Brainwriting
Write separately, pass the sheets and build on earlier ideas while retaining each contribution's provenance.
Use the written form of brainstorming when a group needs varied contributions before discussion. Record ideas independently, pass the sheets, then clarify and screen them in a separate phase.
One shared question, participants and facilitator, numbered sheets, time and passing rules, and known constraints.
An idea pool with sheet, round, author and parent records, plus questions for clarification and later screening.
How to use it
- Agree one open question and explain the separate writing, passing, clarification and screening phases.
- Give each person a numbered sheet and let them write independently without critique or voting.
- Pass in the agreed direction; retain earlier rows and add an extension, variation or new idea with its parent ID.
- Continue for the agreed rounds, retaining blanks and duplicates; clarify and group the records without deleting provenance.
- Identify questions and screening criteria before proposing a small test.
Keep its limits in mind
This guide uses brainwriting, not every form of spoken brainstorming. Three people, three rounds and one idea per cell are this example's settings, not 6-3-5. Record counts and reuse do not measure independent ideas, demand or people's ability. Screen constraints after generation.
Tracing nine records through three sheets
Fictional teaching example: Starbridge Library Service explores after-hours returns. C4 contains 20 synthetic visits on Oct 12–14, 2026 (12 during staffed hours, eight outside), three people's writing records, pilot estimates and a hypothetical monthly model. People, permissions, prices and behavior are invented. No pilot or paid service has run.
On a small screen, scroll across the diagram to read it.
Trace contributions: S1 passes P1→P2→P3, S2 P2→P3→P1, and S3 P3→P1→P2. Each person writes one entry per round. Original and parent IDs survive grouping.
Count carefully: Three people times three rounds times one entry gives nine records and nine writing minutes. Briefing, passing and screening take additional time. Extensions need not be independent ideas.
Result: A–D are candidate groups, not approved solutions or customer findings. The three contributors do not represent all readers; no benefit has been measured.
Three people and three rounds yield nine records and nine writing minutes; provenance remains, with no claim of independent ideas, demand or approval.
Next step: Clarify the candidates and gather cost, capacity and permission evidence before screening or testing.
Three reasons to change the session before collecting ideas
First, pause if participants are answering different questions. “Improve the library” and “make after-hours returns clear and traceable” create different idea pools. State one shared question and retain known conditions. A contribution outside the scope can be preserved separately without pretending it answers the agreed task.
Second, stop if silent writing is being presented as anonymous or free from hierarchy without supporting arrangements. Participants may still recognize handwriting or fear later judgment. Explain who will see the sheets and how contributions will be discussed. Do not promise safety or participation merely because nobody speaks during a round.
Third, keep generation separate from selection. Critique, voting, and budget decisions during passing can narrow what people are willing to add. Known restrictions should remain visible, but a facilitator should not silently discard ideas while collecting them. Later screening needs explicit criteria and authority; the number of entries cannot establish quality or demand.
Choose between generation, recombination, explanation, and screening
This guide uses a written brainstorming variant with its own small-session settings. It does not prescribe the six-person, three-idea, five-minute protocol often called 6-3-5. Select related methods according to the next missing output rather than claiming every workshop needs the same sequence.
| Input and question | Method | Output and limit |
|---|---|---|
| One question and contributors who can write and pass | Brainwriting | Traceable contributions and extensions; not validated demand. |
| A problem can be split into distinct choices | Morphological analysis | Configurations and compatibility checks; not automatic selection. |
| A rough concept needs a coherent explanation | Idea sheet | A discussable concept and assumptions; not permission. |
| Concrete concepts have value and implementation evidence | Value-feasibility matrix | A bounded comparison; no automatic test approval. |
Trace all nine Starbridge records through the sheets
The complete fictional C4 packet concerns Starbridge Library Service after-hours returns. Twenty synthetic visits on October 12–14, 2026 include twelve during staffed hours and eight outside. These invented visits motivate a teaching question; they are not demand research. Three contributors P1–P3 write on sheets S1–S3 for three rounds of three minutes, one entry per cell. Briefing, passing, clarification, and screening require additional time.
S1 passes P1 to P2 to P3; S2 passes P2 to P3 to P1; S3 passes P3 to P1 to P2. The table supplies every contribution, author, and parent. An extension preserves its earlier record instead of replacing it. A new direction may retain the idea that prompted it without claiming it is the same concept.
Three people multiplied by three rounds and one entry produces nine records. The writing phase lasts nine elapsed minutes; with all three writing throughout, it represents twenty-seven participant-minutes. Neither count measures independent ideas or excludes the extra session work. Group A information, B box, C locker, and D partner counter retain their originating IDs. No screening or pilot has happened.
| Round and sheet | Author, record and parent | Contribution |
|---|---|---|
| 1 / S1 | P1 / I11 / root | Hours notice. |
| 2 / S1 | P2 / I12 / I11 | QR information. |
| 3 / S1 | P3 / I13 / I12 | Paper equivalent. |
| 1 / S2 | P2 / I21 / root | Locked box. |
| 2 / S2 | P3 / I22 / I21 | Numbered two-part slip. |
| 3 / S2 | P1 / I23 / I22 | Reconciliation at next opening. |
| 1 / S3 | P3 / I31 / root | Smart locker. |
| 2 / S3 | P1 / I32 / I31 | Partner counter instead. |
| 3 / S3 | P2 / I33 / I32 | Paper handover pouch. |
Preserve the contribution before evaluating the concept
At clarification, ask what an entry means and retain any correction beside its original wording. “Paper equivalent” needs to identify the information it would provide; a numbered slip needs a clear relationship to the returned item. Clarifying these points does not establish that readers understand the arrangement or that the venue permits it.
When grouping the records, keep one-to-many relationships where appropriate. The counter direction arose from the locker chain, but its delivery conditions differ. Do not erase that history to make four neat groups. Repeated ideas, blank cells in a future session, and disagreement can remain in the record without becoming individual performance scores.
For subsequent screening, gather actual estimates, capacity, safety conditions, and permissions. Define who decides and what can be tested. A slip reference, deposit in a box, and confirmed inventory receipt are different events; the proposed reader communication should make that distinction clear. No paid service or demonstrated benefit is supplied.
Review whether the process retained varied contributions and whether the resulting concepts are clear enough to investigate. This is separate from whether a service solves the customer problem. Keep any proposed test, its authorization, execution, and findings as later records rather than writing a successful outcome into the workshop summary.
Check your work
- Question, timing, rounds and passing rules are explicit.
- Original records and parent links survive synthesis; screening is separate.
- Counts are not demand evidence, independent-idea totals or performance scores.
Framework references
References explain the framework and its scope; they do not validate the fictional example's results.
Put the framework to work
Copy the outline and add your question, evidence, and open questions.