Thinking Framework of a Product Manager

by Alex Ho

(Image Source: Storyset)

Disclaimer:

In this article I will share how I think, analyze problems, and make decision.
In this article I will not introduce you the day-to-day product manager work details such as how to write a PRD, how to create Jira tickets, etc.
This article is the English version of another article : 互聯網產品經理思維與方法論

(Last update: Dec 8th, 2022)

Table of contents

“How you think” matters a lot
Critical Thinking
General skills
Some skills in the day-to-day work
Framework
General Rules
Some skills in the day-to-day work
Example: Dissect a product by product canvas. Quickly explain what it does and how it works
Example: What are the input and output?
Example: Comparisons between 2 solutions
Example: Identify which issues are both important and urgent
Example: Figure out what is minimum effort required in order to meet the standard of success. and why?
Example: Classify various items into a few categories
Framework Example - Answer these questions before you start to deal with a problem
Thinking process
Framework Example - Define the product strategy
Product Vision: What problems are we solving by doing so? What is the state after we solve the problem?
What to do & what we don’t do
Product Roadmap: How to make the “vision” happen？
What is the impact?
What’s next?
Leverage data in the decision making
The key points when we use the data
What can you do in practice?
Some examples of “jump to conclusions after seeing the data”
Include user feedback in your decision-making process
Business Mindset
Product success cannot do without business success
There is always more than one method to make a business profitable
Human cognitive bias
Communication
What should we achieve when we deliver a report to our bosses
思維、能力、策略
產品經理
經濟、投資
產業筆記
運動筆記
歷史、文化
日本歷史文化
旅行建議、推薦景點
台灣
上海
電影筆記

“How you think” matters a lot

As a product manager, you have to carefully think before you make the decisions such as

asking your engineer to do A first instead of B in the product
asking your product squad to do X first and do Y later (instead of in reverse order)
asking your team to do task Z by solution 1 instead of solution 2

Your decisions have huge impact on

What values that the product can deliver to your customers (to solve their problems)
How you team allocates resources (we we have limited resources)

💡

It’s hard to come up with “perfect decisions”. But we should try our best to find the “most suitable decision at this moment”

Critical Thinking

💡

“Critical Thinking” helps us identify correctness and effectiveness

General skills

Be a bit skeptical of everything that you hear or read

Do not take “the things that other people take for granted” for granted.
Tell yourself “I myself could be wrong”. Even your boss’s opinion might be wrong.
Use scientific evidence, data, and logic to judge right from wrong, possibility, and impact.

Not everything is either black or white. There is always a gray area

Do not simply jump into “yes or no”
Assess the possibility of “yes” and “no” separately. And then make your decision based on that.

Be able to tell the differences between correlation and causation

Correlation does not imply causation.
Do not misunderstand the correlation as causation

Considering the limited resources (that you have now) and choose the best solution that you can make at this moment

Sometimes the decision that you make right now is not for the immediate effect but for the long-term goal.
Sometimes we might make a decision like this - “Because of reasons A and B, we decide not to do X. We will only do X when Y happens.”

Some skills in the day-to-day work

Put yourself in someone’s shoes

If you were your customer, what would he/she care about? How would he/she think about your idea?

“How will your boss think about this?”

If your boss is going to make the decision for an issue that should be decided by you, what could be his/her thinking process?
When your boss (e.g. a BIG boss) reads your quarterly report deck, what might he/she ask?
If you can expect “the things that your boss would definitely ask you for more details”, why don’t you provide more context beforehand and proactively tell him in the first place? (Before he/she asks)

Framework

Backgrounds:

As a product manager, most of the time you won’t repeatedly work on the same tasks.
You continuously deal with different problems, develop new features, solve bugs that have never been seen before, etc.

Question:

For each task that we solved, is there any tactic that we can reuse in the upcoming new tasks that we encounter in the future?

Among all of the problems that we encounter, there can be some underlying logic (that might rarely change) and some environmental variables (that change depending on what the problems are)

💡

Framework = ( Underlying logics + Environmental variables )

General Rules

As a product manager, most of the time, you won’t have much repetitive work. You continuously cope with different problems, develop different features, solve different bugs, etc.
Create a framework when you think and make decisions.

Based on your own experiences, find out the common/similar/repetitive components in your work. Find out the common principles among them.
It is possible that you can use the process that you solved problem A to solve another problem B that you face in the future.

However, you will not be able to use exactly the same method to solve every problem. You will need to adjust your thinking process, decision-making process to cope with the issues from different products and situations

Some skills in the day-to-day work

Example: Dissect a product by product canvas. Quickly explain what it does and how it works

The components in a product canvas

What’s the Problem
Data as the proof
Hypothesis
Solution (Options)y
KPIs
Actors (client, stakeholders, business sponsors…etc)
Efforts Required
Risks
Impact

Reference: Framework for building high-performance data products

Example: What are the input and output?

When you run into a complex issue or are assigned with a brand-new task. Firstly, figure out what is the required input (e.g. input for a system, a feature) and what the output is (e.g. results, impact, revenue). Then the problem might look easier.

Example: Comparisons between 2 solutions

	Solution 1	Solution 2
What should we do	A, B & C	A, B, D, E & F
Pros	Less expensive	Full functionality
Cons	Less room for scalability	Higher flexibility for future scalability
Engineering efforts	Low	High
Dependencies	N/A	from another team A
Additional IT costs	Low	High
Estimated launch time	May 2023	July 2023

Example: Identify which issues are both important and urgent

	Importance: High	Importance: Low
Urgency: High	Task A	Task C
Urgency: Low	Task B	Task D

Example: Figure out what is minimum effort required in order to meet the standard of success. and why?

Example: Classify various items into a few categories

Source: Defensibility x Scalability = The Marketplace Matrix

For more framework examples, check the following two chapters.

Framework Example - Answer these questions before you start to deal with a problem

Thinking process

What is the status quo?

What are the customer's pain points? Which are the ones with top priorities and why is that?
How big the negative impact is (from those pain points)?

Why should we do this?

By doing so, what customer pain points can be solved?

What options do we have? which option should we go with?

Write down the pros and cons of each option.
Assess the resources required for each option.
Write down the reasons why you suggest (the team) go with a certain specific option

When should we start to do this? and why?

Question: What are the differences between “we do this right now” vs. “we will do this in 6 months from now”? What might we lose? What opportunities might we miss?

Figure out “what we should do” and “what we should NOT do”

Resources are limited. When we decide to do A, you might not be able to do B simultaneously. You need to figure out the opportunity cost of both A and B
“Building many features” do not definitely lead to “bringing out a good product”
There can be numerous “no” before each “yes”.

What are the impacts of doing so?

How will we change or improve the status quo?
What are the before and after?
What values will be generated for each party in our ecosystem?

End consumer
Enterprise clients
Our own company
External 3rd-party partners

What are the sizable impacts? Some examples as follows

Increase the revenue
Decrease the cost
Reduce the unnecessary waste of resources

Ｗhat are the impacts that cannot be quantified? Then try to describe the impacts in qualitative ways

Framework Example - Define the product strategy

The following items are common components that we see in product strategy.

In practice, of course, we won’t be able to answer every single question stated.

Product Vision: What problems are we solving by doing so? What is the state after we solve the problem?

What is the holistic goal of our company?
What are the goals that we (as a team in the company) should achieve in order to contribute to the company’s holistic goal?
Which problems do we plan to solve now? Who are the target customers?
Examples

E.g. To solve problems for end consumers or enterprise customers
E.g. We aim to do X, Y, and Z. So that we can contribute to our company’s goals A, B, and C.
E.g. To increase revenue or make breakeven or reduce the lost

What to do & what we don’t do

Articulate the definition of done

Define the boundary. What is not included in the current phase?
Eliminate the ambiguous and misunderstanding

Remember to explain “why”

Why do you suggest doing these?
Why we do NOT do those (for now)?

Provide enough context

Backgrounds
Proof (quantified data and qualitative information)

Run experiments

Set up proper Successful Criteria & Unsuccessful Criteria in the AB Tests

Product Roadmap: How to make the “vision” happen？

How many different options do we have?

What are the pros and cons of each option?
Which option do I suggest adopting? What are the reasons?

What to do in each phase? (if your overall project will be separated into multiple phases)

Why we should do X first (that means now)?
Why we should do Y later? and when should we do Y then?

What is the impact?

What are the TAM, SAM, and SOM?

TAM: Total Available Market
SAM: Serviceable Available Market
SOM: Serviceable Obtainable Market

What are the sizable impacts?

e.g. We expect to reduce the cost by X%
e.g. We expect to increase the revenue by Ｙ%
e.g. We expect to reduce the time that the customer success team spend on certain task by Z hours per month
e.g. We expect to reduce the possibility that the operation team make a human error by N% per year

Ｗhat impacts cannot be quantified? Then please describe the impacts in qualitative way

What’s next?

After we deliver a new product or feature to the market, what product metrics should we monitor?

Example: If the metric X can be increased or decreased by N%, it means we meet the Successful Criteria

After we deliver version 1 (or MVP) of a product (or feature), what is the next step?
What is the expected end-state?

Leverage data in the decision making

Everyone knows “we need to use data”. But how?

The key points when we use the data

What data metrics should we use or monitor?
How to interpret the data?
What are the action items after we review the data?
Make sure we do not misunderstand the data.

What can you do in practice?

Before you have your data analyst get data for you (of course you can do it by yourself as well), write these on a paper

What questions do we want to get answers from the data?
What are my assumptions (before seeing the data)?
Depending on the results that data tells us, what decisions should we make to solve the problem?

By doing so, you can probably prevent these things from happening:

Not sure how to analyze the data
Not sure what’s the next step (after we are overwhelmed by a long excel sheet)

Some examples of “jump to conclusions after seeing the data”

Example 1 - You saw this in news “Unemployment rate in Q2 is quite high”

How can I dive deeper into this?

Examples:

The cause of the “high unemployment rate in the current year” might be rooted in 20 years ago

Because young people today were born around 20 years ago.
Then, anything happened 20 years ago? (e.g. Baby Boomer)

Compare the number with historical data in the past few quarters or years. What trend do we see?
Separate the population by age, location of the residence, etc.
Find out the external factors or underlying causes. Examples such as:

War
COVID pandemic
Economic sanctions
Birth rate a few years ago.
There are many job openings. But enterprises are not able to find suitable candidates.

Companies in this country require people with

The number of job openings is decreasing.

Example 2 - You saw this in news “F&B GMV increased 75.8% year-over-year in July 2022. The economy is booming for the F&B industry”

You might feel that “wow it’s great! Things looks fine”. How can I dive deeper into this?

Examples:

What happens in July 2021? Was there a low base period?

Was there a peak of the COVID-19 pandemic?
Were the consumers banned from dining in the restaurants?

Is the CPI in July 2022 much higher than in July 2021?
Does the currency exchange rate change a lot in the past year?

References

Factfulness: Ten Reasons We're Wrong About the World--and Why Things Are Better Than You Think

Buy Factfulness: Ten Reasons We're Wrong About the World--and Why Things Are Better Than You Think on Amazon.com ✓ FREE SHIPPING on qualified orders

www.amazon.com

Factfulness: Ten Reasons We're Wrong About the World--and Why Things Are Better Than You Think

Include user feedback in your decision-making process

Remember our own limitations

We don’t have the expertise in many things in life and work.
Product managers cannot come up with a product requirement based on their own imagination without proof/logic/reasons.
People don’t know “what they don’t know”

Listen to the users. But do not fully believe what they told us

Focus on these

User pain points
User’s problems being solved

Do not focus too much on these

“A user said he/she wants A, B, and C features”

Think about “why users said they want A, B, and C features”

Business Mindset

💡

A product cannot succeed without business success

Product success cannot do without business success

Figure out the cost

Examples: salary, cost of production, marketing spending, CAC (Customer Acquisition Cost)、tax、transaction fee, subscription fees for using a SaaS tool, etc

Figure out if it is worth it.

Profit Margin = (Revenue - Cost)
(I know the above state meant is way too rough. But I believe you get what I mean.)

Do not judge a business by limited information or narrow viewpoints.

E.g. A company with low profit might have a super big business scale. So it makes a lot of money.
E.g. merchandise with high prices might not result in high revenue, because the transactions per month might be low.

Assess the TAM, SAM, and SOM
What is the funds-flow?
Assess the ROI
How can we scale the business?
Do not judge a business by the limited data of subjective viewpoints

There is always more than one method to make a business profitable

Example: increase the revenue

Increase the online traffic
Increase the transactions
Increase the price per item
Bundling sale
Upsell ….etc

Example: Increase the profit

Increase the price per item
Reduce the cost…etc

Human cognitive bias

💡

Remember that we all might have some cognitive biases that affect our decision makings. We can consciously try to avoid making these mistakes.

Some common cognitive bias

Most of the time, we are not rational.
Prospect Theory

Risk-averse (e.g. when there are low-hanging fruits)
Risk-seeing (e.g. when faced with potential loss)

The latest information we get tends to formulate our main impression of a man/issue/incident.

E.g. We might remember the recent happy things clearly and forget the sorrow that happened years ago

Make a judgment based on limited information

Communication

Conclusions first

Come to conclusions first.
Then the backgrounds, reasons, thinking processes…etc
Next steps.

MECE (Mutually Exclusive Collectively Exhaustive)
The point is not what you said but “what the audience thinks and feels” after they heard from you
Make sure the information is conveyed to “the people who should know this”
Find out the most efficient ways to convey the information to “the people who should know this”
As long as you are not sure your audience (or readers of your documents) has the same level of context in mind as you do, then you should assume that they do not know enough background context.

So you should provide enough context, definition of terms (being used in the document), simple examples, proven data, etc.

In your statement (no matter in verbal conversation or written document), identify what the subjective and objects are.

Do not get your audience confused that “who are the they that you are referring to”

What should we achieve when we deliver a report to our bosses

(I personally think there are 3 possible scenarios)

When you don’t need your boss to give you suggestions or orders

Hi boss, please read this. All is fine and under our control.
The current situation is blah blah blah.
All you need is to trust me (and my team)
I will get things done by YYYY/MM

When you need agreement from your boss

Hi boss, please read this. All is fine and under our control.
These are the actions that we plan to take. And these are the reasons why we proposed to do these actions.
Boss, please agree with our proposed actions and grant us some guidance

When you need your boss to make decisions for you

Hi boss, please read this.
Here are the problems that we’re trying to solve.
And here are the proposed solutions from the team. And there are pros and cons, resources required, a timeline for each solution.
Boss, we need your order or guidance on which solution should we go with.

About Alex