Question-Led Access Method
Earn insider access by investigating the question your subject already cares about
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 95%
The Question-Led Access Method begins with a contradiction that fascinates both investigator and subject. A shared question creates a relationship based on inquiry rather than publicity: the subject wants the answer too. The investigator remains honest about uncertainty, avoids prematurely fixing the final format, and seeks repeated access to real working environments. Time is essential because useful evidence appears after participants stop treating the observer as an event. Comparative reporting then tests whether the apparent insight is truly distinctive or merely common practice. The method grants subjects factual fairness but preserves the investigator’s independence over interpretation and publication. Its output is not simply richer quotation; it is behavioral evidence, system context, and a defensible explanation of how the subject differs from alternatives.
Origin
Michael Lewis explained on Masters in Business that Billy Beane welcomed his central Moneyball question because it was the same puzzle occupying Oakland’s front office, enabling Lewis to remain around the team and investigate it deeply.
Core principles
- 01Lead with a real question rather than a predetermined narrative
- 02Choose a question that is intrinsically interesting to the subject
- 03Be transparent about uncertainty while the project is still forming
- 04Spend enough time inside the setting for performed behavior to fade
- 05Compare neighboring cases before declaring the subject exceptional
How to run it
- 1
Form the shared question
Express the central contradiction in plain language and present it to the subject. Determine whether it is also a question the subject genuinely wants answered.
Pro tip Questions about unexplained performance gaps often expose more than requests for personal history.
Watch out A flattering premise may secure an interview but will not necessarily sustain candid access.
- 2
Disclose current uncertainty
Explain what you know, what you are investigating, and what outputs are currently possible. Avoid promising a narrative before the evidence exists.
Pro tip Distinguish uncertainty about format from uncertainty about ethical boundaries.
Watch out Do not misrepresent a definite publication plan as casual curiosity.
- 3
Negotiate repeated proximity
Seek permission to observe ordinary work, decisions, and interactions over time. Prioritize access to behavior over a sequence of formal interviews.
Pro tip Ask for low-disruption ways to remain present during routine activities.
Watch out Access does not remove obligations concerning privacy, consent, or off-the-record material.
- 4
Become familiar, not embedded
Stay long enough that participants pay less attention to your presence while maintaining analytical independence. Record concrete actions as well as retrospective explanations.
Pro tip Note when observed behavior diverges from what participants say they do.
Watch out Familiarity can turn into advocacy if you stop testing the subject’s claims.
- 5
Build comparison cases
Observe relevant peers and alternatives to determine whether the focal subject is actually distinctive. Use comparison to refine the causal explanation.
Pro tip Select comparison cases that face similar constraints but produce different outcomes.
Watch out Without comparison, ordinary practices can be mistaken for unique innovations.
- 6
Retain narrative independence
Verify facts and represent people fairly, but keep control over conclusions and final framing. Do not let access become an implicit veto over the work.
Pro tip Set expectations about review and publication boundaries before the material becomes sensitive.
Watch out Surprising a subject with avoidable factual errors damages both trust and accuracy.
In the wild
Lewis asked how Oakland could compete while spending one fifth of what the Yankees spent. Beane said it was the question he thought about constantly, giving both parties a reason to continue. Lewis then spent time with Oakland and several other clubs before concluding that the A’s approach was genuinely different.
→ Sustained access and comparative reporting revealed the system behind Moneyball.
A researcher investigating unusually high customer retention asks the team how it succeeds with fewer staff than competitors. Because the team is equally puzzled by which practices matter most, it permits repeated observation. The researcher then visits three comparable teams before attributing the result to a specific triage routine.
→ The final case study rests on observed behavior and comparative evidence rather than executive claims.
Common mistakes
Pitching a finished thesis
A predetermined explanation turns interviews into confirmation exercises and gives the subject little intellectual reason to collaborate.
Relying on formal interviews
People describe idealized behavior when questioned directly. Repeated observation reveals decisions and contradictions that interviews miss.
Mistaking access for validation
Deep access can make a subject feel exceptional before comparisons have been conducted. Study peers before making the central claim.
Is it for you?
Best for
Journalists, documentarians, biographers, and qualitative researchers investigating a system through its participants.
Not ideal for
Research requiring anonymity, rigid experimental controls, or immediate publication before trust can develop.
From the transcript
“This often happens too, that you have a question and the question is equally interesting to the subject.”
“How you how you get them to let you hang around? That's the important thing. You've got to hang around.”
“I had to do that to know that they were special, like know that nobody else was doing this”
From the episode
Bonus: An Evening with Michael Lewis, From "Liar’s Poker" to Today