Primary-Research Pattern Synthesis
Interview practitioners, compare cases, and extract recurring patterns
- Difficulty
- Advanced
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 98%
Primary-Research Pattern Synthesis begins with a consequential question whose answer is unclear from one person's experience. The researcher identifies companies or practitioners with direct evidence, asks each for concrete stories, and collects cases in a common structure. Rather than beginning with a predetermined conclusion, the researcher compares the cases and allows recurring mechanisms to emerge. Outliers are retained as context but are not presented as universal rules. The final output compresses many hours of interviews, outreach, comparison, and editing into an accessible answer for the audience. This mechanism creates differentiated content because the raw evidence is not readily available elsewhere, while also reducing the risk of generalizing from a single famous company. Its value comes from access, breadth, synthesis, and the willingness to do costly work on behalf of readers.
Origin
After leaving Airbnb, Lenny was repeatedly asked how to build marketplace companies. Unsure whether Airbnb's approach was universal, he interviewed major marketplaces, identified cross-company patterns, and turned the research into a five-part series.
Core principles
- 01Do not assume one company's experience is universal
- 02Ask practitioners who directly experienced the problem
- 03Compare several relevant cases before drawing conclusions
- 04Let patterns emerge from the evidence
- 05Perform expensive research once and make the synthesis accessible
How to run it
- 1
Frame the research question
Choose one question that matters to the audience and cannot be answered confidently from your own experience. Define the scope tightly enough that cases can be compared.
Pro tip A question about causes or mechanisms produces stronger synthesis than a request for general advice.
Watch out Do not embed your preferred conclusion in the question.
- 2
Build the evidence list
List the companies, teams, or practitioners whose experiences would provide the most revealing cases. Seek variation while preserving relevance to the same question.
Pro tip Start with the cases you would most want to see as a reader.
Watch out Convenient contacts alone may create a biased sample.
- 3
Collect concrete accounts
Contact knowledgeable people with a concise question and ask what actually happened. Capture actions, context, and outcomes rather than broad theories.
Pro tip Follow vague claims with requests for a specific example.
Watch out Do not treat polished company narratives as complete evidence.
- 4
Compare cases
Place the stories side by side and mark repeated mechanisms, meaningful differences, and outliers. Delay the conclusion until the evidence has been assembled.
Pro tip Use a table with consistent fields for context, action, and result.
Watch out A memorable anecdote is not necessarily a recurring pattern.
- 5
Extract qualified patterns
Describe the mechanisms that recur and specify the conditions under which they appear to work. Distinguish strong patterns from tentative observations.
Pro tip Show enough contrasting cases for readers to understand the boundaries.
Watch out Do not claim causation that the collected accounts cannot support.
- 6
Compress the value
Edit the research into a concise answer using miniature case studies, bullets, charts, or tables. Make the costly investigation inexpensive for the audience to consume.
Pro tip Preserve concrete details while removing repetitive explanation.
Watch out Excessive compression can erase the context needed to apply a pattern responsibly.
In the wild
Rather than treating Airbnb's methods as the definitive playbook, Lenny interviewed leading marketplace companies and searched for patterns across their experiences. He synthesized the findings into a five-part series about kick-starting and scaling marketplaces.
→ The series provided a broader and more credible answer than a single-company retrospective and performed strongly with readers.
For a post about growth inflections, Lenny listed companies whose stories he wanted to understand, contacted people he knew inside them, asked what caused their growth change, and reviewed the collected cases to see what emerged.
→ The resulting post gave readers behind-the-scenes evidence from multiple companies.
Common mistakes
Generalizing from one company
A successful company's method may have worked only because of its context or even in spite of its decisions.
Starting with the answer
A fixed conclusion encourages selective evidence gathering and prevents unexpected patterns from emerging.
Accepting vague responses
General theories provide little reusable value unless they are grounded in concrete actions and outcomes.
Is it for you?
Best for
It is best for experts with access to practitioners who can contribute concrete examples around a shared question.
Not ideal for
It is not ideal for urgent topics where there is insufficient time or access to collect several trustworthy cases.
From the transcript
“I just talked to all the biggest marketplace companies and see what they did, and see if there's any patterns that emerge across all marketplace…”
“I'm just kind of doing primary research on behalf of people”
“So the way I start there is just make a list. Here's the companies I would love to capture stories from, and then I just…”
From the episode
How Lenny Rachitsky Makes +$500,000/Year As a Content Creator
Lenny Rachitsky