The essence of enterprise management is to improve operational efficiency, or how to re-understand management when management begins to become shackles? What is called “resource allocation and flow efficiency”. How can enterprises improve management efficiency?
Empiricism vs. Positivism
Enterprise management is facing the real world, and the real world is complex and changeable. For a long time, the management profession has responded to this complex and changeable modeling method: to constantly update its own theories and methods according to the needs of management for talents.
So, how does the demand for talent come from the management? Of course, it is a passive change that follows the changes in the market:==
- When in a seller’s market and the market environment is more certain, the way to maximize the interests of the employer is to regard the labor side as an “economic animal”, while management puts forward the hypothesis of “economic man” and puts forward a series of concepts, theories and methods, such as “no measurement, no management” and so on.
- When it began to turn to the buyer’s market and the market changed rapidly, it transformed “economic man” into “social man” and “self-actualized man”, and also put forward a series of human concepts, theories and methods, such as “demand hierarchy theory”, “two-factor theory”, “reinforcement theory”, “expectation theory”, “X-Y theory”, “super Y theory”, “Z theory” and so on.
The transformation of basic assumptions means that management theories, methods, techniques and tools are also transformed. These transformations are correct, necessary, and progressive. For example, the introduction of these scientific theories has been extended to “activity cost management”, “balanced scorecard”, “Inamori accounting” and so on in the field of accounting. It can be said that scientific management has created a new era, and the management of Chinese enterprises is still learning and catching up.
Consulting firms are constantly following these changes by proposing different operating techniques to collect money from companies, while the industry continues to use these constantly updated theories, methods, techniques and tools.==
Specifically, management and managers have been formed into such a paradigm: factual data – > sensory information – > empirical model – > management practice – > perceptual experience.
- Stage 1:The first stage: all kinds of first-hand factual data generated in the business process of the enterprise.==
- Stage 2: some factual data will form sensory information after being observed by humans, and the collectors of this information mainly have three roles: scholars, consultants and business managers.==
- Stage 3: Modeling. Enterprise managers will summarize their own experience, but if they are too busy with practical work, or their ability is insufficient, or their internal opinions are difficult to unify, they need to ask consultants to help enterprises provide methods, technologies and tools to solve management problems in addition to observation and interviews, such as McKinsey & Company. Academics will mainly observe, experiment, and come up with theories for use by consulting firms and corporations.==
- Stage 4: The model contains concepts, methods, and technologies, and finally encapsulates them into standard management tools, such as the Three Pillars, the 7S model, the Balanced Scorecard, the Yang Triangle, etc., which are again applied to management practices. And practice will produce new factual data, and so on.==
- Stage 5: Managers become familiar with the experiential model and form perceptual experiences.==
Therefore, managers, consultants, and academics are constantly moving around this paradigm. Two philosophies were formed: empiricism and positivism.
Empiricism exemplifies Fayol’s “planning, organizing, directing, coordinating, and controlling”, which attempts to summarize a set of universal theories and methods through observation and genius insight.==
Positivism is typical as Mr. Li Yuhui said in the introduction to “Organizational Behavior”, it is not like empiricism to pursue abstract, concise, universal conclusions, but the pursuit of complex, pluralistic, conditional conclusions, scholars from various enterprises to get data, put forward models, to give conclusions.
In this way, management is an empirical discipline, and the way to help managers improve their management ability is to continuously accumulate experience and learn from the experience of their predecessors, and it seems that in addition to starting from perceptual experience, management and managers cannot improve management ability at all.
In essence, these two management philosophies have a fundamental problem, and the commonality is that they have never faced the “essential solution”, but have been working on the “phenomenon solution”.
- Empiricism tries to find an essential solution because it wants to come up with universal conclusions. However, it has two fundamental cruxes: 1) empiricism cannot solve specific problems, for example, no one can say that the five functions of “planning, organizing, commanding, coordinating and controlling” are of no value to enterprise management, but in the specific implementation process, they are ever-changing; 2) It is impossible to avoid the dilemma that the inductive method will eventually be invalid and the experience is unreliable, that is: even if 10,000 enterprises have been verified by a certain empirical management theory, it cannot be said that it is effective for 10,001, so you can’t learn Haidilao.==
- Positivism is mainly responsible for solving the first deficiency of empiricism, so positivists go deep into enterprises to do experiments, such as Taylor and Mayo, and they simply don’t look for the essential solution, isn’t the real world complex and changeable? So, where is the unified theory? It will change with the environment. Therefore, the above-mentioned “economic man”, “social man” and “self-actualized man” are constantly finding solutions to phenomena in the phenomenal world, which are the work of positivists. Although positivism keeps pace with the times, it is still limited by two points: 1) if the real world changes too quickly, the efficiency of experiments cannot keep up; 2) Positivism is still empirical and will still be limited by the shortcomings of the above inductive method.==
Big Data & Machine Learning
For example, the big data of Internet companies is more reliable than the traditional positivists, and they have to go to the enterprise to learn from it, or the business community now has its own research institutes and laboratories.==
Bottom-up data-driven, rather than top-down management-driven, is increasingly the consensus of the business community.==
Again, there are 2 questions here.
Big data is also an inductive method, and it is relevant, but it is still limited by the natural shortcomings of the inductive method.==
At this stage in the development of artificial intelligence, the tide of bottom-up data-driven algorithms such as machine learning has reached a bottleneck, and the solution of replacing causality with correlation has not solved many of the core problems.==
In her book Why: The New Science of Causality, the founder of machine learning, Judea Pearl directly criticizes that data can tell you that patients who take medication recover faster than those who don’t, but it can’t tell you why. Perhaps, those who take the medicine choose to take it only because they can afford it, and even if they don’t take it, they can still recover so quickly.
This is the first point.
Second, especially in the field of management of small data and professional knowledge, the method of using data to reverse train algorithms is basically unreasonable, and it inevitably requires the efforts of expert knowledge systems. For the management field, it is not only the processing of unstructured data that is more important, but also the domain knowledge, and its keywords are such as recruitment, performance, OKR, channel, responsibility, team, senior management, employees, social security, leadership…… of a limited collection.==
Not every manager has an understanding of machine learning and causal science, but it is important to say that machine learning is not a panacea, don’t be intimidated by data-driven AI, and now the AI industry is expecting better algorithms (e.g., causal inference) from academia to help break through this bottleneck in AI development.==
Causal science is a form of rationalism. However, it is not pure Platonism (the world of ideas), but a new application of a Kantian “a priori synthetic judgment of how possible” in the data age.
Rationalism and Causal Science
From the very beginning, rationalism has focused on essential solutions rather than phenomenal solutions, which is a very different starting point from empiricism and positivism. The same is true of causal science, which first establishes an essential solution to causal relationships and then looks for data to populate.==
The scientist who used this method to the extreme was actually Albert Einstein, who first came up with the general relativity formula Rμν−12Rgμν=GN8πc4Tμν (a priori causal judgment) through pure mathematical logic, and then, Arthur Eddington’s 1919 expedition confirmed Einstein’s prediction that the sun would deflect light during the total solar eclipse on May 29, 1919 (this is a piece of data), which cemented the position of general relativity as a true scientific theory. Since then, many observations have confirmed the correctness of the general theory of relativity.==
There is no doubt that Einstein’s view of relative space-time is closer to Newton’s absolute view of space-time than the “essential solution”, because it can explain and solve the problem of microscopic high-speed space, and it can also be said that without Einstein, computers would not have existed at all.==
Therefore, although empiricism and positivism are also effective, they have always been limited to seeking answers only at the level of “phenomenal solutions”. If, in order to find the solution to the essence, it is necessary to resort to rationalism, because, if you have some knowledge of philosophy, it must be easy to understand, we cannot say that logic is the essence, but logic must be closer to the essence of reality.==
For enterprises, is it possible to establish a causal relationship for the enterprise, first reflect its essence, and then enter the data of its business operations, and establish a data structure that can reflect the essence of the enterprise?==
The answer is yes. But there are three major problems, before understanding these problems, it is necessary for us to have a certain understanding of the basic assumptions (axioms) of enterprise management activities, enterprise management is facing the real world, it follows two self-evident basic axioms:
Axiom 1: The real world is interconnected and dynamic.
Axiom 2: The more you grasp the facts that are close to the real world, the more you will have the true knowledge to solve problems.
This is not new, as previous empiricist and positivist scholars have known and are dealing with the complex and changing realities of the world in their own way. Or, let’s try to establish an understanding that:
- Isn’t it complicated to connect?==
- Isn’t it dynamic, but it’s time-related and changeable?==
- Isn’t the facts that are close to the real world factual data?
- Isn’t the true knowledge of problem solving based on fact-based data modeling?
After the above analysis, the paradigm of the traditional method is: management practice – > factual data – > sensory information – > empirical model – > management practice. This expression is not intuitive, let’s put it another way, as follows:
Therefore, they are not based on the original factual data direct modeling, but first form a perceptual view (experience), and then establish an empirical model, what is lacking is the causal relationship that can reflect the true essence like Einstein, so every management practice can not have an explicit connection with the essence, and is scattered in the minds of different managers to hide, forming different intuitions of different people, and even can never be called out. It will also leave with the departure of managers, so it is never possible to accumulate at the essential level, which is the fundamental crux of the traditional management paradigm and has never been solved.==
Therefore, their conclusions are valid, but they cannot be connected and dynamically valid, because they have never touched the more hardcore essential model, and when the environment changes, they need to constantly ask consulting companies to deal with this change, in fact, consulting companies can only continue to put forward new theories, methods and tools in the original paradigm, but they are trapped in the same paradigm, and they cannot solve problems in a connected and dynamic way.
Obviously, this method is extremely inefficient in today’s era, and it cannot keep up with the changes, because the “incompetence” of the model itself makes it impossible to process a lot of factual data that can reflect the real situation.==
To improve the efficiency of enterprise management, there are two major problems:
- How to deal with complexity in response to Axiom 1?
- How to quantify complex and time-sensitive data to cope with axiom 2?
However, the road to simplicity is to model the factual data from the perspective of the whole/network, at this time, the formed network is not linear, but a complex network graph, so that it realizes a kind of “prior judgment”, and then the data is run in this kind of graph, such as:
In this simple graph, between every two connected nodes, there must be a causal relationship between support and support, rather than correlation, and the real operation of the enterprise is the same, if each transaction does not support each other, there is no need to establish an enterprise, and the enterprise becomes strong because of mutual support.==
The data in the diagram are concrete, but their causality is in the form of a “transcendental” logic, which is precisely the application of Kant’s “a priori synthetic judgment of how possible” mentioned above. It has the following features:
- The graph itself is a complex network, so it is inherently connected and reflects the real world.
- The graph can be simulated, so it can process time-dimensional data.
- The graph can even be causal, so it is more essential.
- The graph algorithm is designed to deal with the network, so it can not only deal with explicit and easy-to-network data, but also deal with implicit non-networked data in the minds of managers, and its quantitative results are closer to the truth.
So, Valor is building a new paradigm.
Logical pragmatism
Valor does not reject the theories, methods and tools of the traditional paradigm, but, just as Newton’s classical mechanics is a macroscopic special case under Einstein’s theory, in Valor’s eyes, traditional theories, methods and tools are also special cases adapted to a specific time and space field of the enterprise, but the development of the enterprise is faced with uncertainty, and the special case cannot always be effective.==
Businesses need to face the market cheaper, more connected, and more dynamically.
Valor’s approach is:
- Use rationalist philosophical ideas and causal science to establish a causal map.
- The original, factual data generated in practice rather than the information generated in practice is incorporated into the causal relationship map, and then a dynamic and complex network of business facts is generated as a “data base”.
- Further, the calculation is carried out and converted into an objective contribution value for transactions, people, skills, resources, etc.
- Based on this “pedestal”, the index system generated by different management concepts, technologies and tools at different stages born in empiricism and positivism, such as those described in management, organization, and human resource management theories, is adopted.==
Compare the traditional paradigms: management practice – > factual data – > sensory information – > empirical model – > management practice – > factual data.
Valor’s new paradigm is: Management Practices – > Fact-Based Data and Empirical Models – > Rational Models (Complex Networks) – > Intelligent Algorithms – > Management Practices – > Fact-Based Data. **This expression is also not intuitive, and it is also translated into a graph:==
Under the new paradigm, the factual data generated by each management practice does not have to go through the “distortion processing” of the intermediary of “human sensibility” to form sensory information, but goes straight to the essence (complex network), so that we can say that every practice of the enterprise can be connected with the ontology of the enterprise, and it will integrate the multiple conclusions of empiricism and positivism, and the unique management experience model of the enterprise to unify the factual data. The empirical conclusions will be linked to the essence. Therefore, although it is still the manager of the flow, the management accumulation of the enterprise has always been there, thus forming an invisible but iron-clad camp.==
Think about it, isn’t this kind of accumulation of enterprise management and the way that the traditional management paradigm has been seeking a solution to the phenomenon an efficiency improvement of several orders of magnitude?
In this way, Valor completes a paradigm shift:
Since empirical evidence does not emphasize actual combat activities, this is the biggest difference from practice, which emphasizes that knowledge comes from practice and uses knowledge in practice. Therefore, what has been practiced must be empirical, and the reverse is not necessarily true; At the same time, what is practical must also be experienced.
As Huawei says, actions to achieve goals must be reasonable. The specific target value is the company’s determination and demand, which is determined according to the company’s development requirements and cannot be bargained. What needs to be discussed with the company is how to achieve the goal, that is, the action strategy and resource requirements, which can be discussed, but also must be discussed and deduced. Any target of Huawei must be deduced in advance, and the atomic bomb must first explode on the blackboard before it can explode on the ground.
It’s also completely consistent with Valor’s philosophy, but the atomic bomb exploding on the blackboard is too primitive, so let’s explode on Valor.
Thus, we can simplify the formulation of a new concept: logical pragmatism.
Note: The “objectivity” here is not the objectivity of the “thing itself” mentioned by Kant, it is concerned with the record of the facts that happen in the enterprise. The “ontological world” here is not the “ontology” like the “world of ideas” mentioned by Plato, but the essence behind things, which is often reflected as “intuition” in the subconscious in the human brain, and “reason” in the scientific context.
