The best book I've read recently that relates to work is "Thinking in Systems." I took Professor Huang's course on "System Operations and Value Chain Management" before, but I felt that Professor Huang didn't fully present the very good concept of 'system' in management and supply chain to the students, which led me to not fully realize the importance of 'system'. It wasn't until I read this book that I realized how shallow my previous understanding was. I also specially collected an MIT System Dynamics online YouTube Series, preparing to study it thoroughly when I have time.
are as follows:
: Constants and parameters, such as subsidies, taxes, and standards
Only when parameters enter the range of one of the higher items on the list do they become leverage points. In our daily work, many of the metrics we focus on are vanity or ineffective metrics. Sometimes the designed metrics are overly cautious, leading many PMs to concentrate on achieving short-term goals at the expense of long-term benefits—focusing on trees rather than forests. Therefore,
: The size of stable inventory relative to its flow
For non-physical startups like ours, there are no inventory-related issues, but buffers for us include:
Adequate capital reserves to ensure cash flow doesn't break (most small companies fail because of this); At least 1-2 backups for each position; More than one product line (this contradicts the advice of many investors in startups who suggest focusing, but I actually think focus isn't applicable to all stages and all companies).
: Physical systems and their intersection nodes
For example:
Designing the company's CCC cash conversion cycle (Cash Conversion Cycle) well. Planning the product and business pipeline in advance. Talent reserve planning. We often lament that we don't have enough good PMs, but we're too lazy to invest energy in recruiting and training PMs.
: The speed at which the system responds to change
While reading Wang Dingjun's memoirs, he compared the resistance between the Nationalists and the Communists:
"How to handle such situations, even the superior officers never taught, they pretended not to know, so I had to pretend it didn't happen. Later, when I was captured by the PLA in Tianjin and underwent re-education, I found out that they held group meetings every night to exchange experiences, improve shortcomings, and solve problems. Only then did I realize that back in Shenyang, inside and outside the city, these review meetings might have been held every night. They shortened their sleep time to work, while the military and political officials of the Nationalist government in Northeast China shortened their sleep time to enjoy pleasures. It was exactly as the saying goes: 'Some lie on firewood below the stage while others dance above, pitifully none of them sleeps.' "
If our iteration speed is fast enough and our learning curve steep enough, we can win in the end competition.
: Who can access information structure
The absence of information flow is one of the most common reasons for poor system function. Adding or restoring information may be a powerful intervention method, and it is usually easier and less costly than rebuilding the physical infrastructure of the system. While ensuring information security, we maintain transparency internally. Sometimes, granting permissions means greater authorization and trust.
: Incentives, punishments, constraints
As the CEO of the company, it is more necessary to establish reasonable and fair rules. When we strive to reshape the rules and understand how changes in these rules affect people's behavior, we can truly understand the power of the rules. Besides making rules, we must lead by example and follow the rules. For our relatively grassroots startup, we sometimes don't do well, such as always making "exceptions", not relying on personal connections, but still relying on rules.
: The ability to add, change, or develop the structure of the system
Self-organization is the highest expression of a system's resilience. A system capable of self-evolution can adapt to various changes by altering itself to maintain survival. Previously, I tried to persuade colleagues in the company to learn AI, but the effect was minimal. Later, Lily organized several AI competitions, forming self-organizations, and in this process, those who were already interested in AI and had accumulated knowledge stood out, allowing the company to give them more space to play. Extensive self-organizing computer models show that complex and surprising patterns can emerge from just a simple set of rules. Respecting diversity and encouraging experimentation may allow organizations to play a greater role.
: The goal of the system
A participant in the system can clearly set, articulate, repeat, support, and adhere to new goals, thereby guiding the transformation of the system. I used to not understand the importance of vision and mission, but now I increasingly feel that they are the foundation of a company. When chaos appears in some daily work, vision and mission can better guide individual decision-making and behavior. For each specific organization, the O in OKR is also very important, setting reasonable Os is fundamental.
For example, I once consulted a colleague from Alibaba, asking about the difference in goals between Pinduoduo and Alibaba. Alibaba's goal is "No difficult business in the world," and since they once used GMV as a performance indicator, they kept oscillating between serving small clients, brand clients, and consumers. On the other hand, Pinduoduo firmly serves users. For instance, a merchant selling oranges once became a hit product, and Pinduoduo's operations immediately solicited similar varieties, the same quality but at a lower price from suppliers, redirecting user traffic to this new supplier. This better protected users but indeed somewhat harmed the original orange supplier who initially made the hit product. However, users benefited and stayed on the platform, strengthening the consumer side of this two-sided network effect. Moreover, Pinduoduo once used the interest from supplier payments (not Alibaba's GMV) as the source of salary and bonuses for the category team, so many operators went to more remote places to find suppliers willing to delay payment for longer periods, thus enhancing the diversity of the system and the platform's negotiation power.
: The mindset of the system's goals, structure, rules, delays, and parameters
Paradigms are the root cause of why systems become systems. Rooted in these paradigms, the system's goals and information flows, feedback, various stocks and flows, and everything about the system are generated. The system's goals, structure, rules, delays, and various parameters are all directly influenced by paradigms.
Returning to Liu Run's addition, subtraction, multiplication, and division:
Addition in the business world is same-dimensional cooperation (upper-left quadrant). For example, how do you arrange sales work? Assign 10 salespeople, each independently developing customers and bringing home their own grain? If so, you're managing the company with "addition." Each salesperson contributes to the overall performance of the company in the same way and with the same weight.
But some companies don't operate this way. For example, Beike manages the company through multiplication. Multiplication in the business world is cross-dimensional cooperation (lower-left quadrant). Most real estate agencies have each salesperson working independently. However, in Beike's view, this "scattered soldier" style of "additive management" cannot achieve great things. Why not try multiplication?
Beike divides the work of real estate agents into 10 roles. No one person can independently complete the sales work; they have different divisions of labor and collaborate to complete a single transaction. Their contributions to this transaction come from different dimensions (10 dimensions), and their weights are also different. This is cross-dimensional cooperation, this is multiplication in the business world. The best algorithms in the business world are basically multiplicative.
This change from addition to multiplication is a change in business paradigm.
In addition, we need to learn generative AI, which is also a change in production technology paradigm. Trump's appearance at the 2024 Bitcoin Conference embracing blockchain represents a change in financial paradigm. These are all very powerful levers.
: Transcending mindsets
Compared to changing paradigms, there is another lever point at a higher level, which is freeing oneself from any paradigm control, maintaining flexibility, and realizing that having no paradigm is "real."
This may be a kind of open-mindedness, sincerely respecting different cultures, different voices, and different ideas.
Quoting from the book:
"Approach the whole reality with the heart of a child. Only then can one enter a state of 'emptiness' (not-knowing), entering the 'enlightenment' realm spoken of in Buddhism."