Knowledge Point 11: Agnostic Marketing
Knowledge tier: [General layer]
Layer One: Extracting the Skeleton
Definition: When facing a disruptive technology, you cannot ask customers “what do you want”—because the customers themselves don’t know either. No one can know whether they will like a product, or how they will use it, before they have actually used it. So what you need is not “market research” but “agnostic marketing”—entering the market with the assumption that “we don’t know what customers want,” and discovering the market through rapid iteration and rapid trial-and-error.
Underlying logic: This is essentially a question of “the epistemology of the market.” For sustaining technologies, the market is “knowable”—customers know what they want, and you can learn their needs through research. For disruptive technologies, the market is “unknowable”—not only don’t you know, the customers don’t know either. The only way is to go try it yourself and let the market tell you the answer.
Layer Two: Digging Into the Meat
Why analysis: Mature enterprises fail at disruptive innovation largely because they trust market research too much. They spend millions of dollars on research, asking customers “would you buy this product,” and then the customer says “no,” so they give up. But when a customer says “I wouldn’t buy it,” it isn’t because they truly wouldn’t buy it—it’s because they haven’t used the product yet; they have no idea at all whether they would buy it.
Case 1: How Honda Motorcycles Entered the American Market
The story of how Honda motorcycles entered the American market is a classic case of “agnostic marketing”:
- The original plan: Honda originally planned to sell large motorcycles in America—the same ones they sold in Japan. They did market research, and the results showed that “Americans like large motorcycles.”
- The actual situation: Honda sent several employees to America to sell large motorcycles, and they sold very poorly. For months they could barely sell a few.
- The unexpected discovery: These employees themselves rode Honda’s small motorcycles (the 50cc Super Cub) around Los Angeles to run errands. Many Americans saw them and asked, “where can I buy one of these little motorcycles?”
- The pivot: Honda’s employees didn’t think much of it at first—they were there to sell large motorcycles. But as more and more people asked, they began to realize: maybe what Americans want isn’t large motorcycles, but these small, light, cheap little motorcycles.
- Huge success: Honda began selling small motorcycles in America, and it was a tremendous success. By 1964, Honda already held more than half of the American motorcycle market.
The most ironic part: if Honda had strictly followed the market-research results back then and sold only large motorcycles, they might have exited the American market long ago. Their success came not from planning well, but from the plan failing, and then accidentally discovering the real market.
Case interpretation: Market research can only tell you “what customers wanted in the past”; it cannot tell you “what customers will want in the future.” The market for a disruptive technology isn’t “researched” into being—it’s “blazed” into being.
Case 2: How Intuit Discovered QuickBooks
The story of the financial-software company Intuit is equally fascinating:
- The original product: Intuit’s first product was Quicken—a home-finance software for individual users. Quicken sold well.
- The unexpected discovery: Intuit’s customer-service team kept getting some strange calls—the callers were small-business owners who were using Quicken to manage their company’s books. The customer-service team said, “Quicken is for individuals; you should use professional financial software.” But those small-business owners said, “Professional financial software is too complicated; we don’t need that many features. Quicken is simple, but it’s just enough for us.”
- The key decision: Most companies would treat these calls as “users misusing the product,” and then find a way to “correct” the users. But Intuit’s founder Scott Cook didn’t do that. He said, “If so many users are managing their company’s books with our personal-finance software, that means there’s an unmet market need here.”
- Launching QuickBooks: Based on Quicken, Intuit developed a simplified small-business financial software—QuickBooks. This product had far fewer features than professional financial software, but was far simpler and far cheaper.
- Huge success: After QuickBooks launched, within just two years it captured more than 70% of the small-business financial-software market. Today QuickBooks is Intuit’s largest and most profitable business.
Case interpretation: What customers are doing with your product is 100 times more important than what they say they want. Behavior is always more honest than words.
Case 3: A Personal-Life Scenario—“How Do You Know You’ll Like It?”
The principle of “agnostic marketing” applies just as much in personal life. Think about it:
- When you chose your major in college, did you know you would like that major? Most people didn’t. You just picked it based on some vague impressions, and only after actually studying it did you find out whether you liked it.
- When you were looking for your first job, did you know you would like that job? Most people didn’t. You just thought “this company sounds good,” and only after actually doing it did you find out whether it suited you.
- When you and your partner first started dating, did you know you would get married? Most people didn’t. You just thought “this person is pretty good,” and only after actually spending time together did you find out whether they were the right one.
Most of the important decisions in life are made in a state of “not knowing.” You can’t “research” it all clearly in advance—you must first commit, first try, and only then can you know the answer. Those who always try to “figure everything out before acting” often end up treading water forever. Because for many things, if you don’t actually do them, you will never know the answer.
Case interpretation: Life is not an exam—there is no standard answer, and it’s impossible to review everything beforehand and then walk into the exam room. Life is exploration—you learn as you go, and adjust as you learn. This is the true essence of the “agnostic life.”
Layer Three: Distilling the Essence
Trigger conditions: You should adopt the “agnostic marketing” mindset when the following situations arise:
- You’re making a brand-new product—there’s nothing similar on the market
- You’re entering a brand-new market—you know almost nothing about its customers
- Your market-research results are very vague, contradictory, or say “there’s no demand”
- Customers are “misusing” your product—using it to do things you never designed for
- You catch yourself saying “customers don’t get it; they should be using it this way”—this is a danger signal
Cross-industry transfer matrix:
| Marketing method | Applicable scenario | Core assumption | Key to success |
|---|---|---|---|
| Traditional market research | Mature markets, sustaining innovation | Customers know what they want | Precise questionnaire design, large-sample statistics |
| Focus-group interviews | Product improvement, feature optimization | Customers can clearly express their preferences | An excellent moderator, sensible group composition |
| Agnostic marketing | New markets, disruptive innovation | No one knows what customers want (including the customers themselves) | Rapid iteration, small quick steps, learning from mistakes |
| Lean-startup MVP | Startups, brand-new products | Most of our assumptions are wrong | Validating the most important assumption at the lowest cost |
Execution SOP:
- Admit ignorance: First admit to yourself—“I don’t know what customers want, and customers don’t know what they want either.” This isn’t modesty; it’s a fact.
- Launch the simplest product as soon as possible: Don’t wait until it’s “perfect” to launch. Get out the simplest version (an MVP) as fast as you can and put it into a real market environment to test.
- Observe behavior, don’t listen to words: Don’t ask customers “do you like it”—observe what they’re actually doing. Did they pay? Do they keep using it? Did they recommend it to friends?
- Iterate rapidly: Based on the real behavior you observe, adjust your product quickly. Don’t cling to your “original plan”—your original plan is almost certainly wrong.
- Tolerate “messiness”: This process will look chaotic and disorganized. Accept that chaos—it’s the inevitable price of exploring an unknown market.
Layer Four: Using Up the Leftovers
Limitations:
- “Agnostic marketing” is not an excuse for “doing no research at all.” You still need to do research—but the focus of the research shifts from “what customers say they want” to “what customers are actually doing.”
- It requires a large amount of trial-and-error cost. Although it’s “low-cost trial-and-error,” the costs add up. Not every company can afford multiple failures.
- It can lead to “strategic drift”—if you just blindly follow customers’ behavior with no sense of direction at all, you may end up achieving nothing.
- Its applicability in the B2B market is more complex—B2B customers usually know clearly what they want, so the room for “agnosticism” is relatively small.
Failure boundaries:
- Making sustaining improvements in a mature market—here market research is very effective, and you don’t need “agnostic marketing”
- In heavily regulated industries (such as healthcare and finance)—you can’t just “trial-and-error” freely; every step must go through strict approval
- When the cost of failure is extremely high (such as launching a rocket or making a major investment decision)—you need to do as much research beforehand as possible, rather than “try first and see”
Critical perspective:
“Agnostic marketing” is a very mind-liberating concept—it tells you that you don’t need to be omniscient, that you don’t need to have a perfect plan from the start. But it’s also easily misunderstood as “blundering around blindly.” In fact, “agnostic” is an attitude, not a method. What’s the method? It’s scientific trial-and-error, careful observation, and rapid iteration. That is fundamentally different from “blundering around blindly.” Moreover, this concept doesn’t solve the problem of “a sense of direction”—you can’t just try everything. You still need a rough sense of direction, knowing roughly which direction to explore in. Where does that sense of direction come from? From vision, from intuition, from a deep understanding of human nature. “Agnostic marketing” isn’t telling you to throw away your sense of direction, but rather not to be held hostage by it—stay firm on the broad direction, and stay flexible on the specific path. That’s what a true master does.
Knowledge Point 12: Learning Plan vs. Execution Plan
Knowledge tier: [Metacognition layer]
Layer One: Extracting the Skeleton
Definition: There are two completely different types of plan:
- Execution plan: assumes we know what the goal is and how to get there. The purpose of the plan is to ensure efficient, accurate execution.
- Learning plan: assumes we don’t know what the goal is, nor how to get there. The purpose of the plan is to learn as fast and as cheaply as possible.
The typical mistake of mature enterprises is: using an “execution plan” for things that require a “learning plan.” They demand that innovation projects produce precise revenue forecasts, detailed execution roadmaps, and clear milestones—when in fact these things simply cannot be known in the early stages of innovation.
Underlying logic: This is essentially a conflict between two different management philosophies. The execution plan is based on a “philosophy of control”—we can predict the future, we can control the outcome, we can eliminate uncertainty. The learning plan is based on a “philosophy of adaptation”—the future is unpredictable, we can’t control the outcome, and what we can do is adapt quickly to change. In the known world, the philosophy of control works well; in the unknown world, the philosophy of adaptation is king.
Layer Two: Digging Into the Meat
Why analysis: This explains the innovator’s dilemma from the level of “management mechanisms.” The entire management system of a mature enterprise—budgeting, assessment, approval, reporting—is designed for “execution plans.” When you use this system to manage innovation projects that require “learning plans,” it’s like using a ruler to measure temperature—you’ve used the wrong tool, so the result is naturally disastrous.
Case 1: The Tragedy of Xerox PARC—Managing Breakthrough Innovation With an Execution Plan
We mentioned Xerox PARC earlier; let’s analyze it more deeply from the angle of “plan type”:
Xerox’s copier business was a mature business—everything was predictable. To manage the copier business, what you need is an “execution plan”:
- A precise production plan: how many copiers to produce next month, and what the cost of each one is
- A precise sales plan: how many to sell in each region, and what each salesperson’s target is
- A precise financial plan: what the revenue, cost, and profit are each quarter
- A precise assessment system: the deviation between actual results and the plan must not exceed 5%
But PARC’s work was breakthrough innovation—everything was unpredictable. To manage breakthrough innovation, what you need is a “learning plan”:
- Which 3 most important assumptions are we going to validate this quarter?
- Which experiments are we going to run to validate those assumptions?
- What do we expect to learn from those experiments?
- If an assumption is disproven, what’s our next step?
But Xerox’s management didn’t understand this distinction at all. They managed PARC the way they managed the copier business. They always asked PARC the same questions: “When can this technology be productized? When can it contribute how much revenue? When can it reach what profit margin?” When PARC couldn’t give precise answers, management lost patience, cut the investment, or sold the technology off cheap to others.
The result: Xerox held the future of the entire personal-computer industry, yet it got nothing out of it itself.
Case interpretation: Managing breakthrough innovation the way you manage a mature business is like watering desert plants with fish-tank water—you water them attentively, but the plants still die. It’s not that your intentions weren’t good enough; it’s that you used the wrong method.
Case 2: Amazon’s “Two-Pizza Teams”—Managing Innovation With a Learning Plan
Amazon has been able to keep producing breakthrough innovations largely because it understands this distinction:
- Mature business (e-commerce retail) uses execution plans: Amazon’s e-commerce business has very precise forecasts, very strict KPIs, and a very efficient execution system. In this business, any deviation from the plan gets corrected.
- Innovation businesses (AWS, Kindle, Alexa, etc.) use learning plans: For innovation businesses, Amazon doesn’t use the execution-plan approach at all. It doesn’t demand precise revenue forecasts, doesn’t demand detailed execution roadmaps, and doesn’t demand that every step go according to plan. It asks only one question: “What did you learn?”
Take the development of the Kindle as an example:
- At the start, no one knew whether an e-reader would succeed, and no one knew what it should look like, what features it should have, or what price to sell it at.
- Bezos didn’t demand a detailed five-year plan from the team. He gave the team resources and told them: “Go explore, go learn, go find the right direction.”
- The team made many prototypes, tested many different designs, and tried many different pricing strategies. They learned from each attempt and constantly adjusted direction.
- In the end, the Kindle was a huge success and redefined the entire publishing industry.
Case interpretation: Excellent companies don’t have “only one management method.” They have “multiple management systems”—for different types of business, using different planning methods, different assessment systems, and different management models.
Case 3: A Personal-Growth Scenario—Two Ways of Learning
The distinction between “learning plan vs. execution plan” also exists in personal learning:
- Execution-plan learning: You buy a thick Python programming book and make a detailed plan: “Weeks 1-2 learn basic syntax, weeks 3-4 learn functions, weeks 5-6 learn object-orientation, weeks 7-8 do a project. Study 2 hours a day, do exercises on weekends.” Then you execute strictly according to the plan. What is this learning method suited to? It’s suited to learning knowledge that “has definite answers”—like preparing for a standardized exam.
- Learning-plan learning: You have no detailed plan. You first find the simplest tutorial and spend a weekend making the simplest little project—say, a to-do list. Then you think, “If I want to add a user-login feature, how do I do that?” So you go learn the knowledge about user login and implement the feature. Then you think again, “If I want to put it online for others to use, how do I do that?” So you go learn about server deployment. Every step you take is based on what you learned in the previous step to decide what to do next. There’s no fixed schedule, no deadline that must be met. What is this learning method suited to? It’s suited to learning skills that “have no definite answers”—like programming, entrepreneurship, and writing.
Most people, when learning a new skill, instinctively use “execution-plan learning.” They make detailed plans, buy thick books, and sign up for expensive courses. Then they discover—the plan can never keep up with change, the actual situation never matches the plan, so they feel frustrated and then give up. Real masters use “learning-plan learning”—they don’t pursue the perfect plan; they start acting first, then learn through action, and adjust direction through learning.
Case interpretation: For “unknown” territory, the best plan is “no plan”—or rather, a “plan about how to learn.”
Layer Three: Distilling the Essence
Trigger conditions: You should use a learning plan, not an execution plan, when the following situations arise:
- You’re doing something you’ve never done before—there’s no experience to draw on
- The outcome of this thing is highly uncertain—you can’t predict what will happen
- The path is very unclear—you don’t know specifically how to proceed to succeed
- You catch yourself saying “let me start once I’ve figured everything out”—this is a danger signal
- You’re doing innovation, exploration, or R&D-type work
Cross-industry transfer matrix:
| Project type | Suitable plan type | Core question | Assessment metric |
|---|---|---|---|
| Building a house | Execution plan | How to execute efficiently and accurately? | Cost, quality, schedule |
| Improving a mature product | Execution plan | How to reach the predetermined performance target? | Performance gains, cost reduction |
| New-product development | Hybrid plan | How to explore possibilities while controlling risk? | Number of assumptions validated, learning speed |
| Disruptive innovation | Learning plan | How to find the right direction quickly? | Which assumptions were validated? What was learned? |
| Basic scientific research | Learning plan | How to discover new knowledge? | How much new knowledge was produced? What new discoveries were made? |
Execution SOP:
- Judge the type of situation: First determine whether you’re facing a “known situation” or an “unknown situation.” This decides which plan type you should use.
- If it’s an unknown situation, design a learning plan: Don’t ask “what goal do we want to achieve,” but rather “what are our 3 most important assumptions? How can we validate them quickly and cheaply?”
- Set learning milestones: Don’t set “complete feature Y by month X, day X,” but rather “validate assumptions A, B, and C before month X, day X.”
- Design “exit/pivot” criteria: Before you start, agree on: if an assumption is disproven, what do we do? Do we exit completely, or adjust direction and keep trying?
- Regularly review the learning outcomes: After each stage, don’t just assess “how much of the planned work was completed,” but even more “how much unplanned stuff we learned.”
Layer Four: Using Up the Leftovers
Limitations:
- Learning plans are easily abused—some people will use “we’re learning” as an excuse for “we haven’t produced any results.” You must have clear “learning goals” and “learning milestones,” rather than aimlessly “exploring.”
- It requires extremely high patience from management. Most managers are used to the certainty of “execution plans”—they like to see clear progress and clear numbers. The “uncertainty” of a learning plan makes them anxious.
- It can lead to the scattering of resources—if you try too many directions at once, you may achieve nothing. You need to find a balance between “staying open” and “staying focused.”
- It isn’t suited to large projects—when a project gets big enough, you must have some degree of execution plan, otherwise you can’t coordinate large numbers of people and resources.
Failure boundaries:
- When a project’s success depends heavily on “on-time delivery” (such as an Olympic opening ceremony or launching a rocket)—no matter how much uncertainty there is, you must have an execution plan.
- When the cost of failure is extremely high (such as building a nuclear power plant or a major medical surgery)—you can’t “try first and see”; you must make every possible preparation beforehand.
- Once a project enters the “scaling stage”—as soon as you’ve found the right direction, you must switch from a learning plan to an execution plan, otherwise you can’t scale.
Critical perspective:
The distinction between “learning plan vs. execution plan” is very valuable, but most projects in the real world are “hybrids”—they have both known parts and unknown parts. A pure learning plan and a pure execution plan are both extreme cases. The real challenge is: within the same project, how do you manage the known part and the unknown part at the same time? How do you maintain exploratory flexibility while maintaining execution efficiency? This isn’t an “either/or” question but a “balance” question. Excellent managers can flexibly adjust the ratio between the two in different situations—boldly exploring when learning is needed, and pushing hard when execution is needed. Even more importantly, excellent organizations can build a “dual-track” management system—having two planning systems, two assessment systems, and two incentive systems at the same time. One manages the mature business, the other manages the innovation business. That is the kind of organization that can truly innovate continuously.
Knowledge Point 13: A Failed Idea vs. a Failed Business
Knowledge tier: [Metacognition layer]
Layer One: Extracting the Skeleton
Definition: There are two completely different kinds of “failure”:
- A failed idea: you quickly and cheaply validated that an assumption was wrong. This kind of failure is a good thing—it lets you rule out a wrong option and move one step closer to the right answer.
- A failed business: you spent a great deal of time and money executing a wrong strategy, and failed across the board in the end. This kind of failure is a bad thing.
The typical problem of mature enterprises is: they punish “failed ideas” the way they punish “failed businesses.” The result is—no one is willing to try new ideas, everyone only does “safe” things, and innovation is smothered right there.
Underlying logic: This is essentially a question of “incentive-mechanism design.” If you punish people for “smart failures,” people will learn to “avoid anything that might fail”—that is, avoid any innovation. If you reward “smart failures,” people will be willing to try, willing to explore, and innovation will emerge. The incentive mechanisms of most mature enterprises systematically strangle innovation—they reward “not making mistakes,” not “learning from mistakes.”
Layer Two: Digging Into the Meat
Why analysis: This explains the innovator’s dilemma from the level of organizational behavior. Innovation isn’t a question of “genius inspiration”; it’s a question of “incentive mechanisms.” If quickly trying and validating a wrong idea gets punished, while being cautious and doing nothing is instead safe, then every rational manager will choose to do nothing.
Case 1: 3M’s “15% Policy”—How to Treat “Failed Ideas”
3M has a famous “15% policy”—engineers can spend 15% of their work time on any project they’re interested in, without needing their superior’s approval. This policy gave birth to countless classic products, including the Post-it Note. Let’s look at how the Post-it Note came about:
- Step one: A “failed” invention: Dr. Spencer Silver was a chemist at 3M. He had set out to invent a super-strong adhesive. But he failed—the adhesive he invented was weakly sticky; things could be stuck on and then peeled back off. At the time this looked like an outright failure.
- Step two: A “strange” application: A few years later, another engineer, Art Fry, was a member of a church choir. He kept running into a problem: the bookmarks in his hymnal kept falling out. One day it suddenly occurred to him—couldn’t Silver’s “weakly sticky” adhesive make exactly the kind of bookmark that wouldn’t fall out?
- Step three: More “failures”: Fry made the first Post-it Note samples and gave them to people inside the company to try. Most said “this thing is useless.” The market-research results were poor too—most people said they wouldn’t buy such a thing.
- Step four: Unexpected success: 3M didn’t kill the project because of these “failures.” They kept trying, giving away many free samples to business customers. And they discovered—as soon as people actually used Post-it Notes, they got hooked. Once they started using them, they couldn’t stop.
- The final result: The Post-it Note became one of 3M’s most successful products, bringing in over $1 billion in revenue every year.
The whole process was full of “failed ideas”—but 3M’s culture allowed this kind of failure, and even encouraged it. In most companies, if Silver had been criticized for “inventing a useless adhesive,” the Post-it Note would never have been born.
Case interpretation: Innovation isn’t moving from success to success; it’s moving from failure to failure while continuously learning, until the last failure turns into success.
Case 2: The “Innovation Strangulation” Process at a Large Enterprise
Let’s look at a cautionary tale. This is a true story:
A large traditional enterprise saw the internet industry developing rapidly and decided to “vigorously promote innovation.” The CEO announced at an all-hands meeting: “We must encourage innovation and tolerate failure!” The company set up an innovation fund and established an innovation committee.
The first innovation project was launched—a social app aimed at young people. The team invested 5 million, worked for half a year, and after the product launched there were few users, and it ultimately failed.
What happened next:
- The innovation committee required the project team to write a “detailed failure-review report”—deeply analyzing “why it failed” and “who should be held responsible.”
- The finance department made this project a typical case of “budget overrun, poor results,” and issued a company-wide notice of criticism.
- HR gave the project lead a “needs improvement” rating in the year-end review—affecting his promotion and bonus.
- At subsequent management meetings, this project was repeatedly used as a cautionary tale of “innovation failure”—“we must learn the lesson and avoid repeating the mistake.”
What was the result?
Over the following year, not a single department dared apply for the innovation fund again. No one was willing to risk “failure.” Every department did only “safe” things—continuing to optimize the existing business, continuing to serve existing customers, continuing to do the things that “definitely can’t go wrong.”
The company’s “innovation strategy” fizzled out. Two years later, this company was severely hit by internet enterprises, and its market share kept declining.
Case interpretation: “Encourage innovation, tolerate failure” is easy to say but hard to do. The real test isn’t in times of success but in times of failure—how you treat the people who “did the right thing but didn’t get a successful result.”
Case 3: A Personal-Life Scenario—Two Attitudes Toward Failure
Zhang Wei and Li Qiang are both product managers; each led a new product, and both failed:
- Zhang Wei’s reaction: He spent a lot of time “passing the buck”—“the market changed too fast,” “the competitors were too strong,” “the execution team wasn’t up to par.” He avoided admitting that his own strategic judgment was wrong. The only lesson he learned was: “don’t do such risky projects in the future; better to do safe things.” Over the next few years, he indeed did only “safe” projects—he never failed again, but he never produced a successful product again either. His career development stagnated.
- Li Qiang’s reaction: He wrote a detailed “learning report”—listing the 10 key assumptions he had made, which were right, which were wrong, and what he learned from them. He proactively reviewed everything with the whole team, sharing the lessons learned. He didn’t feel ashamed of the failure—he treated it as “tuition.” Over the next year, based on the lessons he learned from the failure, he made a new product that achieved huge success. His career development also moved up to a new level.
Same failure, so why such different results? Because they defined “failure” differently. Zhang Wei saw every failure as a “failed business”—a disgrace, something to avoid. Li Qiang was able to distinguish a “failed idea” from a “failed business”—quickly validating a wrong assumption is valuable learning, not disgrace.
Case interpretation: Failure itself is not wealth; learning from failure is wealth. But to do that, you first have to be able to distinguish a “smart failure” from a “stupid failure.”
Layer Three: Distilling the Essence
Trigger conditions: Your organization may be confusing the two kinds of failure when the following situations arise:
- “Do more, err more; do less, err less; do nothing, err nothing” becomes the prevailing mindset among employees
- The lead of a failed project is “ruined,” while the person who does nothing comes through unscathed
- People spend a lot of time “packaging” project results to avoid being labeled a “failure”
- All the “successful” projects succeed on PowerPoint, with no real market validation
- New projects get fewer and fewer, and everyone does “safe” projects
Cross-industry transfer matrix:
| Failure type | Cause | Value | How it should be treated |
|---|---|---|---|
| Smart failure (failed idea) | Quickly validated a wrong assumption | High value—ruled out a wrong option, learned something | Reward, encourage, share the experience |
| Understandable failure | Tried the right thing, but failed due to uncontrollable external factors | Medium value—accumulated experience | Tolerate, summarize the lessons |
| Incompetent failure | Failed due to lack of ability or carelessness | Low value—wasted resources, learned nothing | Criticize, train |
| Repeated failure | Made the same mistake again and again | Negative value—wasted resources, learned nothing | Zero tolerance, must be held accountable |
Execution SOP:
- Define clearly: Clearly distinguish the different types of failure within the organization and write out specific criteria. Let everyone know—what kind of failure is a “good failure,” and what kind is a “bad failure.”
- Differentiated assessment: For exploratory projects, the focus of assessment is “learning quality,” not “business results.” For execution projects, the focus of assessment is “business results.”
- Build a “failure-review” culture: Require a review after every project—whether it succeeded or failed. Focus the discussion on “what we learned,” not “who should be held responsible.”
- Publicly reward “smart failures”: For failed projects that quickly and cheaply validated an important assumption, also give public recognition and reward. Set an example.
- Leaders lead by example: Leaders should publicly talk about their own failures and share the lessons learned from them. This is the most effective way to change the culture—if even the leaders aren’t afraid to admit failure, employees will be even less afraid.
Layer Four: Using Up the Leftovers
Limitations:
- “Fail fast” can be abused—some people will use “it was just a failed idea” as an excuse for their own incompetence or carelessness. You must have clear criteria to distinguish a “smart failure” from a “stupid failure.”
- The theory doesn’t give an operable standard for “how to distinguish”—this still requires the manager’s judgment. There’s no simple formula.
- Excessive tolerance of failure can lead to a “failure culture”—people no longer strive for success, but get used to failing. This is a dangerous slippery slope.
- In capital-intensive industries, the cost of “failing fast” is too high—this method’s applicability is limited. You can’t “quickly trial-and-error” by building 10 nuclear power plants to see which one works.
Failure boundaries:
- When the cost of failure is extremely high (such as a medical accident or a safety accident)—you can’t fob it off as a “failed idea.” Any failure is unacceptable.
- When a project has already entered the execution phase and the “assumption” has been validated—if it fails now, it’s a “failed business,” and someone must be held responsible.
- When the same mistake is made repeatedly—the first time may be a “failed idea,” but the second time it’s a “failed business.”
Critical perspective:
“A failed idea vs. a failed business” is a very important distinction, but it’s also a concept that’s hard to put into practice. The problem is—when you’re in the middle of things, it’s hard to judge whether a given failure is a “smart failure” or a “stupid failure.” Hindsight is always easy, but judgment beforehand is very difficult. Moreover, even if the organization can distinguish them, at the individual level “face” and “career-development concerns” will still make people afraid of failure. The real solution isn’t “a more precise definition,” but “a safer psychological environment.” When people feel psychologically safe and aren’t afraid of being punished for honestly admitting mistakes, they’ll be willing to try, to learn, and to share. That is the true foundation of a culture of innovation. Finally, we also can’t go to the other extreme—deliberately failing for the sake of innovation. Failure is never the goal; learning is the goal. We tolerate failure not because failure itself is good, but because it’s a necessary path to success.
Knowledge Point 14: The Evolution of Industry Paradigms
Knowledge tier: [General layer]
Layer One: Extracting the Skeleton
Definition: Every industry has its dominant “paradigm”—that is, the rules of competition, the technological trajectory, the value network, and the definition of the customer that the whole industry commonly recognizes. Disruptive innovation is essentially a “paradigm revolution”—it isn’t doing better within the existing paradigm, but creating an entirely new paradigm and then gradually replacing the old one.
The evolution of an industry paradigm usually follows a fixed pattern:
- Emergent stage: The disruptive technology first gains a foothold in a fringe market or new market ignored by the mainstream enterprises
- Improvement stage: The disruptive technology keeps improving along its own technological trajectory
- Crossover point: The disruptive technology’s performance reaches the minimum requirement of the mainstream market
- Disruption stage: Customers begin to switch to the new paradigm on a large scale, and the enterprises of the old paradigm begin to fail
- Dominant stage: The new paradigm becomes the industry’s dominant one, and then the cycle repeats
Underlying logic: Why does the new paradigm always defeat the old one? Because the pace of technological progress always exceeds the pace at which market demand rises. This means—the technology that is “not good enough” in the low-end market today will always, tomorrow, reach the “good enough” standard of the mainstream market. Once it reaches that standard, the basis of competition changes—shifting from performance to dimensions like price, convenience, and reliability, which are exactly where the new paradigm’s advantages lie.
Layer Two: Digging Into the Meat
Why analysis: Understanding the pattern of industry-paradigm evolution can help you identify the threat before the “crossover point” arrives. Most mature enterprises fail not because they reacted too slowly after the crossover point, but because they were completely unaware of the threat’s existence before the crossover point—they felt “that thing isn’t in the same market as us.”
Case 1: The Photography Industry’s Paradigm Shift From Film to Digital
The photography industry’s paradigm shift is a textbook-level case:
- 1975: A Kodak engineer invented the first digital camera. But Kodak’s mainstream customers (professional photographers, film-developing shops) had no interest at all in digital cameras—the digital camera of the time had only 10,000 pixels, and its image quality was unacceptably poor. Kodak shelved the technology.
- 1990-2000: Digital cameras kept improving, with pixels going from 100,000 to 1 million to 3 million. The mainstream market still said “the image quality isn’t good enough.” But an emerging market appeared—ordinary family users didn’t care about ultimate image quality; they cared about “shoot and view instantly, no developing needed.”
- 2000-2005: The crossover point arrived. Digital cameras’ pixels reached 5-10 million, and for 95% of users the image quality was already “good enough.” Customers began switching to digital cameras on a large scale.
- 2005-2012: The old paradigm collapsed. The film market shrank at a rate of 20-30% per year. Kodak’s revenue and profit fell off a cliff. In 2012, Kodak filed for bankruptcy protection.
- After 2012: The new paradigm dominates. Smartphone photography is now starting to disrupt digital cameras—this is the next round of paradigm shift.
The whole process took 37 years (1975-2012), but the real collapse happened only in the last 7 years. And 10 years before the collapse happened, most industry experts were still saying “digital cameras could never replace film.”
Case interpretation: The rhythm of a paradigm shift is “slow, then sudden.” Most enterprises die in the “sudden” phase—they previously felt “there’s still time,” and then suddenly there wasn’t.
Case 2: The Paradigm Shift Happening Now in the Auto Industry (Gas Cars vs. Electric Cars)
We are witnessing the auto industry’s paradigm shift with our own eyes:
- 2008-2015: Tesla launched the Roadster and Model S. The mainstream automakers said “electric cars are a niche toy—short range, slow charging, expensive; they’ll never amount to anything.” Electric cars gained a foothold in the high-end niche market.
- 2016-2020: Electric cars kept improving. Range went from 300 km to 500 km to 700 km, charging got faster and faster, and prices began to come down. China’s new EV makers (NIO, XPeng, Li Auto) appeared. The mainstream automakers began to “seriously” make electric cars, but still focused mainly on gas cars.
- 2021-2023: The crossover point arrived. In the Chinese market, EV penetration soared from 5% to over 35%. Consumers suddenly discovered “electric cars really are better to drive, better to use, and cheaper than gas cars.” Sales at the mainstream gas-car makers began to fall, inventories piled up, and a price war broke out.
- 2023-2030 (predicted): The old paradigm collapses. The gas-car market shrinks rapidly, and a large number of traditional automakers fail or get merged.
- After 2030 (predicted): The new paradigm dominates. Electric cars make up over 80% of car sales. Smart cars and autonomous driving become the new focus of competition.
This time, many traditional automakers are still repeating Kodak’s mistake from back then—they feel “there’s still time,” “users still prefer gas cars,” and “electric cars still have lots of problems.” But history tells us that once a paradigm shift passes the crossover point, it’s irreversible.
Case interpretation: The most dangerous illusion during a paradigm shift is “there’s still time.” By the time you clearly see the threat, it’s usually already too late.
Case 3: A Personal-Career Scenario—The Paradigm Shift in Your Own Industry
Zhang Wei worked at a traditional print-media company for 10 years. Let’s look at the paradigm shift he went through:
- Before 2010: Print media was the absolutely dominant paradigm. Zhang Wei’s job was stable and his income was decent.
- 2010-2015: New media began to appear—WeChat official accounts, Weibo, Toutiao. Zhang Wei and his colleagues said, “Those are all fast-food content, with no depth. Readers still need professional media like us.”
- 2015-2018: The crossover point arrived. Print media’s advertising revenue began to fall off a cliff. A large number of print publications shut down. Many of Zhang Wei’s colleagues lost their jobs.
- After 2018: The new paradigm dominates. New media became mainstream. Short-video platforms (Douyin, Kuaishou) are now starting to disrupt image-and-text new media.
Where was Zhang Wei’s problem? He always felt “that’s another industry, nothing to do with us.” He didn’t realize—this isn’t “another industry”; it’s a paradigm shift within the same industry. Once the paradigm shift is complete, “another industry” becomes “your industry.” And by then, when you want to transition, it’s usually already too late.
Case interpretation: The biggest trap of a paradigm shift is “that isn’t the same market as us.” Until one day you suddenly discover—it has already become your market.
Layer Three: Distilling the Essence
Trigger conditions: When you observe the following signals, an industry paradigm shift may be underway:
- “Heterogeneous competitors” appear at the edge of the industry—their technology, model, and customers are completely different from yours
- These “heterogeneous competitors” grow at a rate far higher than the industry average, even though they’re still very small
- The profits of the industry’s leading enterprises begin to fall, but they attribute it to the “economic cycle” or “intensifying competition,” not to a paradigm shift
- Customer behavior begins to change subtly but continuously—more and more customers start trying new products/services
- You start saying “that thing is still pretty bad right now”—this sentence itself is a typical early signal of a paradigm shift
Cross-industry transfer matrix:
| Industry | Old paradigm | New paradigm | Crossover point (approx.) | Typical mistake of old-paradigm enterprises |
|---|---|---|---|---|
| Retail | Offline stores, regional chains | E-commerce, national/global platforms | 2012-2015 | “E-commerce is just a supplement; people still like to shop in person” |
| Telecom | Feature phones, voice and SMS | Smartphones, mobile internet | 2010-2012 | “The iPhone is just a toy; business people still need a BlackBerry” |
| Finance | Physical branches, collateral and guarantees | Internet finance, big-data risk control | 2015-2018 | “Internet finance isn’t reliable; risk control still has to rely on people” |
| Television | Cable TV, channel-based | Streaming, on-demand content | 2018-2020 | “Netflix has no original content; it’ll never amount to anything” |
Execution SOP:
- Identify heterogeneous competitors: List all the competitors you consider “not in the same market as us”—these are the potential paradigm disruptors.
- Draw technology-trajectory charts: Draw the technology/performance improvement trajectories of these heterogeneous competitors, along with the demand trajectories of your mainstream customers.
- Predict the crossover point: Estimate when (or whether) the heterogeneous competitors’ performance will reach the mainstream customers’ “good enough” standard.
- Assess the impact: If the crossover point really arrives, what impact will your business suffer? How much reaction time do you have?
- Formulate a response plan: Should you enter the new paradigm yourself (setting up an independent organization), acquire a player in the new paradigm, or focus on the high-end market of the old paradigm?
Layer Four: Using Up the Leftovers
Limitations:
- Not every industry will undergo such a paradigm shift—some industries have very stable paradigms (such as catering and agriculture), with no big changes for decades or even a century.
- The time scale of paradigm shifts varies enormously—some take decades, some only a few years. It’s hard to predict the timing accurately.
- The theory doesn’t explain “why some paradigm shifts failed”—not every disruptive technology ultimately disrupts the mainstream market. Many “seemingly promising” disruptive technologies ended up disappearing.
- Predicting the crossover point is very difficult—most predictions are either too early or too late.
Failure boundaries:
- In industries heavily controlled by government, a paradigm shift may be artificially delayed or even blocked.
- In industries with extremely strong network effects, the leaders of the existing paradigm may integrate the new paradigm into the old one through acquisition, imitation, and other means, avoiding disruption.
- In some industries, the old and new paradigms may coexist for a long time—rather than one completely replacing the other. For example, movie theaters and home theaters.
Critical perspective:
“The evolution of industry paradigms” is a very powerful analytical framework, but it can also easily lead to a tendency toward “fatalism”—as if paradigm shifts are inevitable and mature enterprises are doomed to fail. But in fact, a paradigm shift isn’t a “natural disaster”; it’s a “man-made disaster”—mature enterprises fail not because of the paradigm shift itself, but because they didn’t identify and respond to the paradigm shift in time. They were trapped by their own success, trapped by their own processes and values, and trapped by the opinions of their mainstream customers. Moreover, paradigm shifts aren’t “black and white” either—it isn’t “either disrupt others or be disrupted by others.” In many industries, the old and new paradigms coexist for a long time. Mature enterprises can choose to “retreat to the high-end market”—just like today’s luxury watches, high-end cameras, and luxury cars. As long as the customer group you serve still values the worth of the old paradigm, you can survive in the high-end market for a long time. So a paradigm shift isn’t “the end of the world”—it’s just that the rules of the game have changed. What you need to do is identify the new rules of the game, and then decide whether you want to play the new game, or keep playing the old game (if you still can).
Knowledge Point 15: Matching Organizational Structure to Innovation
Knowledge tier: [General layer]
Layer One: Extracting the Skeleton
Definition: Different types of innovation require different types of organizational structure to match. There’s no “best” organizational structure, only the “best-suited” organizational structure.
Christensen proposed an “innovation-organization” matching framework:
| Innovation type | Suitable organizational structure |
|---|---|
| Sustaining innovation | Functional organization, divisional organization |
| Low-end disruptive innovation | Independent spin-out organization |
| New-market disruptive innovation | Independent entrepreneurial organization |
| Breakthrough technological innovation | Heavyweight team |
Underlying logic: This is essentially the “contingency theory of organizational design”—organizational structures aren’t good or bad in themselves; there’s only the difference of whether they match the strategy. If your strategy is to do sustaining innovation, then a functional organization or divisional system is the best choice. If your strategy is to do disruptive innovation, then you must set up an independent organization—doing disruptive innovation within the existing organization is almost impossible to succeed at.
Layer Two: Digging Into the Meat
Why analysis: This is the “concluding thesis” of the whole book—gathering all the previous knowledge points onto the landing point of “organizational structure.” The ultimate solution to the innovator’s dilemma is “let the right organizational structure do the right type of innovation.” Mature enterprises fail precisely because they try to use “an organizational structure suited to sustaining innovation” to do “disruptive innovation.”
Case 1: Different Types of Teams Suit Different Types of Innovation
Let’s explain this framework with concrete examples:
1. Sustaining innovation—functional organization or divisional system
- Scenario: You want to improve the performance of an existing product, reduce the cost of an existing product, or add new features to an existing product.
- Why it’s suitable: The strength of a functional organization (divided by functions such as R&D, production, and sales) and a divisional system (divided into divisions by product or market) is “efficiency.” Their processes are clear, responsibilities are definite, and communication is smooth—very well suited to continuous improvement along a known trajectory.
- Example: Toyota continuously improving the fuel efficiency, safety, and comfort of its cars—this is typical sustaining innovation, and a functional organization works very well.
2. Low-end disruptive innovation—independent spin-out organization
- Scenario: You want to enter the low-end market, using a simpler, cheaper product to attract those over-served customers.
- Why it needs to be independent: Because your mainstream organization’s cost structure simply isn’t suited to the low-end market. If you make a low-end product within the existing organization, it gets “contaminated” by the mainstream business—the cost can never come down, and the profit margin can never meet the requirement. You need a completely independent organization that can build its own cost structure, its own processes, and its own values.
- Example: The steel company Nucor is a typical example—it used mini electric-arc furnace technology to start from the very lowest end of the rebar market and gradually disrupted the entire steel industry upward. It was a completely independent company, unburdened by the processes and values of the traditional integrated steel mills.
3. New-market disruptive innovation—independent entrepreneurial organization
- Scenario: You want to enter a brand-new market, serving people who weren’t your customers at all before.
- Why it needs to be independent: Because this new market’s customer needs, competitive rules, and success criteria are completely different from your existing market. You need an organization like a startup—small enough, flexible enough, and autonomous enough. It can’t be bound by the parent company’s processes and values.
- Example: IBM’s personal-computer team, set up in Florida, is a typical example—completely independent from IBM’s mainframe business, with its own suppliers, its own sales channels, and its own cost structure.
4. Breakthrough technological innovation—heavyweight team
- Scenario: You need to develop a brand-new technology that requires deep collaboration across multiple functional departments and will change the entire product’s architecture.
- Why a heavyweight team is needed: A functional organization runs into serious “departmental wall” problems when doing breakthrough technological innovation—each department only cares about its own part, and no one cares about the whole. A heavyweight team is a cross-functional team; its members are pulled out of the various functional departments, work full-time on the project, and report to the same project manager. Their goal isn’t “do my part well,” but “make the whole project succeed.”
- Example: When Apple developed the first-generation Macintosh computer, it used a heavyweight team. Jobs moved the team out of Apple’s headquarters and established a completely independent team, letting them focus on developing the Macintosh without interference from other departments.
Case interpretation: There’s no all-purpose organizational structure. You need to choose the appropriate organizational structure based on the type of innovation you’re doing. One key opens one lock.
Case 2: The Tragedy of a Mismatch—General Motors’ EV1 Electric-Car Project
In 1996, General Motors launched the EV1—the world’s first mass-produced pure electric car. This was a full 12 years before Tesla. But the EV1 project ultimately failed, and GM terminated it in 2003. Why?
Let’s analyze it from the angle of “organization-innovation matching”:
- Innovation type: The electric car is a typical disruptive innovation—its original target customers weren’t GM’s mainstream customers (those who want to buy high-horsepower gas cars), but a small group of strongly environmentally conscious people. Its performance (short range, slow charging) also initially fell short of the mainstream market’s requirements.
- Organizational structure: GM didn’t set up an independent organization to do the EV1. It placed the EV1 project within the existing passenger-car division, using the existing team, existing processes, and existing dealer system to run the project.
- Result:
- Cost-structure mismatch: GM’s cost structure was designed for gas cars—using this cost structure to make electric cars made the price absurdly high.
- Dealers didn’t cooperate: GM’s dealers made money from maintaining and repairing gas cars—electric cars need almost no maintenance. The dealers had no motivation at all to sell electric cars.
- Internal resistance: The EV1 project was seen as “heresy” inside GM. Most executives didn’t believe electric cars were the future, and they didn’t want to see electric cars “cannibalize” their own gas-car business.
- Ultimate failure: The EV1 produced only about 1,000 units, all of which were leased, none sold to the public. In 2003 GM terminated the project, taking back and destroying all the EV1s.
The most ironic part—12 years later, Tesla used the same technological route to achieve huge success. And what’s the biggest difference between Tesla and GM? Not the technology, but the organizational structure—Tesla was a completely independent startup, unburdened by a gas-car business, unbound by the processes and values of the old paradigm.
Case interpretation: If GM had set up a completely independent company to make electric cars back then, today’s auto industry might look completely different. Technology wasn’t the problem, money wasn’t the problem—the organizational structure, and the processes and values that come with it, were the real problem.
Case 3: A Personal-Life Scenario—What Type of Innovation Does Your “Personal Organizational Structure” Suit?
The principle of “matching organizational structure to innovation” applies just as much at the personal level. Let’s think about your “personal organizational structure”—your allocation of time, your allocation of energy, your circle of friends, your knowledge structure, your way of thinking.
- If your “personal organizational structure” is “functional”: Your time is precisely allocated, what you do each day is fixed, your circle of friends is all people like you, and your knowledge structure is concentrated in one professional field. Then what are you suited to do? You’re suited to “sustaining innovation”—continuously digging deeper in your existing field, getting better and better. But you may not be suited to “disruptive innovation”—trying entirely new fields, doing completely different things.
- If your “personal organizational structure” is “entrepreneurial”: Your time is fairly flexible, you often come into contact with people from different fields, your knowledge structure is very diverse, and you keep an open mind toward new things. Then what are you suited to do? You’re suited to “new-market disruptive innovation”—exploring entirely new fields, doing things no one has done before. But you may not be suited to “sustaining innovation”—you may lack the patience to dig deep in one field over the long term.
Many people’s problem is—they want to do “disruptive innovation,” but their “personal organizational structure” is still “functional.” They want to change, but their daily allocation of time, the people they come into contact with, and the way they think about problems are exactly the same as before. This is like GM wanting to make electric cars but still putting it in the gas-car division—it’s doomed to fail.
If you really want to do something entirely new and different, you need to adjust your “personal organizational structure”—change your allocation of time, change your social circle, change your knowledge structure, change your way of thinking. Changing only your “ideas” without changing the structure is useless.
Case interpretation: Structure determines behavior. This is true for both organizations and individuals.
Layer Three: Distilling the Essence
Trigger conditions: You may need to adjust your organizational structure when the following situations arise:
- The type of innovation you want to do doesn’t match your current organizational structure
- The innovation project “can’t get moving” within the existing organization—everyone says it’s good, but there’s just no actual progress
- The innovation project’s resources are always “grabbed away” by the mature business
- You catch yourself saying “our corporate culture doesn’t support innovation”
- The innovation project team is always complaining “headquarters doesn’t understand us” and “the approval process is too slow”
Cross-industry transfer matrix:
| Innovation type | Requirement for independence | Requirement for resources | Suitable organizational scale |
|---|---|---|---|
| Sustaining innovation | Low—can be done within the existing organization | High—needs a lot of mature resources | Large organization |
| Low-end disruptive innovation | Medium—needs some independence | Medium—needs a moderate amount of resources | A medium-scale independent team |
| New-market disruptive innovation | High—must be completely independent | Low—needs only a small amount of initial resources | A small startup team |
| Breakthrough technological innovation | Medium—needs cross-functional collaboration | High—needs a lot of R&D resources | A cross-functional heavyweight team |
Execution SOP:
- Judge the innovation type: First determine which type the innovation you want to do belongs to—sustaining innovation? Low-end disruptive innovation? New-market disruptive innovation? Or breakthrough technological innovation?
- Assess the match: Assess whether your existing organizational structure is suited to doing this type of innovation. If it doesn’t match, where’s the problem?
- Choose the right organizational structure: Based on the innovation type, choose the corresponding organizational structure—functional organization? Divisional system? Independent spin-out organization? Independent entrepreneurial organization? Or heavyweight team?
- Design isolation mechanisms: If you need to set up an independent organization, design good isolation mechanisms—physical isolation, financial isolation, process isolation, cultural isolation. Make sure the new organization won’t be “contaminated” by the old one.
- Adjust dynamically: As the innovation project develops, the organizational structure also needs to be adjusted dynamically. In the early stages a project needs a high degree of independence and flexibility; once the project matures, it can be gradually integrated back into the parent company.
Layer Four: Using Up the Leftovers
Limitations:
- This framework is relatively simple—innovation types in practice are often hybrids, not black and white. You may need a “hybrid” organizational structure.
- “Setting up an independent organization” is easy to say but hard to do. Most parent companies find it hard to truly “let go”—they can’t help wanting to “manage” the independent organization, and the result is that the independent organization loses its independence.
- An independent organization may “go out of control”—once it grows, it may come into conflict with the parent company, and may even become a competitor of the parent company.
- The framework doesn’t discuss the problem of “integration”—once the disruptive business matures, how do you integrate it back into the parent company? This is a huge challenge.
Failure boundaries:
- For very small startups (<20 people)—organizational structure doesn’t matter; be as flexible as you like.
- If you only do sustaining innovation—you don’t need to set up an independent organization; a functional organization is fine.
- In some heavily regulated industries (such as banking and pharmaceuticals)—the law may not allow you to set up a completely independent organization.
Critical perspective:
“Matching organizational structure to innovation” is a very practical framework, but it can also easily lead to a tendency toward “structural determinism”—as if, once the organizational structure is right, innovation will happen automatically. In fact, structure is only a necessary condition, not a sufficient one. People are always the most important factor—you need the right people to lead the innovation project, you need the right incentive mechanism, and you need the right culture. Moreover, this framework doesn’t discuss a more fundamental question: how do you build an organization “capable of continuously changing its own structure”? Once most organizations’ structures are set, they’re hard to change—they become rigid and can’t adapt to new types of innovation. Truly great organizations can continuously adjust their structure in response to changes in the environment. They aren’t “one structure”—they’re “a collection of multiple structures,” able to do multiple different types of innovation at the same time. That is the true “innovative organization.”
🎯 Feynman Test · Scenario Simulation Questions
Through the following 5 scenario-simulation questions, test whether you truly understand the core ideas of The Innovator’s Dilemma and can apply them flexibly to real-world scenarios.
Scenario Question 1: A Traditional Automaker’s Electric-Car Dilemma
Scenario:
You are the CEO of a century-old traditional gas-car maker. In recent years, electric cars have developed rapidly, and Tesla’s market cap has already exceeded 5 times your company’s. Inside your management there are two voices:
- Voice A: “Electric cars aren’t mature yet—short range, slow charging, expensive, incomplete infrastructure. Mainstream customers still prefer gas cars. We should keep focusing on making gas cars better, improving fuel efficiency, and reducing emissions. Once electric-car technology matures and market demand is clear, it won’t be too late for us to enter.”
- Voice B: “Electric cars are the future; they’re a disruptive technology. We must go all in now, even if we don’t make money for a while, even if it cannibalizes our own gas-car business. If we don’t do it, someone else will, and then we’ll be disrupted just like Kodak back then.”
The board has given you one year to formulate a strategy. What should you do?
Please answer the following questions:
- Which theory in this book does each of these two voices correspond to?
- Which voice do you think makes more sense? Why?
- What specific strategy would you formulate? Please explain in detail, drawing on at least 3 knowledge points from this book.
- What internal resistance do you expect to encounter during execution? How would you overcome it?
Scenario Question 2: A Bank Facing the Impact of Fintech
Scenario:
You are the head of retail business at a large state-owned bank. In recent years, various fintech companies (such as Ant Financial and Tencent Finance) have grown rapidly, using simpler, more convenient, and cheaper products to take away a large number of individual customers—especially young customers. Your team has proposed three response plans:
- Plan A: Set up a “digital bank” department, invest heavily in developing your own App, and offer features similar to those of fintech companies. This department stays within the bank’s existing system and reports to you.
- Plan B: Acquire a fintech startup that’s already doing fairly well, keep its team and brand, and let it operate independently, not folded into the bank’s existing system.
- Plan C: Focus on high-end customers. The fintech companies are all serving “low-end customers”—young people, ordinary people. We should give up these customers and focus on serving high-net-worth customers, providing them with personalized, high-end financial services.
Which plan should you choose? Or do you have a better plan?
Please answer the following questions:
- Use this book’s theory to analyze the pros and cons of these three plans.
- Which plan has the highest probability of success? Why?
- If you choose Plan A, what problems are you most likely to encounter during execution? How would you avoid them?
- If you choose Plan C, can this “retreat to the high-end market” strategy succeed over the long term? Why?
Scenario Question 3: University Education Facing the Challenge of Online Education
Scenario:
You are the president of a top university. Online-education platforms (such as Coursera, edX, and domestic MOOC platforms) are developing rapidly, offering free or low-cost high-quality courses and attracting tens of millions of students. Many people say “the traditional university model is about to be disrupted.” Your professors have different views:
- Professor A: “Online education can never replace the traditional university. Students need more than knowledge—they need face-to-face interaction with professors, discussion with classmates, the campus atmosphere, and the signaling value of a diploma. These are all things online education can’t provide. We don’t need to worry; just keep doing our teaching and research well.”
- Professor B: “Online education is a disruptive technology. We should go all in on online education right now, put our courses online for free, and let people all over the world learn. Even if this affects our admissions, we must do it. Otherwise we’ll be eliminated in the future.”
As president, how should you decide?
Please answer the following questions:
- Do you think online education is a “sustaining technology” or a “disruptive technology” for the traditional university? Why?
- Does Professor A’s view make sense? What might he be overlooking?
- Does Professor B’s view make sense? What might he be overlooking?
- Drawing on this book’s theory, what response strategy would you formulate?
Scenario Question 4: The Innovator’s Dilemma Inside a Big Company
Scenario:
You are a mid-level manager at a large tech company. You’ve spotted a new market opportunity with great potential, but this opportunity has little to do with the company’s existing business—existing customers don’t need this product, and the profit margin looks much lower than the existing business.
You pitched the idea to your boss. Your boss said, “The idea sounds good, but our focus right now is serving existing customers well and improving the profit margin of the core business. Your project won’t show returns in the short term, so let’s set it aside for now.”
You know your boss has a point—the company’s KPI is revenue and profit growth, and your project really would “divert resources.” But you also believe that if you don’t do it now, in the future this market may be occupied by startups, and by then it’ll be too late for the company to enter.
What should you do?
Please answer the following questions:
- Which core problem in this book does this scenario reflect?
- If you insist on doing this project, what resistance might you encounter?
- Drawing on this book’s theory, what methods could you use to “bypass” this resistance and push the project forward?
- If you were the CEO of this company, how would you design the company’s mechanisms to encourage this kind of “seemingly unprofitable but long-term strategically valuable” innovation?
Scenario Question 5: The “Innovator’s Dilemma” of Personal Career Development
Scenario:
You’ve worked at a company in a traditional industry for 10 years, have already reached mid-level management, earn a decent income, and have a stable job. But you increasingly feel this industry is on the decline and may be disrupted by a new technology or model in the next 5-10 years. You’re considering whether to switch to an emerging industry.
But you have many concerns:
- You’ve accumulated 10 years of experience in this industry; if you switch, most of that experience will be useless.
- You’re already 35; switching means starting from scratch and competing with people in their 20s.
- You have a mortgage and children, and heavy financial pressure, so you can’t bear the risk of a big drop in income.
- You’re not sure whether the emerging industry really has a future—could it be a bubble? Could it also fail in a few years?
Your friends and family all advise you “not to rock the boat”—“Your current job is so stable; so many people would envy it. Switching is too risky.”
What should you do?
Please answer the following questions:
- What similarities does this personal career dilemma have with the enterprises’ innovator’s dilemma in this book?
- Using this book’s theory, analyze: what are the risks of the choice “don’t switch; keep developing in the existing industry”?
- Using this book’s theory, analyze: what are the risks of the choice “switch to an emerging industry”?
- Drawing on this book’s theory, can you design a third plan with “controllable risk”? (Hint: think about the “independent organization” approach)
📝 A Critical Summary of the Whole Book
The Innovator’s Dilemma is one of the most important business works of the past 30 years. The “disruptive innovation” theory proposed by Christensen fundamentally changed our understanding of innovation, competition, and the rise and fall of enterprises. But like all great theories, it isn’t perfect; it too has its boundaries of applicability and its limitations.
Let’s summarize this book from both the positive and the negative sides:
This Book’s Great Contributions
1. It explained a counterintuitive truth—good management is precisely the cause of failure
Before Christensen, people generally believed that enterprises failed because of poor management—bureaucracy, complacency, lack of foresight, weak execution, and so on. But Christensen proved with a wealth of detailed cases that it is precisely those “well-managed” enterprises—earnestly listening to customers, actively investing in high-return projects, focusing on the mainstream market, and continuously improving product performance—that are most likely to fail in the face of disruptive innovation.
This is a subversive insight. It tells us: in the world of innovation, the factors that lead to success and the factors that lead to failure are the same thing. What makes you successful today may make you fail tomorrow. This is a very profound, tragic insight.
2. It provided a complete, operable theoretical framework
The Innovator’s Dilemma isn’t a “chicken-soup book”—what it gives you isn’t empty slogans like “have an innovative spirit” or “embrace change.” What it gives you is a complete, operable theoretical framework:
- The distinction between disruptive technology and sustaining technology
- Value-network theory
- The Resources-Processes-Values (RPV) framework
- The tendency to migrate to the high-end market
- The mismatch between performance supply and market demand
- The independent-organization solution
This framework can be applied directly to business practice. Whether you’re a CEO, an entrepreneur, or a product manager, you can use this framework to analyze your industry, your company, and your product.
3. It changed the entire business world’s view of innovation
Before Christensen, “innovation” mainly referred to “technological innovation”—inventing better technology and making better products. After Christensen, people realized: innovation isn’t just a technological question; it’s even more an organizational question, a structural question, and a question of values. Much of the time, you don’t need to invent new technology—you just need to establish the right organizational structure, letting the right people do the right things in the right way.
You could say Christensen redefined the concept of “innovation.”
This Book’s Limitations and Criticisms
1. It has insufficient explanatory power for “non-technological disruption”
Almost all the cases in The Innovator’s Dilemma come from manufacturing—hard drives, excavators, steel, cars. Disruption in these industries is mainly “technology-driven.” But in today’s business world, more and more disruption is “model-driven,” “ecosystem-driven,” and “platform-driven”—such as Uber’s disruption of the taxi industry, Airbnb’s disruption of the hotel industry, and the platform economy’s disruption of traditional industries.
Christensen’s theory has clearly insufficient explanatory power for these “non-technological disruptions.” He later tried to extend his theory (proposing concepts like “empowering disruption” and “market-creating disruption”), but none was ever as persuasive as the original “disruptive innovation” theory.
2. It has a tendency toward “structural determinism”
Christensen strongly emphasized the importance of organizational structure. He repeatedly said: when facing disruptive innovation, you must set up an independent organization. This is of course right, but he seemed to imply—as long as you set up an independent organization, you’ll definitely succeed.
But in fact, structure is only a necessary condition, not a sufficient one. The human factor, leadership, culture, timing, and luck are all very important. Historically, many enterprises set up independent organizations but still failed in the end.
Moreover, “structural determinism” can also easily lead to pessimism—as if mature enterprises are doomed to fail in the face of disruptive innovation. But in fact, some mature enterprises did successfully respond to disruptive innovation—such as Microsoft responding to the internet, Intel responding to RISC chips, and Amazon responding to cloud computing.
3. It underestimated the “first-mover disadvantage” and the “late-mover advantage”
Christensen’s theory implies that emerging enterprises always have an advantage and mature enterprises always have a disadvantage. But in fact, “moving first” often has disadvantages too—you need to educate the market, you need to bear huge trial-and-error costs, and you may take many detours. Whereas “late movers” can absorb the lessons of the first movers, entering at the right time, and are instead more likely to succeed.
For example, Apple wasn’t the first company to make a smartphone (BlackBerry and Palm were), but Apple made the smartphone work. Google wasn’t the first company to make a search engine (Yahoo and AltaVista were), but Google made the search engine work.
4. It discussed “continuous disruption” insufficiently
The disruptive innovations Christensen discussed were mostly “one-time”—a new paradigm replaces the old one, and then the new paradigm stabilizes. But in many industries today, disruption is “continuous”—a new paradigm is replaced by an even newer one shortly after it’s established.
For example, in the consumer-electronics and internet industries, the pace of paradigm shifts is getting faster and faster. In this environment, what enterprises need isn’t “to respond to one disruptive innovation,” but “to build the capacity to continuously respond to disruptive innovation.” Christensen’s theory discusses this point insufficiently.
How Should We Read This Book Today?
The Innovator’s Dilemma was published in 1997, more than 25 years ago. Over these 25 years, the business world has undergone earth-shaking changes. So how should we read this book today?
My advice is:
1. Treat it as a “thinking tool,” not an “operation manual”
Don’t dogmatically apply the book’s conclusions—“as long as it’s disruptive innovation, you must set up an independent organization.” The world is much more complex than 25 years ago. You should treat the book’s theories as “thinking tools”—use them to analyze problems and inspire thinking, not as standard answers.
2. Focus on learning its “analytical method,” not its specific conclusions
The most valuable thing about this book isn’t the conclusion of “disruptive innovation,” but the analytical method by which Christensen arrived at that conclusion. How did he abstract common patterns from a large number of seemingly unrelated cases? How did he challenge the common sense that “good management leads to success”? How did he dig deeper and deeper, from phenomenon to essence?
Learning this analytical method is far more important than memorizing a few conclusions.
3. Read it together with other theories
The Innovator’s Dilemma isn’t the whole of innovation theory. You should read it together with other theories—such as Christensen’s own later The Innovator’s Solution, Brian Arthur’s The Nature of Technology, Clay Shirky’s Cognitive Surplus, and Eric Ries’s The Lean Startup.
Putting different theories together, comparing and complementing each other, is how you form a more complete and deeper understanding.
Final Words
The Innovator’s Dilemma is essentially a book about “change.” It tells us: change doesn’t come from the outside, but from the inside—your past success is your greatest future obstacle.
This applies not only to enterprises, but also to individuals.
Each of us has our own “sustaining trajectory”—we continuously improve in the field we’re familiar with, getting better and better. But we all also face our own “disruptive innovation”—new technologies, new models, new industries, which at first look “not as good as what we do now,” but which keep improving and will one day disrupt our lives.
Facing this change, what should we do?
Christensen’s answer for enterprises is: disrupt yourself before you are disrupted. Set up an independent organization, and use a new organizational structure, new processes, and new values to do new things.
This applies just as much to each of us.
In this era of rapid change, maybe each of us should set up an “independent organization” within our own hearts—that part of you that can temporarily set aside past experience, past identity, and past success, to try new things, learn new skills, and explore new possibilities.
Don’t let yesterday’s you trap tomorrow’s you.
This, perhaps, is the best inspiration that The Innovator’s Dilemma gives each of us.
评论
0 条评论