THE TOOLKIT
Appendix B — The Question Set
How to build the locked question set your measurement depends on. Target 120 to 200 questions sourced from harvested customer language rather than a brainstorm, shaped across the buying journey, with a defined record kept for each question so re-runs stay comparable month over month.
From Becoming the Answer by Jeremy Osborn · 244 words
Shape
Target 120 to 200 questions.
| Segment | Share | Count at 150 |
|---|---|---|
| Constrained recommendation | 40% | 60 |
| Comparison | 20% | 30 |
| Brand-specific | 15% | 23 |
| Implementation and troubleshooting | 15% | 22 |
| Category education | 10% | 15 |
What to record for each
| Field | Note |
|---|---|
| Question text | Verbatim, in customer language |
| Segment | From the table above |
| Source | Sales call, support ticket, site search, forum, competitor position |
| Commercial value | High / medium / low |
| Current rank for the nearest search equivalent | From Search Console |
| Baseline cited domains | Recorded across at least seven runs |
| Baseline mention, yes or no | Averaged across runs |
| Information we hold that others don’t | Honest answer, or “none” |
| Winnable | Yes / no / not yet |
Rules
Source from harvested customer language, not from a brainstorm.
Include questions you currently lose. A set built from your strengths teaches you nothing.
Include your top three competitors’ strongest positions.
Cover the attribute matrix, not just head terms.
Freeze for a quarter. Version any change and treat the series as broken there.
Run each question at least seven times a day. Report rolling averages only.
Attribute matrix worksheet
For each core buying question, fill in your values and generate the intersections.
| Dimension | Your values |
|---|---|
| Company size or household type | |
| Industry or use context | |
| Integration or compatibility requirements | |
| Budget band | |
| Deployment or delivery model | |
| Compliance or regulatory regime | |
| Buyer sophistication | |
| Specific job to be done |
Three to five dimensions with three to five values each generates more intersections than you can serve. Rank by commercial value and honest winnability, then build the top twenty.