product strategy 101 | 5 types of product managers and their thinking framework, AI impact to PMs

A good product manager is the center of the team. A lot of your cross functional

the team. A lot of your cross functional partners trust you. They trust you to make the decision. They trust you to be the domain expert, to give them a direction, to give them a trustworthy strategy and a good solution to help the team make impact on your customers.

Right? As a good product manager is the center of the team. A great product manager decentralized the team. This is

another thing to tell a lot of you guys to keep you guys up to date how we use AI today. So in big companies, we use AI

AI today. So in big companies, we use AI for coding. We already don't have data

for coding. We already don't have data scientists. For product managers, we use

scientists. For product managers, we use data AI data scientists. Um I think this is when a disruptive technology happened. most of the tech found most of

happened. most of the tech found most of the founders are tech founders and it just leads to a not good product experience. Um now we're not saying not

experience. Um now we're not saying not good not even targeting like conservatives we're just talking about visionaries visionaries don't even want to use it. So a good product is also a key for you to um build up your business advantage. These are the divisions that

advantage. These are the divisions that you probably want to keep an eye on and that's also why a lot of companies that I um use as examples in this class fall into a lot of these divisions because I want you guys to research on these divisions. These divisions might have

divisions. These divisions might have more positions open might have more opportunities in the future. So yeah

keep an eye after you analyze the whole thing you may come up with a solution like this. Right today if you subscribe

like this. Right today if you subscribe premium you can see who are the like where are the recruiters come from. You

can have a list of company uh information and you can check out their jobs as well. So there are five types of product managers and these five type of product managers are um are divided based on the job the skill sets they

have. So today we will talk about how to

have. So today we will talk about how to write PRDS and MVP design and I will expand the topic a little bit and really dive deeper into what a product manager's life look like just to give

you guys an overlook of different types of product management and what type of skill set uh that you guys need to have for different type of product management. Uh we briefly talked about

management. Uh we briefly talked about this on our first class but we didn't have time to dive deeper into it. Most

of the classes in this course were focusing on product strategy. So we talk about industry analysis, company analysis, competitive analysis and all

of these analysis are uh the evidence for you to define where your product is going, right? And after the product

going, right? And after the product strategy uh under that product strategy you want to look at what target users you have what are their job to be done

and if they have a specific job to be done for your product then what are the user pain points today to stop them from uh doing their job that they want on

your product right so identify the user pain point which need your user empathy uh a lot a lot you need to put yourself into the user shoes and then after all

the pain points you um designed your solution. So there might be a lot of

solution. So there might be a lot of pain points from a user and for all of those pain points some of them might not be that painful right and for some of them might only have uh you only

probably have one user having that problem. Some of the user painoint you

problem. Some of the user painoint you probably have a million user having that problem right. So from painoint uh you

problem right. So from painoint uh you already should have a prioritization of which one you are dealing with right now. Every product is not perfect and

now. Every product is not perfect and there will be a lot of problems uh when you work on a product but you have to have a very clear prioritization um in terms of which painoint you're going to

work on first. And after you choose the pinpoint that you are tackling right now and you dive deeper into what type of solutions that you can design um on your

product, right? And um for different

product, right? And um for different solutions, they might have different estimated impact um this depends on how your product works and also how uh big

of a total addressable market or how well it could solve the user pinpoint and that will define how well it could solve the problem for the users. So at

the solution stage, you can also have your prioritization, right? Uh you only have one engineering team. So you

probably want to work on one uh biggest bet. Um and the bet you have the most

bet. Um and the bet you have the most conviction and then you go to uh MVP design. Um basically the whole solution

design. Um basically the whole solution might be a lot, right? And you might have a lot of hypothesis when you come up with the solution. Uh you want to

have data to support your hypothesis. uh

later I will talk about why only data okay uh you want to have data to support your hypothesis if it is a zero to one product you don't have data so you have

a lot of hypothesis within this um solution so you want to tackle this bit by bit right you want to deliver MVP first you want to ship as fast as possible you want to iterate later so

you want to ship the MVP and see how people react to it to validate your hypothesis and that's how you can move further and within the product design phase. We also have market research and

phase. We also have market research and UX re research. What is the difference?

Market research is when you have an idea, you think you want to do this, right? But you don't have design. You

right? But you don't have design. You

just want to um do research from your potential users to make sure that this is a real painoint for them and uh they

would be interested in this idea.

And UX research is when you have a design and you want to see how users interact with your user experience. And

then after all of this, right, it goes to um the product execution phase where you need to define your success metrics.

What are the metrics uh that you want to look at that could define the success of your product, right? Uh every time when you launch a product um specifically on

big products like LinkedIn, Meta, Google, like all of these big firms, right? Every time when you launch a

right? Every time when you launch a product, there will be some of the metrics going up, some of the metrics going down. Right? So you have to make

going down. Right? So you have to make trade-off. You have to define success

trade-off. You have to define success metrics. Success metrics is as long as

metrics. Success metrics is as long as this metrics goes up and all the guard metrics are not breaking the uh benchmark, then you're going to launch this feature. Right? If this success

this feature. Right? If this success metric is neutral, then that depends on other metrics to define whether um whether you launch this feature or not.

If the success metric is negative then you for sure are not launching this feature right that is what we call success metrics. Um and within success

success metrics. Um and within success metric uh the most important metric is northstar and then we have have signpost metrics which are the other kind of metrics that you want to take uh keep an

eye on and we have guardrail and then it comes to uh prioritization um prioritization here I mean like when you uh decide which product to work on or um

quarter planning h this week is the planning week um so you want to have a clear prioritization what to work on and uh also uh make trade-offs offs when you

launch product and do AB testing AB testing how to design AB testing uh what is the trigger logic uh how do you want to divide the users um what type of

metrics are you uh looking at at sitewide uh scale or at a specific like area uh kind of scale so uh you want to define how do you want to do the

experiments in the AB testing state phase and then it comes to go to market when you design a product specifically for zero to one product it is very hard for you to build awareness to your

users. So, uh whether you want to work

users. So, uh whether you want to work with your marketers uh to do campaigns or uh to spread the words out or you want to work with your business

development uh partner to have um partnerships with other companies to uh drive traffic to your feature, right? To

further improve adoption. So, this is how to go to market and also you will have customer feedback. Customer

feedback is qualitative feedback. So

usually um it is very hard for you to like launch a feature that all of the people will hate you for this. Uh but um it is very common for you to launch

something and some people will hate you for this. So you want to know why they

for this. So you want to know why they have a complaint. Um and this customer feedback is not decisive usually unless from customer some customer feedback you

identify a bug uh or it is just a very bad user experience that you didn't think about it uh during the product design phase. Um but usually most of the

design phase. Um but usually most of the customer feedback are like expected behavior. It's just some of the

behavior. It's just some of the customers might not understand why your product is designed this way or some of them might be confused using your product which is very common. Um, and

when you go through the customer feedback, th those feedback are qualitative feedback. So they're not

qualitative feedback. So they're not decisive, but a lot of their feedback can give you inspiration of new features. So why they're saying like

features. So why they're saying like stuff like this, right? So you want to understand the reason behind the reason uh the reason behind the complaint and that is what can give you inspiration.

And lastly, uh, cross functional collaboration. Um when you work on like

collaboration. Um when you work on like your product as a product manager you have to work with all the uh cross functional partners that we mentioned in different stage of the of the of the

skill set for example your designer your engineers your data scientists by the way this is a trained um in Silicon Valley that product managers don't have data scientists anymore so you guys have

to do analysis yourself um and um marketer business development business development u business people um finance

people legal privacy regulation uh researchers, I think those operations, uh product operations. Um I think those are

operations. Um I think those are probably all the cross functional partners I have.

And this is within your team, right? So

we call this team a product team. And if

you want to work with some other product teams, right, there will be a cross team collaboration as well. So if other people want to build on your product or you want to build on other people's

product, we want to have a concept of product owner. So we have alignment um

product owner. So we have alignment um to make sure that uh we don't break things, right? So um that is a cross

things, right? So um that is a cross team collaboration.

So this is basically the different stages of a product manager's um work cycle. And also previous class we talked

cycle. And also previous class we talked about conservatives, right?

Conservatives always have complaints and sometimes they have complaints not because your product is not good, it's because they're just not happy. Um,

which makes sense because I use LinkedIn every single day and right now I still stuck at LinkedIn and doesn't and don't have a new job, right? So I will have complaint too. Um, so it is probably not

complaint too. Um, so it is probably not because your product doesn't work well.

it's probably just because they don't have the exe expected outcome from your product. Right.

product. Right.

>> Right.

>> So there will be multiple factors to affect their customer feedback. Um and

uh previously I work on smartwatchi uh device >> and a lot of people just simply don't like uh the stats show up on their smartwatch because they're not happy with uh their heart rate. They're not

happy with their like health status um and they will complain about your smartwatch. So uh a lot of complaints

smartwatch. So uh a lot of complaints are like this but there are also a lot of legit complaints and you need to filter out the legit complaints and uh some of the complaints might not seem

legit but if you dive deeper and you will figure out a bigger problem. Uh for

example uh I will this class I will use my uh job a lot a lot so you guys can uh know what is on my top of mind on a

daily basis. So if you go to LinkedIn

daily basis. So if you go to LinkedIn right and if you go to who my profile and this is what I'm working on every day right this page is what I'm working

on. So previously when I first um joined

on. So previously when I first um joined the team and one of the customer feedback or this is not even majority a minority like a lot of people have the

feedback is they realize there's a number here right how many profile viewers you have let's say it is not 3,000 it is 30 and we will get a lot of

customer feedback saying there aren't 30 people in this list >> so at first I was like are you crazy. You count every single person.

crazy. You count every single person.

But then I realized that I am the minority because not a lot of people have 3,000 profile views in the past 90 days.

So if they only have 30 profile viewers or let's give a more extreme way, right?

You only have three profile viewers and you only see two people on your list, you will definitely have problems, right? I will have problem too if I'm in

right? I will have problem too if I'm in that case. So right see that customer

that case. So right see that customer service ticket what you realize is not like of course the numbers don't meet up you probably want to explain more that's why we add a little bit question mark

here to answer why there are profile viewers who are private right um and another problem is if the person viewed you twice we will only show once on the list so stuff like this we can explain

all the edge cases to you to make sense of it but if I really give you 3,000 viewers you won't count every single one so this is also an indicator to tell me

that majority of the users on this planet do not have enough profile views and they have to care about every single one of them. Um so it's you can

understand a lot uh about the um the implica uh implications of their complaints.

>> So you finally went up with giving all the people's uh profiles or you went up with the people who have a lot of profile views.

>> You have to earn your profile views. So

to uh tell those premium members how to earn their profile views. Um but yes, we do have a lot of initiative to help premium members to get more profile views.

>> Thanks.

>> No problem. Um it is always a very um complicated journey to analyze users uh complaints uh user complaints or all kinds of like market research, UX research, all of

these are qualitative feedback.

Qualitative feedback is extremely hard to an analyze. It is not as straightforward as data. So it very depends on the product manager themselves to define how to deal with

all the qualitative feedback. A the same qualitative feedback might mean different things for different product managers. Uh for example

managers. Uh for example when um during our like consumer product manager offsite half a year ago uh we

did some user interviews just during the offsite and one of the user was saying that um he doesn't like there are too many uh AI generated content on LinkedIn

feed and when I heard about this I was like my first intuition is um

yes you need to um you need to uh to do critical thinking when you look at customer feedback but also I would always believe that customers are are

always right so if majority of users are using AI generated content on LinkedIn it says something to you right again previously we talked about it if user have a need they're going to to think

about how to use it anyways how to use your product for their need anyways there is no way for you to stop the user need. So if most of the people are using

need. So if most of the people are using AI to generate content on LinkedIn, it says something to to me. Why aren't they not using AI generated content on Instagram? Because they see Instagram as

Instagram? Because they see Instagram as a more personal social media. They see

LinkedIn as a more professional social media. So they're more cautious about

media. So they're more cautious about what they're sending out there. They

want to make sure that their content is well written, right? And also they feel more comfortable to use AI tool in their professional life than their personal life. So uh it can tell you a lot of

life. So uh it can tell you a lot of insights. Well,

insights. Well, >> so like which content content is getting reach on AI content or people who really write the content.

>> So first if uh that is your question then we have to identify what is a AI content. So we can only tell whether

content. So we can only tell whether they use the AI tool on LinkedIn whether it is a AI content or not. If you copy paste chat GPT to LinkedIn then it is

extremely costly for us to identify whether it's a AI content or not. Right?

So if you're saying that our algorithm has a preference in AI versus nonAI content first we have to identify whether it is AI content or not.

>> Right.

Got >> it's very costly is what I'm trying to say. But uh but you guys have changed

say. But uh but you guys have changed the algorithm right because since past three or four weeks I see relevant content getting more views like there are content which was posted two weeks 3

weeks before is now coming onto my feed before it was not it was more about recent I don't know they just have >> well you got to figure it out more uh

and yes we are doing experiments all the time um do you know how many experiments I'm running right now and I think I already closed my quarter. Um,

but I still have like 10ish experiments going on. Uh, before I closed this

going on. Uh, before I closed this quarter, I mean probably half a half a month ago, I probably have 30 experiments going on. So that is only

like product manager and there are um how many product manager do we have? I

think at least we will have 500 product managers. So yeah, we have thousands of

managers. So yeah, we have thousands of experiments every quarter. Um, so yeah there >> I'm also B2C product. I can see if you're running 30 experiments how a day

would be like like your bandwidth would be always be choked like >> well uh it is going to be you guys

future drop so um who is laughing now so uh let's go through this um thinking and to use my product as an example

right uh if we talk about from strategy perspective um what is the trend right now Right? Uh

the job market is not good. The economy

is not good. Uh we are just talking about jobseker now. Okay. There are a lot of audiences out there um target users out there. For the simplicity of this task, we just talk about jobseker.

Economy is not well. The marketplace is not balanced. There are way more

not balanced. There are way more jobseker than employer than positions.

So the job seekers are extremely active on LinkedIn. Uh but they're they are not

on LinkedIn. Uh but they're they are not getting their expected outcome.

uh aka a job from LinkedIn, right? So,

uh that is the job seeker and for a lot of what we call careers accelerators, they want to accelerate their career. Uh

but also the economy is not good.

Companies are laying off. You probably

already have job safety issue, right?

And at the same time, you're not getting as many promotion as previously. So, a

company doesn't really have that many headcounts or opportunities for you to get promoted. So, you probably feel very

get promoted. So, you probably feel very defeated uh at work. So, this is the whole trend, right? economy, job market, everything. And then now we are talking

everything. And then now we are talking about job seekers and job seekers job to be done is to get a job is to find a job on LinkedIn. What are their pinpoints on

on LinkedIn. What are their pinpoints on who view my profile, right? Previously,

let's say if you only could see who view your profile. You can't generate insight

your profile. You can't generate insight from them. You don't know who are hiring

from them. You don't know who are hiring managers. You don't know who can give

managers. You don't know who can give you an opportunity. You don't know whom to reach out to, right? Also, all the recruiters viewing your profile, you can only see how many recruiters view your profile, which is, huh, hate LinkedIn

for this clickbait uh for me to subscribe premium for that. Um, it's not helpful. You only know like three uh

helpful. You only know like three uh recruiters view your profile. How are

you going to take next actions for that?

Right? Uh do you know that all the recruiters are uh the jobs that you just apply for all or they're just like random recruiters that you will not be interested in, right? It's not helpful.

So these are the pain points. Um users

are not uh getting relevant insights from who my profile as a job seeker.

People who can uh cannot take next actions um as a job seeker or who made my profile. Uh they cannot make sense of

my profile. Uh they cannot make sense of who view their profile, right? What is

the solution? We can tell the users where the recruiters come from, right?

What companies those recruiters come from? That might be a very relevant

from? That might be a very relevant insight for job seekers. Whether those

are the job that they already apply for, meaning that, oh, the recruiter saw your resume and they come to see your profile or those are just random recruiters outreaching and reach out to you or not,

you probably still want to check out their positions, right? You're just

curious why does this recruiter or this company interested in you? Maybe you

have a side of your resume that you didn't pay attention to, right? So we

could show the company information um to the profile owners as well.

Prioritization I won't dive deeper into it. We can talk about multiple solutions

it. We can talk about multiple solutions and do a prioritization but northstar metrics for me will be premium revenue right so no matter what feature I design

on the premium side or on the premium side I need to drive premium signup or I need to drive premium retention. Premium

retention is usually you build a really good value for the premium member to stay on premium, right? And specifically

targeting jobseker is a good thing for us. Why? Because a lot of users uh a lot

us. Why? Because a lot of users uh a lot of premium users at least applied for one job. So you know that no matter they

one job. So you know that no matter they are like real job seekers that they just want to land a job within a couple next couple months or they're casually browsing, they are interested in the

recruiter information, right? And on the other side, job seeker are also the type of user segments that are very easily to turn from premium. Why? Because they

subscribe premium, they land the job and they turn. Um they're very goal

they turn. Um they're very goal oriented, right? So in this case, you

oriented, right? So in this case, you probably want to build more values to let the job seekers stay on premium. So

this is how we usually do prioritization. Based on these success

prioritization. Based on these success metrics, you have an estimated impact.

And in this case, the estimated impact is huge. Um I think we're going to go

is huge. Um I think we're going to go for it. And um and then the MVP design

for it. And um and then the MVP design maybe um eventually I want you to directly land a job, right? But right now I just want to

job, right? But right now I just want to see whether you guys job seekers are interested in where this recruiter come from. So I will build that first and

from. So I will build that first and then I will build maybe associate with a hiring manager information or give you AI drafted messages for you to outreach

um or directly tell you that which job you should be um suitable more right so these are the later stages but MVP I'm just showing your company information

of course this is a simplified way of how I analyze this project this is a marketplace right so it is not only jobsekers If you want to show the company information, you got to work with the recorder team as well. So

whether the recorders want to show their company information, right? Um what type of companies want this type of um visibility?

These are also the problems that you want want to solve. Um whether you are giving the recorders the right to change their setting or are you giving the

admins of the contract to be able to change their visibility setting, right?

How do you want to manage the um uh control the uh the setting control on the recruiter side because that is a B2B SAS thinking process right so B2B SAS is

different as how consumer product managers think about this problem are the recruiters feel comfortable um will the recruiters feel comfortable with this setting change right if we have a

default setting how can we communicate with the recruiters better to make sure all the recruiters know that we are changing this feature right so all of these have to be included in MVP because

there will be legal concern there will be uh like the enterprise concern is way bigger than the consumer side. Okay, if

price complain about us and then it is way more than one consumer complaint about us. So you want to design both

about us. So you want to design both sides well right success measurement we already talked about go to market is basically how do you want uh the members the free members premium member free member and recruiter site know about

this new feature right after you analyze the whole thing you maybe come up with a solution like this right today if you subscribe premium you can see who are

the like where are the recruiters come from you can have a list of company uh information and you can check out their jobs as well there are so

things that I haven't tackled in this MVP. For example, if you're seeing like

MVP. For example, if you're seeing like the fourth recruiter uh information in my list, which is a staffing and recruiting company. If you click into

recruiting company. If you click into view jobs, there will be no jobs because that staffing and recruiting company is working for someone else, right? So,

that is not a a helpful information.

That information will not be um helpful for you, right? Because who that who is that staffing company working for, right? So this is not included in MVP.

right? So this is not included in MVP.

The product is not perfect. I will

probably tackle it in the future um but I'm using this example um to tell you how to go through the whole thinking process. So there are five types of

process. So there are five types of product managers and these five type of product managers are um

are divided based on the job the skill sets they have. So this is not from me.

Okay. this division or this like concept is from previously uh there's a uh Lenny's nutter they uh I think he interviewed a product manager at OpenAI

and I think this definition is very um helpful for me because my previous experiences cover all of five um so I like this concept okay and maybe some other product managers say some new

concept that I haven't tackled yet and it is not a strain for me in the job market and I don't like that concept but I like this concept okay so consumer Product managers are the type of product

managers that um for example design the value on holding my profile right uh they're just thinking about what their users need who are their target user how

do you help their lives e make their life easier consumer product managers also have a bunch of principles so for example there will be a lot of things

that you could do on your surface but you don't want to do right so for example Um, right now if you go to your

hu my profile, you'll probably realize um let me show you this. Oh, my profile is very hard to show this. But let me try if I make the number a little bit

smaller. Let's just say the past 14

smaller. Let's just say the past 14 days, right? All the uh full public and

days, right? All the uh full public and semi-private users are here.

But let's say oh this one actually I feel there is a very good example you see someone on LinkedIn here right and if you think about this this

information is not helpful to you at all and why this information is there this person is not full private because if that person is full private at the end of the list you will see a

aggregated card saying how many people view their profile in private and that's why I I use the shortest time duration to help myself to go to the end of the

list. But um since yeah, 460 people view

list. But um since yeah, 460 people view your profile in private, right? So that

person is not a private person. And if

I actually keep his name there, I probably can get drive more profile views. We probably can give users more

views. We probably can give users more insight. We probably can make more users

insight. We probably can make more users stay on premium instead of just being like irritated by why are you showing someone on LinkedIn on my list? Well,

this list I just only want to look at people who are public, right? Why do we keep the someone on LinkedIn there?

Because this is a spammer and LinkedIn as a platform is trying to protect you from not reaching out to that spammer. We think that person is

that spammer. We think that person is suspicious and because how our infra is structured that person is just suspicious very recently and that's why we haven't put it in the aggregated card

yet. Um but given that if I show that

yet. Um but given that if I show that person's name to you I probably can drive more downstream metrics right but I'm I'm not driving it because trust and

safety is the most important principle on our platform. That is how consumer product managers think because metrics are not the most important thing.

Principles are very important. User

experience is very important. Um and

consumer product managers are usually um type of product managers that uh are datadriven but they are more like an artist. Um they have a lot of feelings.

artist. Um they have a lot of feelings.

Okay. Uh they they care about the user experience a lot. They will have a lot of opinions about the their designer's design. Um so that is the first one

design. Um so that is the first one consumer product manager.

>> Uh I have a question. Yes. Um but as a job seeker I don't know the meaning of like someone from like Microsoft. I just

think like LinkedIn did didn't give me enough information.

So I cannot get the uh the design of this part.

>> That's a good feedback. and

>> and if it was like uh someone at Microsoft rather than uh HR at Microsoft or product manager at Microsoft that

would be the oh I guess product managers are we something like that >> that is the right mindset to be a

consumer product manager on my product guys you guys are right okay but I'm not a a consumer product manager mindset on my product. So now I will tell you how I

my product. So now I will tell you how I grow my product as a growth product manager.

>> Oh, are you actually a growth PM?

>> Yes, >> I was a growth team. Damn. Thanks.

>> So most of my time at LinkedIn no matter right now working on subscription or previously I work on user growth. My

mindset is very growth product manager and my manager, my skip manager, my skip skipping manager love me to work on who in my profile with a growth product manager mindset. Why? Because who in my

manager mindset. Why? Because who in my profile and profile appearances and all the premium analytics are the major premium revenue driver. We previously

talked about this. Uh premium as a business is $3 billion. All the

analytics product is probably half of it. So it's $ 1.5 billion on these

it. So it's $ 1.5 billion on these analytics products. So it is a huge

analytics products. So it is a huge revenue on stake and also the email and notifications is our huge daily active user driver. Uh our notification

user driver. Uh our notification click-through rate email clickthrough rate is the highest across LinkedIn. One

of the highest across LinkedIn I think it's top 10. Um so given that all the metrics contribution you have to have someone with a huge go growth mindset to

work on this product otherwise it's very hard for you to get the full potential of this product. Now let me tell you what is a growth product manager mindset.

A growth product manager mindset is what is the metric you want me to drive.

You want me to drive premium signup, I will give you $100 million this quarter.

You want me to drive premium retention, I'll give you $100 million from retention. If you want me to drive daily

retention. If you want me to drive daily active user, I'll give you 2% of daily active user this quarter. Right? The

metrics goes first. As long as I have a northstar metric, I will think about how to reach that northstar metrics. This is

also how our um organization is structured. So premium uh user growth,

structured. So premium uh user growth, both of the teams land under uh consumer growth um or at LinkedIn. So how our OKR

is structured is also like this. So for

user growth team, every quarter you have your DAO OKR, right? So probably this quarter you're going to drive 1% down.

Okay, that power manager is going to drive 1% down. All of you guys add together 5% down. Okay, so this is like top down Microsoft on LinkedIn to drive 5%. All of you guys have to figure out

5%. All of you guys have to figure out how to drive 5%. And

is it important for the leadership to know what projects you are going to build next quarter? Maybe.

But you probably have five projects there and you tell the leadership that you're going to drive 1% out of it and then one of your project doesn't work.

You expect that one project to drive 0.1% doubt from uh out of it. But that

project didn't give you any doubt.

And here is where your growth product management mindset come in. Your OKR

goes first. Okay? There isn't an option for you not to meet your OKR target. SO

YOU HAVE TO THINK about whatever you can build within that quarter to actually meet the 1% down. And it's the same thing on the premium side. It is a very

sales mindset. So that's why I have this

sales mindset. So that's why I have this full course for y'all from the middle of our uh first quarter till the end of this quarter. Right? The whole course is

this quarter. Right? The whole course is almost two months and I started this course with y'all from the beginning of August. Why? because I already met my

August. Why? because I already met my target this quarter. So I can spend more time with y'all. Um yeah, so on the

premium side, every product manager has a a revenue target. So um I have making this up $10 million for this uh quarter.

Some other PMs might have $5 million, $10 million for this quarter, right? No

matter what you do on your product, of course you need to have a little bit consumer product management mindset not to do bad things on your users, right?

But as long as you meet those principles, you do good things for your users, you have to meet your target.

Your metrics goes for first. And also

gross product management are very datadriven. So you tell me you don't

datadriven. So you tell me you don't like this metric. Uh you tell me you don't like this notification, but 80 million other people click into that notification and my notification clickthrough rate is the highest. So I

will probably listen to those 80 million people instead of listening to you.

That's the growth product management mindset. No, I'm happy that you already

mindset. No, I'm happy that you already reached your target.

>> My target.

>> Wait, that's actually crazy. Wait, so

you have like the >> Yeah, actually in in India it's different, you know, like if I reach my targets, they just increase my target for that quarter because it's like >> they did um for the past year, every

quarter, my manager double my target compared with my previous quarter, but I still double my uh delivery uh compared with the target. So, uh, no, I think she just gave up.

>> Wait, do you get like bonuses? That'd be

so cool.

>> JJ just secretly like just makes bank and we just don't know.

>> Yeah, I will trim this uh this part out, but just give you guys a sense of how much uh this senior product manager drive at LinkedIn. Okay, so um for

example, for quarter one, our our target is the highest. Why? Because this is stupid uh how financial year is calculated. um the now we are diving too

calculated. um the now we are diving too deep into business um concept but um that's also why by the way that most of the AI startup B2B SAS companies tell you that they're driving 10 million ARR

that number is wrong okay I will tell you why so there are two concepts one is in-ear revenue one is annualized revenue and the annualized revenue is the ARR

that we talk about okay and the in-ear revenue is how much money actually go into your bank account at the end of this year

and the annualized revenue is after you ramp this feature to 100% how much does this feature drive within the one year.

So if you given these two concepts you will realize a very interesting thing is Microsoft let's use Microsoft as an example. Microsoft is different as the

example. Microsoft is different as the all the other companies on this planet.

Okay Microsoft phys physical year is from July to June. It's not from uh January to December. So right now we're talking about from July to September is the most important quarter we have

because whatever feature you build or launch in quarter one the inear revenue is going to realize the whole year right so rather than you build something and launch something at the end of June you

only probably the money come from your feature in the bank account it's only that one month but if you launch something in July uh on the Microsoft financial report it is

going to realize the whole year so that's why this quarter is the most important quarter uh for a financial year.

Well, now I'm Am I losing you guys? You

guys look at me like you're very confused.

>> Yes. Can you explain it again? I got

confused. Like

>> you can run it back. I won't complain.

>> Little slow.

>> This is subscription one. Okay. And

that's why you can also realize why those startups calculate them wrong. So

um in-ear revenue is how much your feature make till the end of this physical year.

Annualized revenue is what your how much money your feature is going to earned if you launch this feature to 100% um for a year.

>> So like if I launched it like June January 1.

So if you launch the feature in June, your in-ear revenue is going to only be that one month that you ramp to 100%.

Because Microsoft year stop at the last day of June. But if you're talking annualized revenue both like no matter when you launch that feature, ARR will be the same.

>> Did I lose you guys? Give me some feedback.

>> And revenue is a projection, right? like

this might I be able to >> both of them are projections.

>> Both of them are projections because you never know how many people are going to turn from premium. Yeah. So let's say I have a very important feature giving me

$10 million annualized revenue. Okay.

That means that if I have this feature on LinkedIn for one year, that feature is going to estimate it giving LinkedIn $10 million.

But that in-ear revenue and annualized revenue will be the same if I ramp this feature to 100% at July the 1st

because in-ear revenue is going to let my feature realize it for a year.

Annualized revenue calculated for a year, right? So if I ram this feature on

year, right? So if I ram this feature on July the 1st, it will be the same no matter it is in year or annualized.

But if I ram this feature on Druid, in-ear revenue will only be like let's say first day of Druid, then inear revenue is going to be one 12th of the

annualized revenue because annualized revenue is after your feature ramp to 100% how much does your feature earn in that one year after 100%. In-ear revenue is how much your

100%. In-ear revenue is how much your feature is going to earn after it run to 100% till the end of this physical year.

and Microsoft um financial report or any uh company's financial report, they don't care that much about annualized revenue because they only care about the revenue that goes into their bank

account. They don't care potentially how

account. They don't care potentially how much your feature earn after one year.

So that's why we care so much about in year.

Annualize is just giving all of your feature the same starting point. Right

now annualized revenue is just a meth measurement for you to compare between features. This feature is more important

features. This feature is more important that feature is not. But inear revenue is the estimation of how much money going to a Microsoft bank account

till the end of this whole year.

So given said that I just want to say that this quarter is the most important quarter of in subscription and um yeah so they have been increasing my target

every time. So originally when I first

every time. So originally when I first joined the team my target was like $2 million per quarter and um this quarter

I committed $14 million. This is already uh one year after every quarter they'll try to increase my target and uh $14 million I I closed this quarter at the

second week of the July uh and also first week of July LinkedIn is shut down. So I basically prepare all the

down. So I basically prepare all the experiments and rent it rampid exactly after the shutdown. Um, so on the second week I closed this quarter and deliver

$30 million. And right now I still have

$30 million. And right now I still have two experiments holding and ramping uh in the next one to two weeks. And in

total I'm going to give Microsoft $50 million in the first quarter. So yes,

thank you very much. That's how I got promoted guys.

So I'm curious uh like for the growth product manager instead of uh like um being the data driven uh do you also like design some market activities to

drive the growth?

>> Yeah that is um kind of um neutral because it depends on what type of culture your company has. So at LinkedIn or at any social media related companies, uh product management has

product managers has a lot of control.

Um why? Because social media companies need product manager to be very opinionated and uh they need a lot of users uh empathy. They need to understand the user a lot because social

media is very sensitive to a lot of details. Um and that's why for social

details. Um and that's why for social media companies product manager have a lot of say and in that way uh the business development or the marketers uh

or designer or engineers will go with the product manager.

So given said that if you want to go to market for your feature, you probably have a proposal uh to your marketer and tell the marketer you want to do whatever and whatever and the marketer

will give you an actual plan of how they're going to execute this.

But this is only in social media company. In other companies, the other

company. In other companies, the other people might have a bigger say. In B2B

SAS, for example, sales have a larger say than product management uh product managers. So uh we talk about B2B sales

managers. So uh we talk about B2B sales are the people who used to be on work, right? So um they're very tough and

right? So um they're very tough and product managers need to go with their biggest clients and consumer product management or uh a consumer product

manager in consumer social media type of company um the product managers have a lot of say.

your career is always around B2C PM like Tesla and this >> yeah if you uh find a company that you find myself a good fit in it uh please

refer me and thank you very much to analyze my skill set >> but just like curious like do you feel B2B PMs grow faster than B2C

>> define grow faster >> career growth like being into the leadership role like direct >> I don't think there is a difference >> no

>> no I don't think there is a difference >> and even even with the sorry Charlene just one more point um even with the people's impact there is no difference

like for example consumer product managers might have like 10 to 20 experiments per quarter and every experiment drive like $1 million $2 million $3 million some of the experiment might drive you $20 million

$30 million Um that's it and you still every quarter need to have a portfolio because you need to meet your metrics anyways. So you need to have some of the

anyways. So you need to have some of the best that you feel very confident with.

Some of the best are being venture best.

Uh so in general you're driving that much money, right? Um but consumer move faster. Uh consumer also have smaller

faster. Uh consumer also have smaller project because you can divide and concur. Enterprise side is different.

concur. Enterprise side is different.

When I work on LinkedIn ads, I work on a feature for freaking half a year and that feature drive $80 million. So

there.

>> Um yeah, it's just how they work is different and um we will talk about enterprise product management later but um their working style uh the stakeholders they are working with and their thinking process is very

different.

>> Like I'm wondering during your work experience do you have any times you have the conflicts with other PMs like say other people their prioritizations is not the right

>> all the time. all the time, girl. All

the time. It will be a surprise for me not to have a conflict with a PM in a week. Every week there will be some

week. Every week there will be some conflict and it is so so so common. Uh

the third one, business product manager.

Uh let me give you an example of a business product manager. For example,

previously when I work on LinkedIn ads, um as a business product manager, you probably are taking charge of the marketplace of the ads. I don't know how

much you guys know about the AS but AS is a marketplace if you guys don't know and it's the same with Wubber with

LinkedIn drop market with um let's say uh who else um Google Google search results right all of them are marketplace because they're supply and

demand and in ads the demand side is the advertiser side uh the advertiser give you advertisement give you campaigns and you either put their campaigns on LinkedIn in or on third parties like New

York Times or CNN.com, right? Um, and as the product manager for the marketplace, for the ass marketplace, you want to make sure how we bid on ad placement or

like whether to put this ad at this placement to these users are the best result or the best delivered outcome for the campaign managers,

right? this sort of marketplace

right? this sort of marketplace management, you need to think about pricing, you need to think about bidding, you need to think about pro predictions, right? Uh you need to think

predictions, right? Uh you need to think about uh algorithm outcome and these are very businessdriven. You don't really

very businessdriven. You don't really think about how user feel in your decision- making. You just want to

decision- making. You just want to target the lowest cost per click, lowest cost per impression, the highest delivered outcome, no overd deliver. uh

these type of metrics. So you're very metric driven kind of business uh kind of business and you just think about how

to make your business better. Another

example is pricing for LinkedIn premium.

So uh you increase the price, you lower the price. These are pure decision uh

the price. These are pure decision uh business decision making. You don't

really think about oh how people feel when you increase the price. No, you

just think about how much money you get or lose from increasing the price.

Right? So this is business product manager. They are extremely good at um

manager. They are extremely good at um like marketplace supply demand relationship. They're extremely good at

relationship. They're extremely good at analysis at business sense at business strategy. Um it's a business product

strategy. Um it's a business product manager.

And then we comes to platform product manager. I also have an experience in my

manager. I also have an experience in my resume. Huh. To let you guys know what

resume. Huh. To let you guys know what is a platform product manager. Um

besides of all the analytics product I work on at LinkedIn, I also is responsible for a technical foundation a data infrastructure platform. Um and

this data infrastructure platform supports all the analytics products across LinkedIn including creator analytics, page analytics, uh of course home profile, search appearance and

stuff. Um so we want our partner teams

stuff. Um so we want our partner teams to have an experience of plugand play.

So they don't need to design different graphs themselves. they don't need to

graphs themselves. they don't need to think about oh which data set should I get the this data from? This is a unified solution for multiple partner teams to like design their own product.

So as a platform PM um do you do all the things that your partner team asks for?

Definitely no. But if all of your partner teams are asking for the same thing, you better do it right. So uh you want to look deeper into their problem

to see what is the core issue that caused the problem and then you want to see whether by solving this core issue I could enable a a lot more business opportunities for other partner teams

because you don't want to build 10 features on your platform that doesn't matter that don't matter you want to build that one single feature on your platform to make your platform simplified to enable more business opportunities that's platform product

management research research product management. I

briefly have been a research product management when I was founding my first startup from 2016 to 2019. That was a very researchy product because the only job that I'm doing during the maybe

three four years is to find the product market fit to find the right problem to solve using this algorithm. So the

research is already there. My

researchers are designing the best computer vision algorithm in China. But

me as a product manager need to find a use case for that. I also have been a research product manager when I work at Tensson AI lab. It's the same thing. I

they have a research they have paper to publish. they have a study that they are

publish. they have a study that they are um um researching on and I need to find oh maybe for that gain we can use that algorithm for that use case and that

probably is something that the users might want and also research product managers need to do a lot of assistant work as well because researchers are the most important people in this world I

also have a mindset of even if you are like product manager like common product managers engineers time value a lot um So I think this is a mindset that I want

to give you guys and a lot of product managers by the way don't have this mindset like your time as a product manager isn't that valuable the engineers time are the most valuable

people why because if you think about from a company perspective who got paid more per hour right so you want to save the company's time um so that's why a

lot of your job is to try to save the engineers more researchers time um whether to draft a better PRD so the engineers don't need to go back and forth with you to confirm what exactly

do you mean by that requirement or on the research side you want to do a lot of things for your researchers if you have the time like cleaning data data

management um labeling data and uh sourcing data mining data right uh these are the type of job that you want to do for your researchers as well I know there are a lot of questions about

growth product management and Um I want to here talk about the differences between product marketers and growth product managers because a

lot of people don't know the difference.

Product marketers are basically doing two jobs. One is user research and the

two jobs. One is user research and the other is product go to market. So you

can think about this as before you design a product you want to have insights from your marketers. After you

develop the product you want your marketers to help you go to market. And

of course you could have maybe you do a little bit more, you do a little bit less. The boundary is always very uh

less. The boundary is always very uh blurry. Um as long as you guys are happy

blurry. Um as long as you guys are happy working together uh is fine. So um yeah, but product manager uh product marketers are doing campaign like or survey like

research like right and product like growth or growth product managers they do features on the product to increase um certain metric uh or grow your

product. So growth product manager if I

product. So growth product manager if I want to drive a metric growth product manager works on a feature to drive that metric. Product marketer does campaigns

metric. Product marketer does campaigns or enables that metric through various channels.

>> Yeah. So um use a very simple example. I

want to drive uh new user acquisition for LinkedIn. Um my job is probably

for LinkedIn. Um my job is probably build a bunch of thirdparty loginins on LinkedIn. uh or I could think about how

LinkedIn. uh or I could think about how to improve uh the indexing of LinkedIn website to make us rank high rank higher on Google right that's a product

manager's job a product marketer is now they have a lot of paid campaign on Tik Tok on YouTube on a lot of other platform to tell those people through ads that use please use LinkedIn

>> okay but even uh retention would be part of the product marketer Right? Because

>> it depends on through what channel.

Let's say uh if you want to drive retention through product features, you want to recommend 10 really good people that that person can connect with on LinkedIn. It's a product manager's job.

LinkedIn. It's a product manager's job.

If you want to drive retention just by sending an email to the user, let's say right now you want to send one email to the user to remind them to come back to

LinkedIn. You can use product marketer's

LinkedIn. You can use product marketer's job. um because it is not a product

job. um because it is not a product email, right? If you want to write a

email, right? If you want to write a well programmed email every week to that user to remind that person come back, that is the product manager's job.

>> Um how do you decide which approach do you want to go to? Whether it's through like >> same question.

>> Yeah.

Uh what do you mean what approach like from email side >> like the way you want to grow the engagement or retention?

>> So um a very easy way to distinguish how you want to spend uh the effort is marketers are usually a very exploratory uh journey. So marketers just have

uh journey. So marketers just have campaign themselves right if the campaign works then you probably want to productize the whole campaign. um that

is through like email notifications. Um

so marketers want to try it out. They

want to send an email reminding you of all the premium features and see whether it works. If it works then I'm spending

it works. If it works then I'm spending more engineering effort on it. Again

engineers time are the most important people's time. So you marketers or

people's time. So you marketers or product managers you guys are trying to have a really rough MVP to test the idea out in order to save the engineers time.

You don't want to spend your engineers on this for a quarter and then it doesn't drive metrics. So you can let the marketers go f go first and see whether it works from a campaign and if it works you productize this whole

thing.

Marketers a good marketer always have a lot of cheap ways of doing things and they could always like help you test

out a lot of stuff. Um, for example, um, I have a really good marketer on the user growth side that always just twist a little bit on my email and

notification copy and she just like do experiment herself and give me free money. A lot of times like after her experiment, it drives so

much money I need to revisit what copy she has.

And I think a lot of times she really grabbed the user psychology uh really good.

>> So you when you build a feature the messaging part the positioning part is done by a product marketer.

>> That one I think you want the product marketer's input. Um so it depends on

marketer's input. Um so it depends on how your team uh how your company is structured. For example at LinkedIn we

structured. For example at LinkedIn we have a internal platform called Chameleon. So basically all the

Chameleon. So basically all the non-engineering people including product manager or product marketers could change any copy exist on this product to do AB testing and you don't need

engineers involved.

Um, you guys feel confused. Let's say

notification copy, right? Um, is three people view your profile, see details the best copy for click-through rate?

Probably not, right? And your marketer probably have 10 more options for you.

The marketer doesn't need to go through engineers to help her to make the to uh to do the AB testing. She can just use this platform to do AB testing herself.

So, in that way, the marketers can be very self-s served, right? But a lot of other companies probably don't have that. So they have to have engineers

that. So they have to have engineers involved and everything. Um so as I said the boundary is blurry but usually marketers are very exploratory and they

do campaigns instead of product um stuff. Um in terms of how you message

stuff. Um in terms of how you message your product, you want message um marketers input. Um and also it depends

marketers input. Um and also it depends on how established your company is, how platformized your company is to see whether the marketers can do copy

testing themselves. Um but regardless of

testing themselves. Um but regardless of product development uh user research and the product go to market are like very heavy on product marketers.

>> They take approvals from you for the messaging or like you are open like you can freely >> it also depends on the company culture at LinkedIn again the product managers have a lot of say in products. So pretty

much every single thing on the product goes to product managers. Um

>> okay >> but also I trust my marketers a lot.

Later we will talk about uh I trust uh my cross functional partners a lot a lot and a lot of product managers don't do that. Um so I pretty much just look at

that. Um so I pretty much just look at whether like out of the 10 copies whether there is one single copy that I know it will have negative impact. I

would just tell her if there is no then I'm open to test. Um, yeah.

>> Uh, I'm interested in the idea you say like productize a marketing campaign cuz that reminded me of like Dolingo. I was

wondering how to see their marketing uh like campaigns. Do you think like those are already like part of their product or still like uh like marketing?

>> Not part of their product. Like an

easier way to see this is whether you see there is engineering team involved.

You could think about product manager's job in this company as how do you spend your engineering effort better right if there are engineering team involved then yes product managers need to make a

decision whether you want to spend your engineers on it um so back to Dolingo's case if it is like social media right that's pure marketer that doesn't need

engineering team involved that doesn't have R&D um if it is email then marketers probably try it out if it works then engineering team involved and productize

Um so it's pretty much just um see whether there are engineering team involved uh or it involve the product decisions but it again it depends on the company

culture a lot like Google it is very engineering driven culture so uh product managers do not have as much say as product managers at meta and LinkedIn um if you meet the software engineers at

meta or LinkedIn they will always complain about their product manager and of course every company complain about their product managers But these two companies especially we complain about the front managers because their product

manager make a lot of decisions. Uh some

examples previously we talk about improve who my profile for job seekers right uh how I analyze the whole thing I was using a very consumer product

manager uh mindset for you and if we are thinking about what is the job to be done what is the problem users have how to solve the user's problem better to help them land the job better right um

here I give you an example of how a growth product manager think growth product managers think about two things.

One is how to increase top of funnel. So

you probably want to have either marketing uh campaign, BD relationship uh whatever or um building like your entry points on a bigger surface

whatever right uh to increase your top of funnel and then other than that all of the funnel um is focusing on conversion. So how can you

make the conversion better? How can you make uh more people from top of funnel to become your real users?

So um for example users coming from app store using users come from SEO from Google user come from getbdt right all of those sources uh whether you can increase new sources or you could

increase um new traffic or how to you uh how to increase the conversion of your existing funnels. These are like how

existing funnels. These are like how growth product managers think, right? Um

what are the conversions? Why is the conversion high or low? Is there member need that we didn't satisfy? Right? If

the users forget their password, um how to give them a easier way to log in.

Right? If the users log out, uh how we help them to remember them next time when they log in. Right? These are the ways to improve the conversion um

conversion rate on the users's um on the user's sign up or um the subscription signup questions.

So I think growth product manager is like the entry level growth product manager is a very easy job. It's usually

just like focusing on how to improve the funnel and how to improve top of funnel.

But if you dive deeper um into like higher level um growth job, you probably can close the partnership that other people cannot close or you can have a

better negotiation power over your partners. Um or you uh know the user

partners. Um or you uh know the user psychology so much that you can have like uh you can have uh like the best proposal uh on the marketing campaigns

with your marketers, right? A proposed

product manager is everywhere. Um so

they're very um sensitive on how the user uh feel and think in every single step of the conversion and they can exactly target your uh thinking and

convert you. So another example is once

convert you. So another example is once the user sign up for LinkedIn right how to re-engage the users is also a question if the user only come to

LinkedIn once then what are the reasons they are not coming back so the growth program manager will think about exactly the reason that they're not coming back maybe they're not looking for a job they do not look for their career growth they

probably care about something else how can LinkedIn satisfy their new uh problem and give a new value prop to convert them Right? Uh what are the channel that I use? What is the messaging I want to send out to the

users? Um these are the thinking process

users? Um these are the thinking process for growth product managers and growth plan has to be very good at data.

Yeah. So in my job uh this is my uh famous or notorious email and uh if you um are a growth product manager and you look at this email then you think about

wait if I want to uh improve who will my profile page view let's say right improve who my who my profile premium to pay to pay sign up the first job for me

is to improve top of funnel of that page right how to improve top of funnel half of our traffic come from email notification is it bigger in email is it bigger in notification Which one have the higher potential? Right? Let's say

just email. On this email, you can realize that it is a very simple email.

One thing is it only has one number, right? It doesn't tell you who viewed

right? It doesn't tell you who viewed your profile. It doesn't have any other

your profile. It doesn't have any other information and it only have one CTA.

All of these are specifically designed for a higher conversion because we don't want to provide two CTAs to make people feel confused. On this page, where to

feel confused. On this page, where to click at is very obvious.

Okay. And also the title um need to be like AB testing a lot of times maybe adding your name there makes you it feel more personal um maybe the copy you are

getting noticed feel more personal than you have whatever profile viewers right so all of this copy change when to send out this email whom to send out this email how this email is designed whether

this email can fit in your phone's first screen all of these matter for your conversion Right? And then this is an

conversion Right? And then this is an example of a um business product management, right? If you want to

management, right? If you want to increase LinkedIn ads, deliver revenue, then the questions that you want to think about is what is the marketplace?

How is the supply demand relationship?

What are the difference uh different objectives of campaigns? Um if you guys haven't done ads um so different campaigns have different objectives. So

you're optimized for impressions, you're optimized for reach, you're optimized for clicks, you're optimized for website visit, you're optimized for like uh install of your app, uh you're optimized

for um u buying your product, right? Uh

on the whole funnel, there are a lot of uh different um different points that you want to target. Um for those who want to target reach uh or impression,

those are brand marketing, right? uh for

those who want to uh target the downstream uh clickthrough or downstream conversion uh you have different objectives. So what are different

objectives. So what are different objectives for campaigns? What are the cost per impression, cost per reach, cost per clicks, couple per website visit? What are the CTRs? How are

visit? What are the CTRs? How are

different AI algorithm perform? Accuracy

loss. Is there new data source for us to improve performance? Is there new model

improve performance? Is there new model for us to experiment on? And what is our current margin? Should we raise the

current margin? Should we raise the price? Right? These are the problems

price? Right? These are the problems that a business product manager will think about. And then this is the

think about. And then this is the platform which is also uh the example that I used before. Build a LinkedIn data infrastructure platform. The value

of the platform team is always enable partner teams to do their job more effectively and efficiently. Plugandplay

experience. Uh interviewer partner teams understand their strategy, vision, problems to be solved, their road map, observe how they use the platform and find common patterns and improve the

platform foundation.

This also works on like this uh here I use data infrastructure platform. There

are also other platform like for example meta or LinkedIn or Google these type of big tech companies they have internal experimentation tools right those tools also have product management product

managers they have those tools in house inhouse build so those product managers will follow up with like product managers data scientists how to use their platform uh how I can improve the

platform to make you m your job easier um those are the platform PMs and lastly the research PM what is the research team capable of Is there anything I could do to faster the research? Is

there any possible application to the research? Are there asks from partner

research? Are there asks from partner teams? Uh prioritize the asks for my

teams? Uh prioritize the asks for my researchers for product execution question is basically like all the skill sets that we previously mentioned um in the product execution, right? So success

measure, prioritization, AB testing, go to market, customer feedback, cross functional collaboration. So here um

functional collaboration. So here um usually the type of question that you get is um like define the success metrics of a certain product of a certain feature uh goal setting uh a

metric might go down a metric might go up uh you might want to uh like make decisions based on your metrics fluctuation uh and also evaluation trade-offs right there are two product decision decisions just get back to what

we previously discussed on the product design question uh you want to have shorter billing cycle or you want to have discount on your first subscription Right? Um, some measures might go up,

Right? Um, some measures might go up, some measure might go down. Revenue

might not be your only criteria to evaluate which uh solution is the best.

So, you probably want to evaluate trade-offs and make trade-offs um uh to uh decide which feature to launch.

I will uh put this into following courses uh if we have time and also the examples too. So uh also another type of question is diagnose

which is a very uh common thing that you have to deal with at work uh and also if your product is very important. So if

your product is important the diagnos is called a war room. Okay. If your product is not important or if your metrics uh up and down is not significant that is called an investigation. Okay. So we

don't really want to escalate an investigation to a war room. That is

very chaotic. Um and a lot of leadership eyes on it. A lot a lot of um like execution involved in this but basically a metric down uh or up um you want to

figure out the reason and that is basically how we started a war room or investigation and in a lot of execution like interviews they will also um evaluate you on whether you could f um

what is your thinking process on diagnose questions.

Okay. How to write a PRD? So first of all add everything you mentioned before right your strategy your target user job to be done your pain points your solution prioritization as uh MVP design

success metrics and go to market why because you want to also tell your cross functional partners why you designed this certain feature out of all the options you have why is this feature why

do you want to design this solution right you want to communicate all the context with your cross front partners the better your PR is the more offline back and forth uh there will be and make

sure to cover all the edge cases so your engineers and designers don't feel confused. How to make sure AI coding

confused. How to make sure AI coding tool is a good way to test. So if you are able to input your PRD directly to a AI coding tool and it directly come up with the product that you want that

means that you explaining yourself pretty well. Um so you could use AI tool

pretty well. Um so you could use AI tool today to make sure that you understand the current uh implementations. uh maybe

you just throw your code into AI tool to help uh to like explain different use cases for you or you can use AI tool to make sure that even AI understand what

you said. Okay. So PRD is for cross

you said. Okay. So PRD is for cross functional communication and scaling impact.

If you only have two engineers one designers work on your product then you have a meeting you can explain everything as well. But if your product feature right now is affecting uh five

teams, different teams and different orgs are looking at your PRD to understand what are you doing, you want to better have documentation or the 100 people are going to chase you down. So

the PRD is for the simplicity of communication.

If you want to write a PRD using AI tool, you can use my GPT. Um, another

point that I want to talk about is a good product manager is the center of the team.

A lot of your cross functional partners trust you. They trust you to make the

trust you. They trust you to make the decision. They trust you to be the

decision. They trust you to be the domain expert to give them a direction to give them a trustworthy strategy and a good solution to help the team make

impact on your customers. Right? That's

a good product manager. It's the center of the team.

A great product manager decentralized the team. If you are centralizing, if

the team. If you are centralizing, if you are the center of the team, you're the blocker. Okay? And you are one

the blocker. Okay? And you are one single person. You want by adding all

single person. You want by adding all those contacts in your PRD and in your strategy, you want to make sure that all of your cross functional partners are on the same page with you. If they have a

question with your logic, they can point it out. They can challenge you. You want

it out. They can challenge you. You want

to make them feel comfortable to challenge you. You want to educate them

challenge you. You want to educate them how you think about the whole process.

So if there is a very little decision to be made by engineers or designers, they can make the decision themselves. They

don't need to reach out to you by simply what color is this button or how big this button should be, right? These are

the decisions that cross functional partners shouldn't make by themselves.

They also as you become more senior, you realize that they also could make a lot of product decision themselves. A lot of times my engineers feel like oh this is something I want to fix. I think there

is a gap. Sounds good. Sounds good. Go

fix it. Right? Like the only response from me is sounds good. Uh they can make decisions themselves. So this is how you

decisions themselves. So this is how you can decentralize the team so the team can move faster and you can scale your impact better.

And also different crossunctional partners have different skill set. Then

if your crossf function partner are not good enough then that's their manager's job. Okay. So tell their managers how to

job. Okay. So tell their managers how to help them to become their better self.

But in the ideal stage you need to trust your crossfunnel partners. What do I mean by this? If there is a design provided by your designer you don't like their design. You better have a very

their design. You better have a very good business reason why you don't like the design. If you just don't like how

the design. If you just don't like how the button look like also tell your regions as well. But you need to trust your customer partners to be the expert of their own domain. If you are trying to do your designer's job, your

engineering's job, your bisop's job, you are again make being the center of the team. It is very hard for you to scale

team. It is very hard for you to scale the impact. You can probably just work

the impact. You can probably just work with five people maximum. But if you help educate your crossunctional partners to be the experts of their

domain and also they know how you think then you can scale your impact. So this

is crossunctional collaboration.

some problems for PMS. First one, research bias. We talk about market research. We talk about user uh

market research. We talk about user uh experience research. The whole research

experience research. The whole research uh process is a collaboration between marketer and product manager. Especially

for zero to one products, product managers have huge hypothesis. You don't

have enough data to support your hypothesis. So you want to validate

hypothesis. So you want to validate those hypothesis through market research. The whole process might be

research. The whole process might be recruiting candidates, questionnaire, interview, right? These are different um

interview, right? These are different um ways to get your research done.

Most of the research on this planet are directional and are very biased. The

nondirectional and not biased research are very rare. Those are really good research. Why? Because usually it's the

research. Why? Because usually it's the product team themselves conducting those research. If you have a centralized team

research. If you have a centralized team let's say usually marketers are not centralized but researcher might be centralized then they probably have the better nonbiased way to conduct the

research but a lot of product team when they conduct research they will literally ask the user do you want that feature?

Of course users will want that feature.

You don't increase your freaking subscription price right? So why not? Um

users will always say yes. So these are the directional questions that you are basically trapping your users into a hole and they will exactly give you the answer you want. That's a very dangerous

thing. Instead, you should never ask

thing. Instead, you should never ask your users whether they want one thing or not. You should not direct them into

or not. You should not direct them into your hypothesis. You should just observe

your hypothesis. You should just observe your users how they use the product or observe the users how do they solve their pain points today. and you have your own understanding of why um this is

happening and ask them to validate like why are you doing this right why are you doing that to understand their behavior and their thinking process behind it if a user tell you what feature they want

just ignore it okay just ask why instead of really focusing on what features they want um so you want to observe facts you want to observe how they use product you want to observe how they live their life

instead of asking them questions whether you want that product.

So ask more their user journey and thinking. Um ask why they do what they

thinking. Um ask why they do what they do, not ask what they want. And also

another reason why you cannot trust users is users lie a lot. Why? We

previously talk about it, right? Users

might talk about their feedback, might talk about their feeling not because of your product, might be because of something else. Um they probably talk

something else. Um they probably talk say certain things um have like other reasons um for saying certain things, right? Um if you're saying that oh um

right? Um if you're saying that oh um your feedback might be like you're not getting enough profile viewers right no user want to admit that they don't have enough profile viewers uh all of the

users want to think that they are very popular on all the social media so um this is just a human nature they want to look nice they want to um feel nice so they say things that they don't mean uh

or not valuable for your user interview so um for product managers uh psychology means a uh you always need to ask why why why people say this why the user say

this why the user do this thing in this certain way we also previously talk about datadriven versus experience driven there is no right or wrong in this most of the

companies are datadriven because if you're experienced driven then there are just numerous escalations and debates happening you feel it this way I feel it that way it is very hard for you to quantify what exactly is the path has to

go. So most of the company are data

go. So most of the company are data driven if they have enough data but experience is the fundamental uh quality product quality right is the fundamental

principle of the product you don't want to blast your uh users with notification and emails right um I could send you all the like notifications in the world 1%

of you click in I would say good I have more doubt the quality is too bad right I could give all the people who open LinkedIn app a premium upsell, right? A

a full page premium upsell. I don't do that because the quality is too bad. So

datadriven and experience driven is a balance.

That's also why B2B products are so hard because a lot of times B2B products do not have enough data. If you only have several you several customers, enterprise customers, you don't have

enough data even for like size of LinkedIn, right? If you're doing as

LinkedIn, right? If you're doing as product a lot of detailed features uh that you launch probably not not giving you static results so you don't have data to support your hypothesis you have

to conduct research inter uh research interviews right those qualitative data those qualitative feedback matter more your sales people tell you these are what the users want you have to take that into consideration you don't have

the enough data to support your hypothesis so data driven and experience driven is a balance This is another thing to tell all of you guys to keep uh you guys up to date how

we use AI today. So in big companies we use AI for coding. We already don't have data scientists uh for product managers we use data AI data scientist um we

draft PRDS using AI we draft emails using AI we do copy recommendations using AI. So um originally a lot of

using AI. So um originally a lot of email campaigns where notifications are having like thousands of copies and we just use supervised learning to choose which copy work for which users but in

the future there will be more copy recommendations drafted by AI. Um so

this is the trained and the workplace uh trend is going and what are the roles will be replaced by AI. Huh I don't think these are uh

by AI. Huh I don't think these are uh surprising uh information anymore.

entry- level developers, product analysts, data analysts, business operations, and performance management.

Uh the more like uh predictable your job is, the more likely you're going to be be replaced. Will product managers be

be replaced. Will product managers be replaced? Yes, I think absolutely. Yes.

replaced? Yes, I think absolutely. Yes.

Um especially for growth product managers, guys. I do think that's why I

managers, guys. I do think that's why I pointed out entry-level growth product managers thinking process is very straightforward. They're optimizing the

straightforward. They're optimizing the funnel. They're optimizing the

funnel. They're optimizing the conversion. So um it's also very um

conversion. So um it's also very um doable with AI. Um if AI did all the analysis from your competitors, if AI know all your previous experiments,

their performance, they should also have good proposal on what is the next growth hack for you to pursue. There are

incremental innovations, right? In

incremental innovations are easier for AI to replace on the product manager.

Bigger BS are probably harder. This is

uh from our chief product officer.

Everyone needs to be a full stack builder with AI expertise. Um this is just where the whole industry is going.

Breaking to product management today. We

also talk a lot about interviews. If you

guys are more interested, go to my YouTube channel. Okay? We literally

YouTube channel. Okay? We literally

don't have more time to go through all of that. And uh the APM is extremely

of that. And uh the APM is extremely competitive. So is it worth it? Right?

competitive. So is it worth it? Right?

I'm leaving my words here. I don't think it is for everyone. Uh so there are also a lot a lot of other product management positions that you want to go through.

Go go go to maybe AI startups hardware physical AI is a huge thing right now other industries or related functions right you don't want to just be very competitive APMs all the time uh for

your like all for all the uh position that you apply for when you are doing job search or you can create your own company which is one of the reason why I started this course for a lot of you

guys. um what is your competitive

guys. um what is your competitive advantage right we talk about this that's why we started from product strategy because I want all of you guys to build product leaders no matter you are in big companies or startups so your

model is a competitive advantage your data or your operations right for AI application startups we previously also talked about it's really really hard for you to have technical advantage build a

good product is the key right now a lot of AI application startups have very shitty product experience that is also um an consequence of most

of the founders are tech founders. Um I

think this is when a disruptive technology happen. Most of the tech

technology happen. Most of the tech founder most of the founders are tech founders and it just leads to a not good product experience. Um now we're not

product experience. Um now we're not saying it not good, not even targeting like conservatives. We're just talking

like conservatives. We're just talking about visionaries and visionaries don't even want to use it. So a good product is also a key for you to um build up

your business advantage. a lot of AI landscape. Um these are the divisions

landscape. Um these are the divisions that you probably want to keep an eye on. Um that's also why a lot of

on. Um that's also why a lot of companies that I um use as examples in this class fall into a lot of these divisions because I want you guys to research on these divisions. These

divisions might have more positions open, might have more opportunities in the future. So yeah, keep an eye. So how

the future. So yeah, keep an eye. So how

to stand out on LinkedIn? just some tips for you guys um to stand out on LinkedIn for your job search or for your startup.

Um just do a screenshot. I won't go through everything of them. Now I'm open for questions.

Um I have a question. So we've been talking about impact a lot in this class and I guess I wanted to know I wanted to get a more exact um definition of impact

like what exactly are we talking about when we talk about impact? you talk

about your uh potential impact to your true north, right? In our example, it is uh revenue, then you talk about how much money you can earn.

If your uh primarily true north is a daily active user, then you're talking about uh how much daily active user you can drive to your company. If your true north is let's say the number of

reactions on feed, then the impact is how many reactions on feed you can drive to the uh to the company. One of the advice on how to stand on Lingo

mentioned open to work. Do you think it is good to have that batch or people or leaders or product managers whom we want to talk to they consider they are

desperate enough and the conversation will lead to getting a job.

>> There could be but different recruiters think differently.

>> Um all I can tell you is from a product side I don't know how the individual recruiters think. If you want to target

recruiters think. If you want to target meta, I don't know how meta recruiters, single recruiter think, right? But all I can tell you is on the recruiter product, there is a filter called open

to work. So they could uh very easily

to work. So they could uh very easily use that filter filtering out all the people who are not open to work.

>> There is one feature right open to work but for private you don't get the badge but the recruiters can see.

>> Yeah.

>> Open only to recruiters.

uh can you give us some uh suggestions about what kind of product manager we can choose for our first product vendor job?

>> Oh that is uh very complicated question. So

um we talk about in terms of the five product manager managers right I do have a very weird hypothesis of

what are the personalities for different product managers. Um so I thought you

product managers. Um so I thought you said you were all of them.

>> Yeah, because I'm so good at product management, so good at my job.

>> Personality disorder.

>> Yes. Um, so for the consumer product managers, they are very very feeling based. So they are very sensitive to um

based. So they are very sensitive to um to users feeling and how they think.

They're also very empathetic. Um, so I do think that one person's personality is pretty much that. So if you are very creative, if you are very empathetic, if

you are very go with the flow kind of person, then consumer product manager is a good role. On the other side, in contrast, right? Um if you are a very

contrast, right? Um if you are a very organized person, if you are very good at um understanding out of a whole like complicated system where your product

land at and how to improve your product, you are for enterprise. Okay. Enterprise

product manager are not like as like creative as consumer product managers but they are very good at organization and communication because for

enterprises I always like to use this as an example. Consumer product managers

an example. Consumer product managers make you feel like you're playing basketball. Basically you can give your

basketball. Basically you can give your ball to your um to your teammates or you can just shoot the the ball yourself.

But enterprise product manager feels more like playing soccer. It is very hard, extremely hard for you to take one ball from one one side of the field to the other side. You probably have to

work with your player to shoot the goal.

So this is how I um want to separate these two. Consumer product managers

these two. Consumer product managers need a lot of creativity because you have a lot of room to play in.

enterprise product management you need to work with a lot of stakeholders because everything you change on your product um means a lot to the whole ecosystem to the whole system as well.

Um so yeah that's why the organization and the creativity uh and also the communication versus um versus user empathy and for the growth product manager.

Yeah, if you look at me I'm a growth product manager. Uh I'm very data

product manager. Uh I'm very data driven. I look at numbers. I'm um yeah

driven. I look at numbers. I'm um yeah very like like believe in data. Data is

my strengths. Data is my evidence. I

don't believe in any other people's I just believe in data. that

is growth product managers. Um they're

very good at um numbers and um they're also very like growth product manager give you a feeling of every quarter you are um

uh gambling like it has that feeling.

So, if you are the kind of person you're crazy about metrics, like if you look at your experiment report, if you it tells you that you drive $10 million, you're going to be happy for no reason for the

next three weeks. You are a growth plan manager. I'm that kind of person. Okay?

manager. I'm that kind of person. Okay?

My experiment report is my daily dopamine. So, you feel happy from the

dopamine. So, you feel happy from the impact that you drive. That's the growth plan manager. Um,

plan manager. Um, also, what do I mean by data sense?

That's a very interesting observation I have. So our VP on growth uh has a very

have. So our VP on growth uh has a very unique skill set. So um for example in the wheelie update you say that oh you

did this feature and you drove let's say 0.1% daily active user his data s his data sense is so good he will point you out saying no that feature cannot

drive.1% down

drive.1% down like he knows what type of feature can drive how much of an impact in a rough sense without estimation that's such a

good data sense he has previously ramp a feature And I said however like we had user I drove and then he was like there's no there's no way and he's just so sure

about this and my data scientist backed me up. My data scientist was like no

me up. My data scientist was like no this is not a direct read out from the report but we did all all this analysis and we lend on this number and this number is also reviewed by all the data

scientist leadership and my VP was like no that's not right that number doesn't sound right. So after the weekly re uh

sound right. So after the weekly re uh meeting we just reviewed the numbers and we did make a mistake in that uh in that couple.

>> Wow. So yeah, that's the data sense and that's usually when you are very high up as a growth product manager. That's how

good you are with the whole ecosystem.

And business product manager, they're very logical. They're very good at math.

very logical. They're very good at math.

Um they're just very business driven.

They're bit good at business strategy.

Uh they do not really need to care about how user feel a lot. Um they just care about how much money or how much numbers they drive. uh platform product manager

they drive. uh platform product manager very good at observation very good at um communication as well also if you want to be a very successful platform product

manager you also have to tell a good story about how to quantify your impact because your impact is very hard to quantify if I only own the data infrastructure platform at LinkedIn it

is very hard to prove to the leadership that I actually am a hardworking person >> because it is very hard to associate what my what the feature that I build on

my team with business metrics of the company. You can't say that because you

company. You can't say that because you do this on the platform uh the internal employees like their efficiency increase whatever because you're not tracking it.

There is no way of you actually tracking an employee efficiency, right? Uh so you can only do an estimation. Um so that is very hard for you to eventualize your

work to the leadership research for product managers is um first of all you have a passion on research not a lot of people have that passion you want to

read a lot of re uh study and like paper yourself and you probably was a researcher and you just don't want to continue doing research anymore. So uh

but you have a huge passion on the new technology coming up um and you are also very technical driven. You are good at coding as well uh usually. So you can

kind of match the researchers and being their assistant as well in their uh research.

>> Got it.

>> So coming back to your question as a new which path do you want to go? Uh I would say based on your personality. Um, for

example, I went through the whole LinkedIn APM program and um, I don't need a rotation to tell me that I like consumer and growth, but I need to go

through that rotation anyways. Um, so

it's just based on my personality and how I would want to work and where my passion is. I'm really interested. Um,

passion is. I'm really interested. Um,

but eventually if your scope is becoming bigger, uh, you will know how to do different type of product management.

Anyways, but this job market just act like this like however you started your vertical your employ your future employer is going to think that that is your expertise. So that is uh what I

your expertise. So that is uh what I think a weakness of today's job market.

Um but that is just people's perception.