Research-First Nonfiction Writing
Gather deeply, find the ending, then write toward the story
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 98%
Michael Lewis separates long-form nonfiction into a long material-gathering phase and a shorter writing phase. He reports and researches for roughly three times as long as he drafts, while continuously considering the eventual shape of the story. Before beginning the manuscript, he tries to know where the narrative ends and may already have its final sentence and title. Drafting then becomes a test of the imagined structure: material that seemed important can feel boring or unnecessary in context, while a minor observation may expand into a major section. Research gradually tapers as writing takes over. The mechanism converts abundant real-world evidence into a focused narrative by delaying prose, defining a destination, and allowing discoveries on the page to override the original outline.
Origin
Extracted from Masters in Business during Michael Lewis's explanation of how he researches and writes narrative nonfiction.
Core principles
- 01Spend substantially more time gathering material than drafting
- 02Keep shaping the story mentally throughout the research
- 03Know the likely destination before drafting the opening
- 04Let the written story determine what expands or disappears
- 05Treat titles as tools for crystallizing the central idea
How to run it
- 1
Gather the World
Spend the majority of the project reporting, interviewing, observing, and collecting source material rather than drafting from assumptions.
Pro tip Budget roughly three units of research time for every unit of writing time on a substantial book.
Watch out Starting too early can lock the story around incomplete evidence.
- 2
Shape the Story Mentally
Think continuously about the narrative while gathering material. Look for the characters, tensions, causal relationships, and unanswered questions that can organize the evidence.
Pro tip Treat every interview as both information gathering and a test of the emerging story.
Watch out Do not confuse a large collection of facts with a coherent narrative.
- 3
Find the Destination
Determine where the story is likely to end before beginning the manuscript. A prospective final sentence can provide direction without requiring a rigid outline.
Pro tip Use the ending to judge which events and characters genuinely belong.
Watch out The destination should guide the draft, not prevent revision when the evidence points elsewhere.
- 4
Crystallize the Idea
Develop a title or concise phrase that captures what the book is fundamentally about. Use it as a conceptual filter during drafting.
Pro tip Look for a phrase that carries both the literal subject and the deeper transformation or conflict.
Watch out A clever title is not useful if it misrepresents the story.
- 5
Draft and Taper Research
Begin writing while finishing only the remaining research required by the developing manuscript. Gradually shift all attention to drafting.
Pro tip Research specific gaps exposed by the prose instead of continuing open-ended investigation.
Watch out Endless research can become avoidance once the story is ready to be written.
- 6
Let the Page Reallocate Space
Cut material that loses relevance in context and expand details that reveal unexpected narrative importance. Allow the manuscript's needs to override the initial plan.
Pro tip Judge every section by what it contributes to the reader's experience, not by how difficult it was to report.
Watch out Do not preserve material solely because it once seemed important.
In the wild
Lewis spent years developing access to Daniel Kahneman and learning unfamiliar material about psychology and Israel. Before drafting, he had identified the intellectual relationship at the center of the story, found an appropriate title, and had the last sentence substantially in mind. During writing, planned sections fell away while seemingly short material expanded.
→ A complex intellectual history became a character-driven narrative about collaboration, friendship, and the transformation of ideas.
A writer spends twelve weeks interviewing a founder, employees, customers, and critics before drafting. She identifies the company's failed launch as the ending, selects a title that captures its strategic reversal, and then writes. A planned technology section is cut when it stalls the narrative, while a short customer anecdote expands into the central scene.
→ The final profile follows a clear evidence-based arc rather than reproducing a chronological research dump.
Common mistakes
Drafting before reporting
Beginning with insufficient material encourages the writer to defend an early interpretation rather than discover the real story.
Treating the outline as fixed
Material can shrink or grow once it appears in narrative context, so preserving the initial allocation can make the book dull or distorted.
Researching forever
Once drafting reveals only narrow factual gaps, continuing broad research delays the transition to a finished manuscript.
Is it for you?
Best for
It is best for journalists and nonfiction authors developing complex, character-driven books or substantial articles.
Not ideal for
It is not ideal for time-sensitive reporting that must be published before extensive research and narrative development are possible.
From the transcript
“I spend three times in much time I'm gathering the material as I do actually writing it.”
“by the time I sit down to actually start a book, I at least think I know exactly where it ends.”
“Things drop that I thought were important you see in print, and they no longer resonate”
From the episode
Interview With Michael Lewis: Masters in Business (Audio)