this are like, yeah, we're gonna make sales

25:13

and product and end very

25:15

close. I imagine many people don't

25:17

actually do this too well. How do you actually execute

25:19

that? Is it they report to the same

25:22

leader, they sit next to each other, or

25:24

I don't know, zoom next to each other? What

25:26

actually makes that work? Yeah,

25:28

I think a couple of things. Number one is a

25:31

structure, right? Everything has to

25:33

go at other structure. So we do have

25:35

a meeting structure that we call it like,

25:38

used to be by month, and now it's actually

25:40

a quarterly level. We

25:42

get everybody together, engineering leader,

25:44

product leader, and also not necessarily

25:46

only the leader level. Some of

25:48

the team members, we're joining

25:50

the team, we're joining the force together to have

25:53

a big meeting. That meeting is 180

25:55

people-ish. It's

25:58

crazy to have a meeting at the... exercise,

26:01

especially that there are different

26:03

kind of functionality there. But one thing

26:05

we keep it really well is actually we

26:07

are using a reading

26:09

format of meeting. So

26:11

it's a doc reading. We just read

26:13

and comment and understand the context. Again, it

26:15

is the doc bringing everybody together. And

26:18

then we discuss the things that we want to make a

26:20

decision with or the things that we feel

26:22

is a blocker or things that we need to celebrate. So

26:25

that meeting structure keeps

26:27

everybody together in consensus. Again,

26:30

not necessarily only for the top leaders.

26:33

It's normal for the engineering leader and product leader

26:35

and self leader at the company level they talk

26:37

to each other. But we made

26:39

that happen for their core team

26:41

members. And the very beginning

26:43

of my time here, that was literally getting

26:46

to the IC level. So it

26:48

is pretty eye opening for me to join

26:50

that meeting first time because I was so

26:52

used to your level of different meetings at

26:54

Google. But here is like,

26:57

okay, everybody read one documentation. And

26:59

then you just understand what are people talking

27:01

about or thinking about. It is

27:03

intentional. But I do think that that structure

27:06

is a very big secret sauce,

27:08

I would say. Not necessarily we invented

27:10

it. We also learned from the other

27:12

companies. So it is actually

27:14

one of the things that we

27:16

actually deployed pretty well today here to

27:19

keep that structure running. And

27:22

the other thing is we feed

27:24

those, I would say, first-hand

27:26

market information to our PMs

27:28

and RDs. That means

27:30

we took them out with us. We just invited

27:32

them together to join the force

27:34

together to meet the clients. And

27:36

a lot of the company, if you want to meet

27:39

APMs, if you want to meet the engineering leaders, it's

27:42

literally like once

27:44

a year maybe. And also if you're investing a

27:47

ton with some of the platforms. For

27:49

us, I think it's always on. To

27:51

junior PMs, senior PMs and engineering

27:54

leaders, we invited them together

27:56

to this immersion trip recorded

27:58

to really get face time. with our clients

28:01

to really feel the heat they

28:03

are actually really facing and challenged

28:06

by using our own product. So that

28:08

kind of a I would say the

28:10

aha moment is bringing a lot

28:12

of outside insight to them and they also get

28:14

them to feel the heat of

28:16

the paints the sellers may feel. So

28:19

that worked really well too. I

28:21

think oftentimes is

28:23

a battle. It is not necessarily the general you

28:26

have to stay in the back. You

28:28

sometimes have to go to the front but we

28:30

just make that make sure that the general go

28:32

to the front quite often you know a company

28:34

too. I love that concept of

28:37

having them feel the heat. An

28:39

interesting trend I've noticed is there's a lot of Amazon

28:41

influence on the way you all operate. There's

28:43

the it's always day one idea there's the

28:45

memo culture you just described. Any idea where

28:47

that comes from is there like a senior Amazon person

28:50

they came in and helped influence

28:52

those sorts of things is it just a

28:54

Amazon's killing in there. I've noticed

28:56

interestingly Amazon has influenced the most companies in all

28:58

of their ways of working so it's not a

29:00

surprise. I'm just curious if there's anything else there

29:02

that's interesting. I think we have

29:05

the we have the benefits to standing on

29:07

the shoulder of all the Giants right so

29:09

we learn definitely always they want the culture

29:11

that you know Amazon was always championing I

29:13

think we learn from them right

29:15

so this is something that we I

29:17

would say like you know always trying to listen

29:19

and trying to learn from the industry. The

29:22

dark fashion is also learn from from Amazon

29:24

right so we kind of a studied oh

29:26

this is maybe the one of the best

29:28

practices we can employ here or deploy here

29:31

so we try it. Not even

29:33

mention we have the OKR system. So

29:35

it is actually a very good learning from

29:37

early stage from Google. So all

29:39

these I think definitely we do have

29:42

some of the I would say benefits being

29:44

the newcomer to the market and then learn

29:46

a lot of the best practices coming from

29:48

our industry peers and

29:50

really deploy here hopefully successfully and some of

29:52

the things that we just tweak it right

29:54

so for example our culture

29:57

always day one is definitely very similar to

29:59

Amazon but the implementation of that could

30:01

be different. And also the context no

30:03

control pieces. I believe other companies may have a

30:05

similar idea, but for us, I think

30:08

we just need really need to implement it in a way

30:10

that gonna be fitting to us. I

30:12

happened to listen to your podcast

30:14

with the Airbnb co-founder the

30:16

other day. He also mentioned that,

30:18

you know, how he break out the IELTS. I think

30:21

it is very similar approach among industry right now,

30:23

trying to really make sure the

30:25

team is able to talk to each other, because I

30:28

think a quote from him, if

30:30

your PM doesn't know how to sell the product they

30:32

are creating, you won't be able to

30:34

do your job better. So this

30:37

is literally how we're thinking about it too, in a

30:39

lot of way. I know that you all

30:41

move very fast. I don't wanna actually talk about that next. And

30:44

with that, it feels like your value should be,

30:47

it's always the first half of the day, instead of

30:49

it's always day one. It's always the morning

30:51

of the first day. I think

30:54

the value if I put it

30:56

in a very reactive way is

30:59

like, how can I run my second half of

31:01

my marathon faster than the first half? So

31:04

that's how I think

31:06

about it. And how do we really like continue pushing for

31:08

it? Wow, that sounds very

31:10

hard and painful, but I'd like that

31:13

metaphor. Okay, so let's

31:15

talk about how you set up the product org

31:17

to move as fast as you move. I

31:19

think there's this idea of just like running fast. I don't know

31:21

if that's a phrase you use, but just how

31:23

is the product org set up, especially

31:26

different from other teams that you've seen that

31:28

allows it to move as quickly as

31:30

you move and innovate as often as you all

31:33

innovate? Our product teams

31:35

are setting, I would

31:37

say like very importantly is global. Right,

31:39

so we wanna actually, like I said, the

31:42

number one step is if we really wanna

31:44

do global business, have to go global. So

31:46

we set up the teams really across

31:49

the board in the global locations to

31:51

really acquire global talent, who knows the

31:53

market and who knows the competition too.

31:56

So we're able to really getting the

31:58

jumpstart. in the local market. So

32:01

for example, we have the majority

32:03

of the engineer and also PMs

32:05

currently located in the west coast of

32:07

the North America, right? So

32:09

Los Angeles and also San Jose,

32:12

these are the key hubs we have for

32:14

our tech folks. And

32:17

also for North America-wise, we do have

32:19

our majority of the go-to-market leads sitting

32:21

in the New York to get closer with our

32:24

seller and also with our clients at the same

32:26

time, right? Also, you know,

32:28

it is not necessary only for North America,

32:30

like I said, we heavily invested in Southeast

32:32

Asia. So you can see that

32:35

a lot of our engineering and also PM

32:37

resources are deployed over there in Singapore to

32:39

enable them to get closer to our clients over

32:41

there as well. So really

32:43

deploy your resources globally and also

32:46

focusing on the key markets you

32:48

wanna penetrate. That's the commitment

32:50

I think we're doing pretty

32:53

good in this case. And

32:55

the second one is to really, again, I think

32:58

the PMs and the product team of settings are oftentimes,

33:02

I would say like, because

33:04

we are growing so fast, oftentimes we have to do a

33:06

lot of minor team

33:08

adjustment to catering for that, right?

33:10

So it is very usual

33:13

or common for teams to do a

33:15

little bit reorg on

33:18

annual basis or even like on a two

33:20

years or three year cycle. The

33:23

stability is important. Don't get me wrong, but

33:25

I do think that is a faster growing

33:28

company. We need to consistently to reiterate not only

33:30

the product but also our teams. So

33:33

how can we do reiteration on

33:36

the PM side, on the go to market side?

33:39

It is actually something that, you know,

33:41

I have seen this company doing really, really well. Not

33:43

necessarily we're bonding to one

33:46

team structure. We're actually bonding to the

33:48

marketing need and we're bonding

33:50

to the growth we're looking for. So

33:53

we're not afraid to break our things. And

33:56

actually I literally break out my team last year

33:58

to. make sure that

34:00

my team having more go-to-market mindset

34:02

to actually immedit them

34:05

with seller directly. So these are

34:07

the things that you know very I would

34:09

say incommemorial to a

34:12

size of this company right but I

34:14

do think that's necessary and also

34:17

that's a good mentality for

34:19

the team to really run

34:21

faster with this kind of a rigid approach.

34:23

So yeah these are the two things I think

34:25

very unique to us I think

34:28

it could also be continuously

34:30

helping us in the next phase of

34:32

the growth. Today's episode is brought to

34:34

you by one schema the embeddable CSP

34:36

importer for SaaS. Customers always seem to

34:38

want to give you their data in

34:40

the messiest possible CSV file. In

34:43

building a spreadsheet importer becomes a never-ending

34:45

thing for your engineering and support resources.

34:47

To keep adding features to your spreadsheet

34:49

importer the customers keep running in C

34:51

issues. Six months later you're fixing yet

34:53

another date conversion at Facebook. Most tools

34:55

aren't built for handling messy data but

34:57

one schema is. Companies like

35:00

ScaleAI and Pave are using one

35:02

schema to make it fast and

35:04

easy to launch delightful spreadsheet import

35:06

experiences from embeddable CSV imports to

35:08

importing CSVs from an SSP folder

35:10

on a recurring basis. Spreadsheet

35:12

import is such an awful experience in so

35:15

many products. Customers get frustrated

35:17

by useless messages like error on line

35:19

53 and never end up getting started

35:21

with your product. One schema intelligently corrects

35:23

messy data so that your customers don't

35:25

have to spend hours in Excel just

35:27

to get started with your product. For

35:30

listeners of this podcast one schema is offering a $1,000

35:33

discount. Learn more

35:35

at oneschema.co.led. I

35:38

know you mentioned earlier when we were chatting offline

35:41

is when you were trying to build

35:43

the go-to-market org for this stuff you

35:45

failed in some ways and there's some things you

35:47

learned from that experience what went wrong when you

35:49

first tried to approach this? Yeah when

35:52

I when I joined a company there were

35:55

only two people on a go-to-market side.

35:57

For the advertising business? Yeah,

36:00

there are only two people and by that

36:02

time the US and plus, I was a

36:04

Europe business, together we're having like

36:07

less than 80 people. But

36:10

the business needs to grow and we need

36:12

to hire really fast. The

36:14

first mistake I made was, by the way, the goal

36:16

is to hiring 100 people

36:18

in the sixth month to support

36:21

the go-to market. That's the speed we're into,

36:23

right? So that is early 2020 to mid

36:25

of 2020. So

36:29

within six months, I need to hire, I would

36:31

say, 100 people to support

36:33

in the global go-to market structure and

36:35

build everything. Then

36:37

the first mistake I made just

36:40

right at the right point because we're

36:43

trying to grow too fast and

36:45

sometimes as a hiring manager, I have

36:48

to compromise the

36:50

standard we're trying to return to hire.

36:53

So that's the mistakes I think I made first

36:56

and I think nobody should beat that mistake is

36:58

you need always run for the quality

37:01

rather than the quantity. So

37:03

it's an easy mistake you can

37:05

fall into the trap because the business

37:07

demand you to go faster. If you don't have the

37:09

manpower, you won't be able to but I

37:12

always believe me, when I say this, you

37:14

do the pain, right? When you have the wrong people on a

37:16

team, it's not necessarily going to

37:18

make you move faster. It's going to actually

37:21

slow you down. So that's one of the biggest mistakes

37:23

I made for my first year when

37:25

I created the team and not

37:28

necessarily myself only. So my

37:30

reporting manager, sorry, the

37:32

manager reporting to me, they are facing

37:35

the same pressure and then

37:37

it's casting down. So it's

37:39

definitely the mistakes we made at

37:41

early stage. The

37:43

second thing I can think about is

37:45

really on the context, no

37:47

control. It is not necessarily I'm born into,

37:50

to be honest, because I was trained really

37:52

like, you know, hey, this is your box, finish

37:55

your work here and then you're good. But

37:58

The reason why I... The

38:00

Meadow.really like the attitude more.

38:03

Today is literally I felt.

38:05

At. The very early stage of my time

38:07

here. Because. I was trying

38:10

to creating that can have a very black and white. the

38:12

something for my team. You hate. You can do this. You

38:14

cannot do that. But. Typically speaking,

38:16

that's literally slowing things down. Because

38:19

a lot of times you can see that. Hey we're

38:21

delivering. I'll go to like a strategy. And.

38:23

Were good, But. Literally. Would.

38:26

You don't know is your goal is not

38:28

to deliver to go to market strategy. Your

38:30

goal is to land. and go

38:32

to market strategy was self together. So.

38:34

If your job only is delivering. Know.

38:37

Your your failed Oftentimes because you're You're not

38:39

really getting a market context. You're not even

38:42

talking to your clients. So.

38:44

That was the literally another mistake I think.

38:46

taught me how to read it. I to

38:48

embrace the couch or here is context. no

38:50

control. By. And.

38:53

The So peace I think. It's also

38:55

a mistake eyes and really are a big

38:58

ah ha moment for me as well. Is

39:00

for the past couple years now and be

39:02

managing. A such big global

39:04

organization. Sometimes. Even

39:07

not myself, my managers. They don't have

39:09

time. To. Go detail and to

39:11

go talk to the clients. Which.

39:14

Is very scary. Because.

39:16

If you don't know, Again, You don't

39:18

hear. What is happening in a

39:21

market? You will know the details. In a market

39:23

you won't be able to. Take.

39:25

The. Right movement or take the right approach

39:27

to go to market. I'll even give the

39:29

see back to the ingenuity. So.

39:32

It's very important that you know the leader.

39:35

At. Any level. Needs

39:37

to be situation. You

39:39

can Not. Always don't have the wheat.

39:41

And you cannot read. Distance yourself from

39:44

the reality. So. You need

39:46

to find a balance to really get

39:48

engaged and also serious self out there.

39:50

To getting I was a getting

39:52

deeper into the problems. Identified A

39:54

Problems. And then you're able to

39:56

perform even better. because i don't believe

39:59

one thing is you're the pure, I would

40:01

say, people manager. You cannot

40:03

do that because when you do that, you're

40:06

at a very, I would say, position

40:09

to really thinking about your career because

40:13

you're losing your competitive edge from

40:15

the other, I would say, equivalent talents

40:17

in the market. I love these

40:19

stories. I love stories of things not working out, so I appreciate

40:21

you sharing these things. When someone

40:23

doesn't work out at TikTok and they have

40:25

a bad time and they let go or

40:27

they leave, what's the most common reason? Other than

40:30

just they're not good enough, is there

40:32

something that just doesn't stick with people that often

40:34

leads to, this is not the place for me?

40:36

Yeah, I would try to really think about it

40:38

in a different way. I can tell how people

40:40

can be more successful here. So I

40:42

definitely can see, we just talked about

40:44

people being very curious and people

40:47

very being nimble. It

40:49

can be more successful here. At

40:51

the same time, I think we have to admit one thing.

40:55

Join a startup and join a

40:57

rocket ship is a lifestyle. It

41:00

is not necessarily a job you're working

41:02

on from nine to five,

41:04

right? So it is a different lifestyle

41:07

and it is not beautiful for everyone, right?

41:09

So if you are not able

41:12

to adjust your mentality towards some of the

41:14

work that we are here to do, and

41:16

it's maybe not right fit for you, I'm not

41:19

saying that candidates is incapable. I

41:21

think they could be capable in

41:23

the other scenario for sure, but

41:25

it is a right fit. I think that

41:27

is a, I would say, very much towards

41:29

the situation or the company status in

41:32

the market. I can see a lot of people that

41:34

they left and become

41:36

very successful too, right? So it is

41:38

not necessarily like, oh, we think you're not good

41:40

and then you're gonna be not good for

41:43

everything about the company. That's not the case. And

41:46

one thing, and also this is my team culture,

41:48

I try to create is, I'm

41:51

happy to say that when

41:53

an employee reach out to me, say,

41:55

hey, Ray, I'm actually leaving the

41:57

company. As long as they're telling me

41:59

that. you're going to a better

42:02

place or a place that they can continue to

42:04

grow their career, I'm happy for

42:06

them because oftentimes

42:08

my last question during

42:10

my interview is what

42:13

is actually your goal

42:15

in the next three to five years?

42:18

And also like, you know, I'd be really honest with them,

42:20

say, hey, I don't think this is

42:22

the job for you forever, right? Nobody gonna work in

42:24

this forever. If you can, great. But

42:26

what is really like, you know, you're a North Star.

42:28

I think that's the part that I would love to

42:30

co-partner with you because I always believe

42:33

one thing is it is not only about

42:35

achieving the company goal, it's also achieving your

42:37

really like career goal or your employees

42:39

career goal together. So I

42:42

want to create in that culture here as well.

42:44

So yeah, I think I'm doing so far

42:46

so good. Most of my team members,

42:48

when they actually moving on

42:51

internally or externally, I'm able

42:53

to say that, okay, that's a good choice. If I were you,

42:56

I may probably do the same thing. It

42:58

is actually a very good culture, I think

43:00

I would love to champion across.

43:02

On that first point, I'm also a huge advocate

43:04

of just you will be successful if

43:06

you work very hard. I know there's a

43:08

bit of a backlash at working along and

43:11

thinking too much about work-life balance and I feel

43:13

like it's actually really important to work

43:15

a lot and work long hours often to be

43:17

successful, especially at a company that's going

43:20

through this because it's not gonna last forever. I

43:22

think at the end of the day, it's a personal choice, right?

43:24

It's a very much like a personal choice. If

43:27

you are excited about this, if

43:29

you want to grow together, yeah, this maybe

43:31

is a good thing for you and

43:34

also depends on the life stage, right? So

43:36

some of the people they want to

43:38

actually get in more family time, I

43:40

think that's also the right choice too.

43:43

But it just depends on your personal,

43:45

I'll say, any of their personal choice

43:47

rather than if the company

43:50

demands that. I cannot

43:52

force my team to work in long

43:54

hours. I don't want them to work in hours.

43:57

I think it's more about if you're able to

43:59

deliver. If it requires a bit longer

44:02

time to contribute, I

44:05

think it's okay, but you will also get rewarded

44:07

very well too. So what's getting and what's

44:09

get out? So I think

44:11

it's, again, I do believe that this

44:13

is the quality and also the value we're

44:16

evaluating here as well. And

44:19

even though it's hard in the moment, I find that

44:21

those are the times you remember most fondly in your

44:23

career when you just go all in, I'm going to

44:25

work really hard and do the best possible job I

44:28

can do. Assuming that

44:30

doesn't last forever, those end up

44:32

being the most impactful, helpful to your career,

44:34

most proud moments when you're just like, look what I

44:36

accomplished. And so I'm

44:39

on the same page. I want to

44:41

talk about being successful on TikTok

44:43

as a creator, as a business, as an

44:45

advertiser, but a couple more questions real quick

44:47

on how TikTok operates. You mentioned you do

44:49

OKRs. Just briefly, is there anything

44:51

that you've learned about being successful

44:54

doing OKRs within TikTok? Is there anything different that

44:56

you all do versus how other companies think about

44:58

OKRs? It is definitely

45:00

a company alignment that we are using

45:03

OKR as our, basically,

45:05

the system to make sure that everybody is

45:07

working towards the same goal. Implementing

45:10

wise, I think definitely we have a lot of room to

45:12

improve. So how often do you

45:14

actually see your team able to go

45:16

the OKR at the end of a quarter and

45:19

also putting the OKR really two weeks or one week

45:21

before the beginning of a quarter? I

45:24

have to say that shame on me, I sometimes

45:26

delay it a little bit. But

45:28

I think the goal is always there

45:30

to using OKR system as our no

45:32

star to drive the behavior and also

45:34

to align. Again, it's

45:36

very important to align on the OKRs

45:39

because I can see a lot of times the

45:41

OKRs are putting in, but they

45:43

are very siloed. And that

45:46

is not really necessarily helpful for

45:48

the company want to achieve in really high

45:50

growth. So I think it's

45:52

very important that we don't take OKR as a

45:55

shell, but we take OKR as its core is

45:57

cross-function alignment, cross-function alignment. goal

46:00

settle. So these are

46:02

the things we're still continuing

46:04

improving. Is

46:06

the way that OKRs work at TikTok, is there an

46:08

OKR per team and they all kind of trickle up

46:11

to a company level OKR? Is it less structure

46:13

that way and teams kind of decide if they want to use

46:15

OKRs or not? How does that roughly work? The

46:18

structure is basically the guidance is you're using

46:20

the key result to evaluating and then you

46:23

put the steps in between. Right. So that's

46:25

how at least my team has been using

46:27

this. I think the things that we

46:29

can improve is the input and output. So

46:32

the output is very clear. But what

46:34

is actually the input sometimes is debatable, sometimes

46:37

I have to say. And

46:39

also oftentimes your output is

46:41

other people's input. Are you

46:43

able to connect in the dots over there too? Then

46:46

that's actually the part that requires a

46:48

lot of I would say reinforcement on

46:51

alignment. Definitely we're getting better.

46:53

Don't get me wrong. We're totally not perfect

46:55

for sure. But I do

46:57

see there was a lot of outside momentum

46:59

to leveraging the system better. If you know

47:01

other companies doing this really, really good, please

47:04

shoot in my way. I would love to

47:06

learn from them. One last

47:08

question here. You do planning, you have OKRs. Just

47:10

briefly, how often do you all do planning?

47:12

Is there a yearly plan that you put together and then a

47:14

quarterly detail plan? Yeah,

47:16

we do have annual planning cycle.

47:19

But I have to say that, you know,

47:21

our annual planning cycle is the baseline. We

47:23

often do a lot of iterations in the middle

47:25

of the year. And also

47:27

on a quarterly basis that we're able to pivot

47:30

in really nimboli to really catering to

47:32

the things that we see in the market. Right.

47:34

Some of the longer term strategy won't change

47:36

just like the platform we want to always

47:38

creating, you know, inspiring and also a

47:41

frictionless and immersive experiences for users. It

47:43

won't change. But anything into

47:45

the core of like, how do

47:47

we realizing that you're always a

47:49

consistent experience over there? I

47:52

cannot speak for the user product side, but at

47:54

least from advertising product side that, you know, this

47:56

is always the rigorous approach we're taking. And

47:59

for the go to market. part that's also creating

48:01

a very different behavior for us, right? Because oftentimes

48:04

if you have a solid and kind of

48:07

a static protocol map, you

48:10

can do go-to-market relatively easy, I would say,

48:12

because everything is planned. But with

48:15

a environment like that, that basically

48:17

make the go-to-market and

48:20

also the product feedback loop much

48:22

more shorter and faster. So there's

48:24

a lot of, I would say, pressure or

48:27

actually put it nicely, there's a lot of

48:29

innovative things that on the go-to-market side and

48:31

also on the sell side, the company or

48:34

the teams need to actually do to make sure that

48:36

we're able to catering for that. But again,

48:39

this is a teamwork rather than only one

48:41

side of the work. So far

48:43

so good, I would say a lot of things that we've

48:45

been able to achieve within the

48:47

past couple of years has been

48:49

already proven that this approach has been working for

48:52

us. But not necessarily

48:54

they are always like, it's perfect already,

48:56

always room to improve to make sure

48:58

that we have more structural

49:00

approach as well, so that the market able

49:03

to keep the patient with us. We don't

49:05

want to overwhelm our advertisers or our users

49:07

either. So that's also the other part that

49:09

we need to continue optimizing to. Okay,

49:12

let's talk about a different topic, which is being

49:14

successful on TikTok. So the way I think about

49:16

it in my head is there's how to be

49:19

successful as just a regular human creator person, how

49:21

to be successful as a business trying to just

49:23

create viral content, and then being successful as an

49:25

advertiser, which I know is where you spend a

49:27

lot of time. So let

49:30

me just ask, is there a tip you

49:32

could share for someone to be successful, say,

49:34

aka go viral on TikTok?

49:36

I imagine your answer will be just produce something

49:38

people love and want to share and like, but

49:41

I guess is there anything that could be tactically

49:43

useful when you're creating content TikTok to help you

49:46

go viral? I think if I know

49:48

that, I definitely will already become

49:50

a very successful creator, I have to say our

49:52

system is very much smarter than I

49:54

am. I cannot trick the system,

49:57

but I have seen a couple of good cases. Right.

50:00

Number one thing is that you have to really

50:02

like be unfiltered. I mean you

50:04

don't really need to be perfect on this platform.

50:07

I mean that's the beauty of it. You can be yourself, you

50:10

can really share the things that you like and

50:12

if you really master at one thing that you're

50:14

really really good at and you want to showcase, this

50:17

is the platform for you to shine. Because

50:19

not necessarily that you know, we are fully

50:22

saturated, right? And also all algorithms distributing

50:24

the content in a very different way, right?

50:27

Some of the other platforms they are I

50:29

would say like a people-based, a friendship, a

50:31

friend-based. I think for us it's

50:33

purely based on actually you're

50:35

creating something that everybody want to see. So

50:37

let's see if we can distribute it more,

50:40

right? So I think continuously to bring

50:42

new content to this platform and testing

50:44

and finding your own competitive

50:46

edge is going to be very important

50:48

as a successful creator. And most

50:51

of our creator has been doing that. And

50:53

I can see some of our biggest

50:55

TikTok stars. They are literally practicing

50:58

this every single day. And

51:00

I just think creativity

51:02

and that part of I would

51:05

say getting the nuances is

51:07

the key part that to be more successful

51:09

on the TikTok community. And

51:11

the second thing is it's

51:14

including also for brands as well because I

51:16

consider brands as our creator as well. They

51:19

really need to embrace the culture and

51:21

the community here. So

51:23

really listen and understand what are the user behaviors

51:25

on the platform to understand what do they like

51:27

to see and also the

51:30

messages or the presence could

51:32

be very different from your other

51:34

media channels or as a

51:36

creator it could be very different from your

51:38

other outside platforms. So that's

51:41

the other thing that it's going

51:43

to be challenging because for them to shift in the

51:45

mindset. But I do think that

51:47

definitely was trial, right? So some of

51:49

the I would say our early adopters has

51:52

already been proven that when you

51:54

do embrace the culture here you're

51:56

able to acquire a ton of different kind

51:58

of a User or you

52:01

know the audience to your channel and

52:03

you can show a different side of yourself as well So

52:06

yeah, I've been trying to do that. I have

52:09

not really like finding my competitive edge. I have

52:11

to say but I keep trying Is

52:14

there an example you could share if someone that has done that really

52:16

well either be really authentic and also

52:18

embrace the community of a business Specifically

52:21

that has done this really well and taken

52:23

off not as an advertiser. There was

52:25

a one creator. I remember called Shaba

52:30

she's a singer and she

52:32

is able to call my eyes because she

52:34

was able to basically wrap and also doing

52:37

some of the songs cover in a very

52:39

different way because She's a

52:41

minority and she was able to basically

52:43

using her minority identity as actually a

52:45

everybody was thinking I'm supposed to be

52:47

doing like Bollywood music, but actually you

52:49

know what I'm not I'm doing a

52:51

lot of very like You know

52:54

just hip hop and also the music that you

52:56

know people may think like I'm

52:58

not good at right so it is pretty

53:00

fun to watch that kind of a comparison

53:03

or the contrast between a

53:05

creator and also she's able to put a lot of Original

53:08

music on the platform to really inspire more

53:10

people to do the same thing. There's

53:13

another music. I Was

53:15

a take a creator. So he was

53:17

pretty big on the other platforms but the total

53:19

approach from him is He's

53:22

basically changing the lyrics make it a

53:24

very relatable as a personal life because for example

53:26

He can totally change the lyrics

53:28

from a old battery boy song or instinct

53:31

song to make it like a related with

53:33

his Daily communication with his

53:35

wife making the really like relatable

53:37

and fun So these are the things I

53:39

think is very unique to us Like

53:41

if you are able to test and find something

53:43

like new like that You're able to find a

53:46

new batch of audience and even go viral on

53:48

the platform Okay, so then

53:50

switching to the advertising network a

53:53

lot of listeners here are thinking about I imagine

53:56

Advertising on tick-tock. There's kind of classically

53:58

been Facebook and

54:00

Google are the two places to

54:02

do run paid ads. Paid

54:05

ads are a huge growth driver for tons of companies.

54:07

It's one of the easiest, you

54:09

could say, or one of the most traditional way to grow.

54:12

TikTok obviously is emerging and has already emerged

54:14

as one of the newer advertising networks. So

54:16

there's a lot of people thinking about how

54:18

do I succeed as an advertiser on TikTok.

54:21

So what advice do you have for people? One,

54:24

who is it best for? I imagine TikTok

54:26

isn't the best place to advertise for every

54:28

sort of business. So what sort of businesses

54:30

are best aligned to be successful on TikTok?

54:32

And then just what advice can you share to do

54:35

well as an advertiser on TikTok? Yeah,

54:37

I see a lot of really different types

54:39

of advertisers already find their success on the platform.

54:42

One thing that they actually can

54:44

do that is really due to

54:46

a couple of things that they're

54:48

doing. Number one is,

54:50

like I said, they're embracing this

54:52

platform. They actually do a

54:54

lot of things is TikTok first. I have

54:57

a couple of advertisers they

54:59

have actually creating their own internal

55:02

creative team, just dedicated

55:04

for TikTok. So

55:06

they actually produce a ton of

55:08

creative every single day to

55:10

actually test and learn to understand the

55:12

platform and understand the community they're engaging

55:14

with. So also leaning

55:16

in is the first part. It's harder,

55:19

but it is not that hard. As long as

55:21

you try it, you will feel that every single

55:23

day is getting easier. And

55:25

also we make a lot of tools to make things

55:27

easier for them as well. Like creative, we have also

55:30

a lot of resources on the platform, the

55:32

creative hub, and also we have creative

55:34

analytics help you. So these are the

55:37

things that we're able to basically help

55:39

the advertiser to leaning in more. The

55:42

other angle to leaning in more is test and

55:44

learn. A lot of times

55:46

that people don't know how to

55:48

really run ads on

55:50

this platform. Google is very much

55:52

search front. They are really leading

55:56

on the intent graph and

55:58

meta is really really on the

56:01

people graph they're making. I

56:03

mean, TikTok is the content graph. It's

56:05

very different, I would say, machine compared

56:07

to the other two. And it

56:10

requires different way to optimizing and

56:12

to leveraging the tools we have.

56:15

So if you're applying the

56:18

same logic from meta or

56:21

Google into TikTok, not

56:23

necessarily you are able to see great

56:25

success, I have to say. So you

56:27

have to really get to

56:29

the detail and to learn how to operate

56:31

in this platform at the very beginning. Of

56:34

course, like I said, we're trying to make

56:36

things as simple as possible, because we strongly

56:38

believe that the advertisers job is to

56:40

taking care of their own business, and

56:42

our job is to service them. So we

56:44

definitely make things a bit easier and along

56:46

the way. But still, it's

56:49

a little bit learning for advertisers to change

56:51

their mindset when they engage with us for

56:53

the first time. And I

56:55

can see that, again, for example, last Q4,

56:58

I can see a lot of advertisers taking

57:00

this approach to really listen to

57:02

us and understanding what is our

57:05

best practices. They actually see a

57:07

very successful Q4 on the platform. So

57:10

I do think that if you want to

57:12

do more, just do more tests and learn with us

57:14

and to really understand the impact from TikTok. Just

57:17

to understand this point about versus Instagram, because I think a

57:19

lot of people probably run on them on both platforms and

57:21

try to see which one's working better. Your

57:24

point is the same content won't work as well on

57:26

one versus the other. So just so people understand what

57:28

the main difference there is, I know you talk about

57:31

there's the friend graph versus TikTok just spreads it all

57:33

over, and anyone can see it. You don't have to

57:35

be friends. And it's really good at getting content

57:37

out. So what is it that you

57:39

would do differently if you're making an ad

57:41

video for Instagram versus TikTok? I

57:44

think the TikTok video, it's more about

57:46

the back end settings, right? So how

57:49

often do you actually changing creatives?

57:52

I think for us, it is

57:54

actually pretty rigid. You want

57:56

to actually test more creatives on this platform and

57:58

see which one is actually working. working. And

58:01

then we also have really detailed guidance on how

58:03

do you set up your campaign structure to

58:05

make sure that you're able to be more

58:07

successful on a platform. So these are

58:10

the basic hijinks we talked

58:12

about. You can see these

58:14

guidance are very different from what

58:16

Meta has today or even Google has

58:18

today because we're just basically different platforms.

58:21

And oftentimes you can also hear that we

58:24

require a bit more real time

58:26

react on the platform due

58:28

to some of the trends we have seen. So

58:31

that is the part I feel like if

58:33

an advertiser wants to engage more with

58:36

our really the sales team and

58:38

they're able to provide more guidance to you, you're

58:40

able to see more success there. But

58:43

a lot of things will be counterintuitive,

58:45

I would say, because the intuitive you have

58:47

learned is coming from the other platforms. But

58:49

technically we're not. So a lot

58:52

of things that, oh, this doesn't make sense to me,

58:54

but why don't you try it? And we

58:56

make actually that really easy because

58:58

we're showing a lot of, I would say,

59:01

out of credit to incentivizing

59:03

our advertisers to try it. And

59:07

even though hopefully they can see the result

59:09

is proven itself. God, I

59:11

think that's such an interesting point, this idea of

59:13

testing more, which basically you're saying with Instagram, certain

59:16

people will see it and they're not going to

59:18

be, that's not going to

59:20

be shown to tons of random people. So you basically

59:22

have one shot at getting this in front of the

59:25

Instagram crowd versus TikTok just tries it,

59:28

this explore and exploit kind of approach

59:30

is like, we'll just keep trying stuff

59:32

until something sticks. Yeah, I think exactly

59:34

like that. I

59:36

think a lot of times that I

59:38

think advertising, especially when digital advertising becomes

59:40

a thing, right? So

59:43

we kind of think everything can be calculated because

59:46

you have the data. But

59:48

the beauty of advertising is never

59:50

like that. The core

59:52

value of advertising is to tell people

59:54

don't know you exist and

59:57

tell them the what you're doing for them and

59:59

then creating these. demand. Discover

1:00:01

is the core of advertising to

1:00:03

me because I was never expecting

1:00:05

my wife telling me that was she gonna buy when

1:00:07

she walk into a shopping mall. If

1:00:10

I know that I will stop her already. She

1:00:12

oftentimes that you know get out something different right

1:00:15

so this is not planned. I think

1:00:17

that's literally one of the behavior I

1:00:19

would love to emphasize

1:00:21

more is you want to

1:00:23

be open up your door to more consumer

1:00:26

because we are a

1:00:28

digital version of word of mouth. I

1:00:31

always compare us to that because

1:00:33

it is the way that how the

1:00:36

digital era becomes more human because

1:00:39

it is actually helping the user

1:00:41

to discover new things just

1:00:43

like what they used to do. There's

1:00:45

a new place in a certain area you

1:00:48

just go explore. It is just

1:00:50

like that. So I think that's the reason

1:00:52

why I think at the very beginning continue

1:00:54

doing this kind of open-minded testing with

1:00:56

us will be a very good approach

1:00:59

to get some early learning and

1:01:01

eventually that you can refine your approach. But

1:01:04

at the beginning I would highly recommend that just

1:01:06

be open up and also take some

1:01:08

risks because with us together and we're

1:01:10

able to show you how

1:01:12

much we can actually benefit in the business. Awesome

1:01:15

and on that point that was the other piece of advice

1:01:17

you shared is pay attention to the trends so that

1:01:20

you can connect your ad to things that people

1:01:22

are already laughing at or

1:01:24

finding really interesting. I feel like Duolingo

1:01:26

is incredible at this. Their videos are hilarious

1:01:29

and I think they're all just organic videos

1:01:31

and a lot of them connect the trends that are.

1:01:33

Yeah it's funny you brought up

1:01:36

Duolingo because I'm actually now I have

1:01:38

become a heavy user of Duolingo myself

1:01:40

because I found I watched

1:01:42

a video on the TikTok. I

1:01:45

think just basically like kids like

1:01:47

just randomly learn a different language

1:01:49

and make a lot of mistakes and it's really

1:01:52

funny and then I just download the app because

1:01:54

I didn't know. I've been using Duolingo

1:01:56

for the past 40 days as a

1:01:58

New Year resolution. I'm going to

1:02:00

be myself to learn Japanese. Wow. 30-day streak?

1:02:03

Yeah. Amazing. I'm at 25 days. Okay,

1:02:07

great. We're on par pretty much. Are

1:02:09

you in the Ruby League or Emerald League? Which league

1:02:11

are you in right now? Emerald right now. Emerald.

1:02:14

Okay, I think I'm in Emerald too. So,

1:02:16

we're on par. Just to close

1:02:18

the thread on this. So, you're talking about one

1:02:20

of the benefits of TikTok ads is

1:02:22

awareness building. Basically, more top of funnel.

1:02:24

I know you also focus a lot on taking

1:02:28

action, not just brand awareness. There's also

1:02:30

a lot of... So, maybe talk

1:02:32

a bit about that, just like that's also a big part

1:02:34

of advertising in TikTok. Yeah, I think the

1:02:36

beauty of word of mouth is actually... You

1:02:39

know that word of mouth leads to actions. Right?

1:02:42

So, I think TikTok, we oftentimes,

1:02:44

people are thinking that, oh, TikTok

1:02:46

is really good for building awareness,

1:02:48

building upper funnel, or some

1:02:50

of the discovery funnel. But, I really

1:02:52

want to say that we want to prove, and

1:02:54

also we already proved a lot from the studies

1:02:57

we have seen from third parties that we're

1:02:59

driving actions at the same time. And

1:03:01

this is literally the ambition we're trying to

1:03:03

really talk to all over the hizars. Especially

1:03:06

on the commerce front

1:03:08

that, you know, shopping and TikTok

1:03:10

shop and shop ads. It

1:03:13

is actually the proven point that we see.

1:03:16

And also, this is not necessarily coming off of our

1:03:19

illusion, right? Because we

1:03:21

see there was a biggest trend on TikTok, is TikTok maybe

1:03:23

buy it. We have billion

1:03:25

level views on that. It's continued

1:03:27

growing. And this literally

1:03:29

inspired us to do these products. Like

1:03:32

I said, you know, one of the

1:03:34

very important things here is we drive

1:03:36

our product by listening to our user and see the

1:03:38

behavior from them. And we see

1:03:40

the behavior and now we're trying to capture that

1:03:43

and provide the best service to our user and

1:03:45

also help our advertisers to really

1:03:47

shipping their product. So I just

1:03:49

think that, you know, this year people

1:03:51

will see us more as a FullFunnel

1:03:53

solution platform rather than only, you know,

1:03:55

building the brands. Because we want to

1:03:57

actually impact on FullFunnel for our advertisers.

1:03:59

Again, driving their business without is more

1:04:01

important to us. Say a

1:04:04

startup is starting to think about advertising

1:04:06

on TikTok. Maybe they've done some Google

1:04:08

ads and Facebook ads.

1:04:10

What do you recommend they plan

1:04:13

for in order to just see if this could work

1:04:15

for them? Like, how much time should they give it?

1:04:17

How many ads should they run? How much budget should

1:04:20

they allot to just explore this as a growth

1:04:22

channel for them? I would say at the

1:04:24

very beginning, the investment

1:04:26

will be coming from they're leaning into creating

1:04:29

a business account with us. So

1:04:31

this is actually how you're engaging with your

1:04:33

community. But even before that, I think

1:04:35

just do some research on a platform

1:04:38

and be the user as a TikTok to

1:04:40

really experiencing it and see the differences. And

1:04:43

then you are thinking about how can you

1:04:45

actually connecting your behavior or your

1:04:49

desired behavior coming from the user with your

1:04:51

business. And then you're creating

1:04:53

content around it. And

1:04:55

that's the moment I think this first step is creating

1:04:58

your business presence on TikTok. And the

1:05:00

idea there is just an organic account

1:05:02

you create, let's say, Lenny's podcast,

1:05:04

which actually have my Lenny's podcast is on

1:05:06

TikTok. So we can use that as

1:05:08

an example, maybe. So you're saying start off

1:05:10

just creating free business account on TikTok and

1:05:12

posting videos just to see how it feels

1:05:14

and how it goes. Yeah. Just see how

1:05:17

it feels, right? So maybe some of the

1:05:19

videos you don't get any views and some

1:05:21

of the views you get more views, right?

1:05:23

And the other day, you can test some

1:05:25

of the advertising products, drive those awareness, and

1:05:27

see if it's actually driving impact for you,

1:05:29

right? And then you have to do more

1:05:31

maybe testing with us or A-B

1:05:33

testing or geospitting testing eventually. It

1:05:35

depends on how big the investment

1:05:37

is. You can see there is

1:05:40

actually a directional impact on

1:05:42

your business. And

1:05:44

also, we are giving you reporting and insights on

1:05:46

how you're doing on the platform. So

1:05:48

you can optimize in towards that. But also, the

1:05:50

very important part is trying to get

1:05:53

a feeling of the platform by creating your organic

1:05:56

presence. And then try to launch

1:05:58

the ads account. to make

1:06:01

sure that you're able to drive more

1:06:03

traffic to your desired destination or to

1:06:05

a desire actions that you want a user to take

1:06:08

and continue refining that. Along the way there are a lot

1:06:10

of things that you're going to learn for example how

1:06:12

can you leverage in automation solutions on the platform

1:06:15

and how can leveraging some of the of the

1:06:17

creator trends you detected on the platform and

1:06:19

also some of the tools that we're creating

1:06:21

to help you to generating those scripts right.

1:06:24

So these are all the things that you can learn

1:06:26

from the platform. In terms of time

1:06:28

investment I think at the beginning of the month

1:06:31

definitely it's going to be I hope it'll be a

1:06:33

little bit more intense of learning

1:06:36

so that you're able to get a rhythm in there and

1:06:38

along the way that as long as

1:06:40

it's been become more automated and also

1:06:42

get more understanding towards the business you're

1:06:44

able to actually creating I would say

1:06:46

more relevant content for the platforms by

1:06:48

leveraging our creators or by leveraging some

1:06:50

of your own I would say resources

1:06:53

from their third party for example. So

1:06:56

I think yeah it takes a little bit a

1:06:58

learning curve but I do think that the result

1:07:00

will surprise you. And with the

1:07:02

implication there give it a month like spend a month

1:07:04

running ads or is that not what you're

1:07:07

saying? I think oftentimes we say

1:07:09

like a month minimum to run ads

1:07:11

because I think it's actually a learning

1:07:13

curve for advertisers to really get into

1:07:15

understanding the behavior in the platform. And

1:07:17

how many ads would you suggest and I know

1:07:20

there's not like a rule of thumb but just

1:07:22

how many ads would you suggest they try to

1:07:24

run in that month to give you a real

1:07:26

sense of like this is could work or like

1:07:28

no. The more the better I would say at least

1:07:30

10 different ad creatives

1:07:33

it will be ideal per week and

1:07:35

the more the better. 10 per week?

1:07:37

Oh wow okay. 20.40 potentially. Yeah 10

1:07:40

per week. I would say like we

1:07:43

can see that it is a little

1:07:45

bit of I was like a new audience is there because

1:07:47

lot of it oh I don't have that

1:07:49

resources. But as simple as

1:07:51

possible it can give you a tool right we have

1:07:53

CapCut as a tool. I created

1:07:55

my anniversary video For

1:07:58

my wife by using that tool. Don't

1:08:00

tell her what minute Now everybody knows. That

1:08:03

she think that he might alison all the

1:08:05

way as Landsea to this episode using it

1:08:07

takes a lot of toxicity in the production

1:08:09

is amazing Week reading natural specifically for our

1:08:11

creator and us of our of has any

1:08:13

user in general. So. You're able

1:08:15

to do a lot of our say. Automated.

1:08:18

And customize the way in. So.

1:08:21

You're able to germany those contents are the

1:08:24

finger tips. So. It it

1:08:26

will be a really good how for on of tethered

1:08:28

Wanted to be more self service back. On.

1:08:30

The other hand, we also have the

1:08:32

parties to making acidified. To

1:08:34

talk for a service providers. On

1:08:37

a creative side to help you as well

1:08:39

so depend on a level of out advertiser

1:08:41

you are. There most common mistake

1:08:43

people make when they try this out where you're just

1:08:45

like often been like you for. Here's what you did

1:08:47

wrong. There's. Something there's just like just

1:08:49

don't do this thing as lot of people make this

1:08:51

mistake in the muscle and tic tac. Yeah the the

1:08:54

first one is I I can see a lot of

1:08:56

a taser wanna do. Instantly

1:08:58

the want to do like remarketing.

1:09:01

Or they want to do. and I guess very

1:09:03

small niche target on a tough one because you're

1:09:05

limiting yourself. Like. Us as it is. And

1:09:07

what about again of? Getting. To the rhythm

1:09:09

to understand why most affluent. So.

1:09:11

A broader tugging approach is actually recommend

1:09:14

it. As an early stage and Mozilla

1:09:16

them sizes are already doing that today because for

1:09:18

the see I can see date for the first

1:09:20

two years in a business. Especially

1:09:22

where we acquire new advertisers. Often

1:09:24

times together platform say i want to do. This.

1:09:27

And dad's like our really refine my

1:09:29

targeting etc. And then we

1:09:32

just recommend Hate. By the way to this

1:09:34

comparison is have a complete set up like

1:09:36

this going on. But. This is a

1:09:38

recommendation and you can see the difference. A

1:09:40

little lot most of them go see a very

1:09:42

big difference over there on it. May.

1:09:45

Think ran, I have to run. I'm going to stuff the

1:09:47

lightning round the let me ask you one question from like

1:09:49

Lightning round. Your. Favorite

1:09:51

tic toc account that you have been just like

1:09:53

really loving these days. I'll share mine rock like

1:09:55

him and see if anything comes to mind. There's.

1:09:57

A lady I found recently who. is

1:10:01

silent baby product reviews

1:10:03

where her baby's like sleeping in the room. And

1:10:05

she's like, shh. And then she just goes through

1:10:08

20 different baby products very quietly.

1:10:10

And it's hilarious. I'll link to it in the

1:10:12

show notes. If you have a kid, you'll love

1:10:14

it. Is there anything that

1:10:16

you love or want to highlight? I

1:10:19

do have one creator I am

1:10:21

actually actually following. It's on, like,

1:10:24

he's a magician. He basically

1:10:26

used very, I

1:10:28

would say, very normal things. It's

1:10:31

just handy around him to

1:10:33

make something that looked like very cool magic.

1:10:36

I always would like, how did he make

1:10:38

that? So I'm actually following that and getting

1:10:40

more inspiration on myself. He's like, can I

1:10:42

do that? No. I think that's

1:10:44

more about my personal hobby to

1:10:47

see something like that. It's very, very cool

1:10:49

to see people can do these kind of

1:10:51

tricks by using normal stuff around them. Ray,

1:10:55

thank you so much for being here. Two last questions. How can

1:10:57

folks reach out if they ever want to learn more about this

1:10:59

stuff, if they can? And how can listeners

1:11:01

be useful to you? I

1:11:03

think feel free to reach out to me

1:11:06

on LinkedIn if you want to discuss more

1:11:08

about some of the go-to-market challenges you're facing.

1:11:10

I think we're facing a

1:11:12

lot of, I would say, similar challenges

1:11:15

every single day. And also, in

1:11:17

terms of on a product setting point,

1:11:19

different companies have a different product philosophy.

1:11:22

I don't think we are always right.

1:11:25

I was always recommending to receive a

1:11:27

lot of feedbacks or

1:11:29

recommendations. And that would be

1:11:31

really, really nice to have, to form these kind

1:11:33

of leveraging your audience in my community to

1:11:36

teach me a lesson sometimes. That would be even better. Amazing.

1:11:40

Ray, again, thank you so much for being here. I feel

1:11:42

like people don't have a ton of

1:11:44

insight into the way TikTok operates. And I appreciate

1:11:46

making time to do this. No, it's a pleasure,

1:11:48

Lenny. Thank you very much for having me. Bye,

1:11:50

everyone. Thank

1:11:53

you so much for listening. If you found

1:11:55

this valuable, you can subscribe to the show

1:11:57

on Apple Podcasts, Spotify, or your favorite podcast

1:11:59

app. Also, please consider giving

1:12:01

us a rating or leaving a review as

1:12:04

that really helps other listeners find the podcast.

1:12:07

You can find all past episodes or learn

1:12:09

more about the show at lenniespodcast.com. See you

1:12:11

in the next episode!