The biggest misunderstanding about AI is to use it for layoffs.
Is AI solely the concern of the technology department? Is it just about cost reduction and efficiency improvement (layoffs)?
Is it impossible to develop AI without big data?
Does an organization become an AI organization just because all employees can use AI?
If you hesitate even a little in your answers, then you might be standing on the verge of being eliminated by the era.
This morning, Shen Pan, an instructor at the Chaos Innovation School and a tutor at the AI Innovation Institute, dissected the AI implementation and growth path from L1 to L5 for everyone, and thoroughly explained the transformation methods for six major business scenarios. This is not only an efficiency revolution but also a paradigm reconstruction related to survival.
01 Core Framework for AI Implementation - The Growth Path from L1 to L5
Professor Li Shanyou once said in a large - scale lecture in Hangzhou that we are in history without realizing it. We are used to understanding the past by looking back, but often fail to perceive it in the present. Ten years ago, an era with far - reaching influence was just beginning. In 2010, Meituan, a group - buying tool that "took advantage" of merchants, emerged. In 2011, WeChat allowed users to post on Moments and make voice calls. In 2012, Didi, which imitated Uber, was launched. In the same year, Toutiao, a content website that recommended content based on users' interests, came into being. In 2015, Pinduoduo, an e - commerce platform that offered discounted goods through group - buying, was established.
How many people predicted that they would become giants disrupting business models a few years later? In fact, looking back at the consumption map of China in the past twenty years, we can find that the Internet industry has penetrated into all aspects and has become an important driving force for China's consumption development. In 2010, Alipay launched QR code payment. In 2012, Tmall was established. In 2013, WeChat Pay was launched, and the number of Alipay users exceeded 100 million. In 2014, WeChat Red Envelopes were introduced. In 2015, Pinduoduo was founded. In 2016, Douyin and Taobao Live were launched. In 2018, Li Jiaqi's live - streaming sales exceeded 100 million yuan per day. In 2020, the live - streaming e - commerce market exceeded one trillion yuan. In 2023, self - live - streaming by stores has become a daily routine. Back in 2015, we should have started doing live - streaming because we already knew that the era of short - videos and live - streaming was about to begin. So, can we predict the future at present?
Many people say that the world is constantly changing and VUCA (volatile, uncertain, complex, and ambiguous), making it impossible to predict. However, many seemingly uncontrollable problems stem from insufficient knowledge or shallow thinking. Just like being stuck in a traffic jam while driving, you may not know what's going on ahead. But if you climb to a high - rise building and overlook the whole city, you will naturally know where the traffic jam is and how to get through. During the period of discontinuous development, the fundamental reason why we feel jumps and breaks is often a technological revolution.
In the early stage of each technological revolution, the existing value network - positions, businesses, products, and technologies - is first changed. In the initial stage, since people can't clearly see its long - term potential, it is often first used to increase revenue and improve efficiency. For example, after Watt improved the steam engine, it first drove spinning machines, serving the then - developing fields. L1 is to regard it as a new tool to empower job functions. L2 is to regard it as a new ability and apply it to business scenarios. L3 is to regard it as a new technology to achieve product innovation. After its first - principle becomes clearer, L4 native products and L5 native platforms begin to emerge, creating a new value network.
02 L1 - Job Empowerment: AI as a New Tool to Improve Personal Efficiency
L1 is based on positions and aims to improve personal work efficiency by mastering AI tools. One of the misunderstandings about AI implementation is to focus on cost reduction and efficiency improvement first. If employees first see it as "cutting" and "compressing", they will naturally resist. However, if the revenue is increased first and employees see new opportunities, then improving efficiency and reducing costs will be like "icing on the cake", and they will embrace the change. A company doesn't succeed by saving money but by making money.
Therefore, when implementing AI, it should start with revenue - generating departments such as sales, marketing, and brand planning. Once there is revenue, employees will embrace it. Many companies are used to starting with AI - efficiency - improving departments such as HR, design, or finance. The HR department undertakes the functions of organizational management and internal lubrication. Excessive cuts will hinder the company's operation. The design department is indeed likely to be replaced by AI, but it is not a wealth - creating department, and cost - saving from it is limited. Instead, it may lead to discussions about large - scale layoffs. The finance department holds a large amount of data, and at present, AI still has deficiencies in accuracy. Replacing it rashly will bring many troubles.
There are three important aspects in L1. First, master some common position - level AI tools and organize them into your own toolset. Some of these tools are used for image - making, some for data analysis, and some for PPT production. After repeated iterations, they are very easy to use and can greatly improve the efficiency of daily work.
Second, master multi - dimensional tables. Multi - dimensional tables integrate data aggregation, AI analysis, and visualization. For example, when analyzing various popular copywriting and imitating it, just import the copywriting links from different platforms into Feishu's multi - dimensional table. The system can extract data such as likes, comments, and shares, and use large - models to analyze the selling points, usage scenarios, user groups, and popular structures behind them, and batch - replicate popular content. Another example is in the store - inspection scenario. In the past, store - inspectors had to understand the store situation offline, take photos and send them to the group, and it might take a week to organize. Now, just synchronize the photos to the multi - dimensional table, and AI can automatically recognize images and check information within a few minutes. Multi - dimensional tables can also create a "clone" for sales managers, quickly organize sales reports, and real - time aggregate customer follow - up records, and provide guidance on adjusting sales scripts. At the same time, multi - dimensional tables can automatically generate visual charts based on data, allowing companies to see trends at a glance from bar charts, pie charts, line charts, etc., greatly improving the quality of decision - making.
Third, master RPA, public opinion or competitor monitoring robots. As long as you select the types of data to be collected, it will, like a robot, simulate human actions such as clicking, copying, pasting, searching, aggregating, downloading, and checking, and automatically repeat the execution according to the steps. Using RPA, we can easily collect store data, analyze live - streaming room bullet - screens, and even automatically reply to comments in the live - streaming room, realizing an unattended live - streaming room.
03 L2 - Business Empowerment: AI Drives the Transformation of Six Major Business Scenarios
L2 is about mastering scenario applications, mainly including six major business scenarios: business strategy, customer value - added, product innovation, brand marketing, omnichannel operation, and organizational effectiveness.
I. Business Strategy
(1) Insight into the Business Essence
An important module in the AI - empowered business strategy is the insight into the business essence. The understanding of the business essence determines the entire business logic. Drucker said that strategy is not about studying what to do in the future but what to do now to have a future. He had a classic set of three questions: What is my business? What will my business be in the future? What on earth is my business?
The market boundary of a business is determined by customer choices and value perception, which is a strategic judgment. The growth boundary is determined by the execution capabilities such as marketing coverage, service operation, product extension, and ecological connection. The implicit boundary is related to value assumptions and depends on a company's values and the leadership of its founder.
Taking the essence of the pet food business as an example, the first question to ask is: Why do people keep pets? Some people say that when they come home from work and hug their running dogs, all the stress and fatigue disappear. Some people say that when they are very tired from overtime work and look back at their cats in the distance, they feel inexplicably happy and seem to have the energy to continue working. Others say that when their 5 - year - old son takes good care of a sick kitten, they feel that he has suddenly grown up. Companionship and stress relief, love and responsibility, education and cultivation... These are the reasons why people keep pets.
If pet food is only regarded as dog food, then consumers pursue cost - effectiveness and convenience. However, when you see the essence of pets as family members, pet food needs to be nutritious, healthy, and diverse, just like choosing food for your own children. Saving money is no longer the core. This is highly similar to the maternal and infant industry. In both industries, the users and consumers of products are separated. Those who buy don't use the products directly, and their judgment is based on brand and reputation. Many dog foods have a DHA formula. The demand for puppies to eat DHA comes entirely from pet owners. In other words, the real consumers of pet food are pet owners. Understanding this, the essence of the industry is completely different.
Another example is selling phone cases. A phone case costs thirty yuan, and selling one million phone cases a year results in an income of about thirty million. If you want to reach one hundred million, you must go global. After going global, the same phone case can be sold for more than ten US dollars, and the same sales volume can reach over one hundred million. What if the target is over three hundred million? The unit price must be raised to five or six hundred, or even seven or eight hundred yuan. At this price, it is no longer just a phone case. Like CASETiFY, its patterns are mostly created with artists and have become a platform for young people to express themselves and show their creativity. Its sales exceeded three hundred million US dollars in 2022. This shows that the business essence determines the business ceiling.
Therefore, the core questions for insight into the business essence are: Why do consumers pay for the product? Ask "why" again for the above answers. Fundamentally, what is the product exactly, and what is it not? What is the core of value creation? On what key assumptions is the business logic mainly based? How is it different from traditional consensus? Which assumptions may be broken in the AI era? Summarize it into one sentence: What is the essence of XX product?
Many companies find it difficult to think about the business essence, but with the assistance of AI, this problem becomes simple. A student in the non - ferrous metal profiles and composite materials business, whose target customers are consumer electronics and new energy companies, got the conclusion from AI that the essence of this business is to provide "high - performance, customized" non - ferrous metal profiles and composite material solutions for consumer electronics and new energy companies, serving as the "underlying carrier for function realization" of terminal products and empowering their core competitiveness, such as lightness, battery life, and energy - efficiency ratio. This student was greatly inspired. In the past, they constantly emphasized cost - effectiveness and cost issues to customers and got caught in price - cutting competition. But if the business direction is set to help the other party improve their core competitiveness, the situation will be completely different.
(2) Industry Research
After having a clear understanding of the business essence, we enter another key module in the AI - empowered business strategy - industry research. In the past, people often used the column - thinking method, predicting revenue based on a fixed growth rate. If the performance last year was eighty million and this year it is one hundred million, then next year's performance data is probably one hundred and twenty million. This algorithm makes some sense. However, if a company had a revenue of one hundred million three years ago and calculates that it should reach one billion in five years based on a 50% compound growth rate, but the entire industry scale is only seven billion, it obviously cannot occupy one - seventh. Therefore, column - thinking cannot guide long - term thinking.
Many companies are like small fish that can only see their peers around but don't know whether they are in a fishbowl or the sea. This requires using pie - thinking to deduce the future from a spatial perspective. First, observe the market scale and see how big the "cake" determined by user needs is. Second, think about the market share. A company's core capabilities determine how much of the cake it can eat. Third, plan for the deadline. The task plan determines how to eat the cake. Pie - thinking requires us to study the entire industry beyond a single business. Otherwise, we may either leave the industry too early thinking it has no prospects or stick to a declining industry and find it increasingly difficult.
Based on the mission, vision, and strategic intention, industry analysis usually focuses on six aspects. First, industry definition. What are the criteria for defining the industry? Clearly define specific categories and sub - fields. Second, driving factors. What are the underlying laws of the industry? Is it policies, technologies, consumption trends, or historical paths? Third, development space. How big is the future space of the industry? What are the growth rate, concentration, maturity, and penetration rate? If the concentration of an industry is very high, then the opportunities for entrepreneurs are few. Fourth, industrial chain. Where are the key bottlenecks in the industry? Analyze the components of the industrial chain, the value chain of the business model, the proportion, profit margin, and key capabilities of each link. Fifth, external environment. What are the favorable factors for the industry? Macroeconomic policies, economic trends, and geopolitics will all affect the industry's development. Sixth, typical companies. How do the benchmark companies in the industry operate? Just like a physical examination, compare your key indicators such as sales, costs, personnel structure, and R & D investment with those of industry benchmarks to quickly identify gaps and improvement directions. Today, AI tools have greatly lowered the threshold for making industry analysis reports. Taking Mita Search as an example, as long as you input an industry or business, it can automatically generate an analysis framework, fill in data gaps, and convert data into visual charts.
(3) Meeting Decision - Making
In the AI - empowered business strategy, another key module is decision - making. A common scenario in decision - making is meetings. Many company meetings end without any follow - up, and what is said is quickly forgotten. Each meeting consumes time, energy, and collaboration costs. The discussions lack structure, the problems are not focused, and the tasks cannot be implemented. More unfortunately, the real good ideas, suggestions, and insights are not retained, and finally, the company falls into the vicious cycle of "meeting - training - meeting again", and work efficiency becomes worse and worse.
I designed a meeting review intelligent agent. By importing the original meeting records into the intelligent agent, it can generate a complete meeting summary,