Research Idea Discovery Flywheel
Turn community conversations into sharper research and continuous feedback
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 97%
The Research Idea Discovery Flywheel converts trusted relationships into a recurring research process. One-to-one conversations, online discussions, and in-person gatherings expose what informed people are noticing, fearing, or struggling to explain. Those signals help the researcher choose questions worth investigating rather than reacting indiscriminately to daily events. When expertise is missing, the researcher consults a knowledgeable community member while retaining responsibility for the final judgment. The resulting analysis is then shared back with the community, whose criticism reveals omissions and generates the next round of questions. Each cycle strengthens relationships, improves topic selection, and makes subsequent research more relevant without requiring the researcher to pretend to know everything.
Origin
Extracted from Masters in Business, where Jawad Mian described how Stray Reflections combines writing, salons, events, group discussion, and one-to-one relationships.
Core principles
- 01All participants collectively know more than any individual.
- 02Questions and concerns reveal valuable research gaps.
- 03Experts should supplement rather than threaten independent judgment.
- 04Published analysis improves when it returns to the community for challenge.
- 05Strong relationships create better information flow.
How to run it
- 1
Listen for consequential uncertainty
Use candid conversations to identify what investors are concerned about, debating, or taking for granted. Prioritize questions whose answers could materially change decisions.
Pro tip Track repeated concerns across several independent conversations.
Watch out Do not confuse the loudest opinion with the most important question.
- 2
Locate informed perspectives
Find community members with direct expertise or differentiated evidence. Invite them into a focused conversation instead of asking them to pronounce on everything.
Pro tip Seek practitioners with different geographic, professional, or ideological vantage points.
Watch out Expert status does not eliminate bias or guarantee a correct conclusion.
- 3
Investigate independently
Test the community's claims against data, history, and competing explanations. Preserve analytical independence rather than adopting a bullish, bearish, consensus, or contrarian identity.
Pro tip Write down what evidence would invalidate each leading explanation.
Watch out Do not turn community intelligence into consensus imitation.
- 4
Publish a clear synthesis
Explain the question, evidence, mechanism, and conclusion in a reflective research piece. Focus on what readers need to understand rather than reacting to every event.
Pro tip State uncertainties and knowledge limits explicitly.
Watch out Avoid presenting borrowed expertise as firsthand knowledge.
- 5
Return for challenge
Share the conclusion with the community and invite disagreement. Use the response to strengthen the analysis and identify the next research question.
Pro tip Ask critics specifically what you may be missing.
Watch out Feedback only improves the flywheel when dissent is genuinely welcome.
In the wild
After a Singapore dinner revealed unexpectedly strong agreement that a US recession was inevitable, Mian revisited his labor-market thesis. He investigated sectoral hiring, construction employment, job openings, labor supply, and demand conditions before returning to the community with a more robust argument for secular labor tightness.
→ A surprising consensus became the input for deeper analysis rather than an opinion to accept or dismiss reflexively.
An investment writer repeatedly hears portfolio managers asking about a new biotechnology platform. The writer interviews a specialist, tests the specialist's claims against clinical and commercial evidence, publishes a synthesis, and asks experienced readers to challenge the assumptions.
→ The writer produces relevant research without pretending to be a biotechnology expert.
Common mistakes
Treating the community as a poll
Counting opinions produces consensus, not insight. The researcher must investigate why informed participants disagree and independently evaluate the evidence.
Performing expertise
Pretending to know every domain weakens trust and introduces hidden errors. Admit knowledge gaps and find someone closer to the subject.
Ending the cycle at publication
Publishing without inviting criticism removes the feedback stage that makes the process cumulative.
Is it for you?
Best for
It is best for analysts, advisers, writers, and community leaders who can maintain trusted relationships with informed practitioners.
Not ideal for
It is not ideal when participants cannot speak candidly or the researcher merely repeats consensus views.
From the transcript
“I'm a big believer in building strong relationships, and that's kind of what develops the research idea of Flywheel.”
“Again, this idea that all of us are smarter than any of us,”
“And so that's kind of how the process works, the research idea discovery flywheel works.”
From the episode
Jawad S. Mian on Global Investment Research
Jawad S. Mian