TL;DR: Jack Ma’s story is one of the most circulated pieces of motivational content, and the way it is retold is nearly always identical: a run of rejections, then the largest public offering in history, therefore do not give up. That inference commits a specific error the management literature has named. Jerker Denrell’s 2003 paper in Organization Science showed that the organisations we can observe at any point in time are the survivors of a selective process that has eliminated a large fraction of the underlying population, and that because books and the business press focus on the successful, our samples systematically undersample failure. The consequence is technical and devastating: risky practices with no relationship to performance in the full population can still appear positively related to performance within a sample of survivors. So “he did not give up and he won” does not show that not giving up wins. It shows that the winners did not give up, which is a different claim and useless when you are deciding what to do. This update does not remove the story. It separates it into two columns: corporate facts the public record verifies, and biographical figures that rest only on the narrator’s own account. Below: that claim audit, an analysis of exactly how survivorship bias applies here, and the three lessons the story can genuinely support.
The difference between managing yourself like a CEO and being moved by an inspiring story comes down to one question: should this information change my decision, and if so, why?
Jack Ma’s story is an ideal test case, because it is genuinely extraordinary and most of the lessons drawn from it are logically invalid. Both things can be true at once, and here they are.
What the public record holds
Start with what is not in dispute. Alibaba was founded in Hangzhou in 1999. The company listed on the New York Stock Exchange under the ticker BABA in September 2014. Priced at 68 dollars a share, the offering initially raised 21.8 billion dollars; when the underwriters exercised their overallotment option the total reached 25 billion dollars, making it the largest initial public offering to that date and surpassing the Agricultural Bank of China’s 22.1 billion dollar Hong Kong listing in 2010.
These figures sit in exchange records and in the company’s filings with the securities regulator. They are verifiable. This is the kind of information a decision can be built on.
What the public record does not hold
The story’s real circulating power lies not in those figures but in the rejection list: 30 job applications refused, 10 applications to Harvard and 10 rejections, 24 of 25 KFC applicants hired and not him, four of five police applicants accepted and not him.
What these accounts share is that all of them come from the narrator. They were recounted in interviews and talks, repeated widely, and cannot be confirmed against an independent record. That does not make them false. It means they belong to a different class of evidence, and an article that presents both with equal confidence misleads its reader.
Table 1. Audit of the story’s claims by class of evidence, CEOtudent editorial framework
| Claim | Class of evidence | Independently verifiable | Usability for a decision |
|---|---|---|---|
| Founded in Hangzhou in 1999 | Corporate record | Yes | High |
| 2014 New York Stock Exchange listing at 68 dollars a share | Exchange and regulatory record | Yes | High |
| 25 billion dollars with the overallotment, largest to date | Exchange and regulatory record | Yes | High |
| Rejected from 30 job applications | Own account | No | Low |
| Ten applications to Harvard, ten rejections | Own account | No | Low |
| The only applicant KFC did not hire | Own account | No | Low |
| Learned English by guiding tourists for free | Own account, consistently repeated | No | Low, though the mechanism is plausible |
The point of this table is not to diminish the story. It is to show which row you can build a decision on. The top three establish a fact. The bottom four establish a narrative. Both are interesting; only one is evidence.
The real error: survivorship bias
The evidence-class distinction matters but is secondary. The deeper problem is that even if every claim were true, the lesson drawn from the story would still be invalid.
Denrell’s work sets out the mechanism clearly. The organisations we can observe are the survivors of a process that eliminated most of the population. Because business books and the business press focus on the successful, the sample in front of us undersamples failure. The result Denrell demonstrates is that under these conditions, risky practices unrelated to performance across the full population can appear positively related to performance within the surviving sample.
Apply that here. The trait “rejected repeatedly but did not give up” looks associated with winning when you only examine winners. You do not see how many people carried the same trait and did not win, because nobody writes about them. Until that number is known, you can say nothing about whether persistence raises the probability of success.
Table 2. Which question the story actually answers, CEOtudent editorial framework
| Question | Does the story answer it | Why |
|---|---|---|
| Were the winners persistent | Yes | Direct observation |
| Do persistent people win | No | The persistent losers are absent from the sample |
| Does persistence raise the probability of success | No | There is no comparison group |
| Is this individual’s path repeatable | No | Single case, uncontrolled conditions |
| Was there a real timing opportunity | Partly | Founding date and market conditions are verifiable |
The third row is what motivational content sells and cannot deliver. Learning like a student means being able to make exactly this distinction.
The three things the story can genuinely support
None of the above makes the story useless. It only narrows what you take from it. Three things survive, and they are real.
First, investing early and for a long time in a skill that compounds. The most concrete detail in the story is years of sustained language practice with no direct payoff. That is not a lesson about motivation; it is a lesson about compounding. English proficiency was a filter that, in late-1990s China, drastically narrowed the pool of people able to understand internet business models. The investment paid off years later in an entirely different field.
This generalises because the mechanism is not personal: if you invest in a skill before it is understood to be valuable, you become one of few people holding it when it is.
Second, observing a gap directly. The turning point in the story is typing a word into a search engine and not finding what was expected. That is the most primitive and most reliable form of market research: seeing for yourself that the thing you are looking for does not exist. Not a gap a report told you about, but one you walked into.
Third, treating timing as a variable separate from personal quality. Founding an internet marketplace in China in 1999 is not the same undertaking as founding one in 2015. The unrepeatable part of the story lives largely here, and any account that ignores it credits a personal quality with an outcome that personal quality did not produce.
Apply this to your own decisions
The practical value of this piece is not about Jack Ma. It is about the next success story you read.
When a story recommends a behaviour to you, three questions follow. How many people performed this behaviour and failed, and where could I learn about them? Are the reported facts independently verifiable, or do they rest solely on the protagonist’s account? How much of the outcome belongs to the person, and how much to the time and place?
If you cannot answer all three, the story may be inspiring but it is not a decision input. Confusing the two is one of the more expensive mistakes in managing yourself.
What changed in this piece
The previous version of this article presented the story as a single narrative, giving self-reported figures and verifiable corporate facts the same level of confidence. Two changes were made.
Classes of evidence were separated, and which claim belongs to which class is now stated explicitly. Figures with no traceable source are no longer presented as facts.
The “never give up” inference was replaced with the management literature explaining why that inference is invalid, and with three mechanisms the story genuinely supports. The aim is neither to defend nor debunk the story, but to extract what is valid in it.
Frequently asked questions
So is the story fabricated?
No, and no such claim is made here. The corporate facts are all verifiable. The biographical rejection figures rest on the narrator’s own account and cannot be confirmed against an independent record; that does not make them false, it places them in a different class of evidence.
Is reading success stories pointless?
No, but it depends what you read them for. Reading to learn a mechanism is useful: which skill was invested in and when, how the gap was noticed. Reading to learn a probability is misleading, because the comparison group is invisible by definition.
Is being persistent a bad thing?
That does not follow. What follows is that a sample of survivors tells you nothing about the returns to persistence. When persistence is rational is a separate question, and the answer lies in comparing its cost against its expected return, not in the inspiring tone of a story.
Why is the Alibaba offering considered so significant?
Its size. It priced at 68 dollars a share, initially raised 21.8 billion dollars, and reached 25 billion dollars once the overallotment option was exercised, making it the largest initial public offering to that date; the previous record was 22.1 billion dollars in 2010.
How do I apply this framework to my own work?
Before copying a success example, ask the three questions: do I know how many people did this and failed, can I independently verify the reported facts, and how much of the outcome belongs to timing. If you cannot answer all three, keep the example as inspiration, not as a plan.
Sources
- Denrell, Vicarious Learning, Undersampling of Failure, and the Myths of Management, Organization Science, volume 14, issue 2, 2003, pages 227 to 243
- Alibaba Group Holding Limited, initial public offering prospectus and related regulatory filings, United States Securities and Exchange Commission, 2014
- New York Stock Exchange, trading records for the security under the ticker BABA, September 2014
- March, Sproull and Tamuz, Learning from Samples of One or Fewer, Organization Science, volume 2, issue 1, 1991
- Taleb, Fooled by Randomness: The Hidden Role of Chance in Life and in the Markets, Random House, 2004
This content was compiled with the support of AI following in-depth research, then written and prepared for publication by the CEOtudent editorial team.
This post is also available in:
