| name | michael-margolis-user-research |
| description | Michael Margolis's (UX Research Partner, GV) bullseye customer research system — find the comically-narrow subset of your market most likely to adopt, then validate with 5 interviews, 3 comparison prototypes, in 1 day |
| version | 2.0 |
| format | completion-engine |
| workflows | 3 |
| source | claude.ai export 2026-07-01 |
Michael Margolis — Bullseye Customer Research
Michael Margolis was the first UX Research Partner in venture capital (GV, since 2010) and has run 300+ hands-on research sprints across biotech, healthcare, security, fintech, and consumer — after compressing 30 years of ethnographic research technique through Walmart.com, Gmail, and the Design Sprint. His system replaces months of guessing with the "five and three in one" formula: five bullseye customers, three simple comparison prototypes, one day of team-watched interviews. The output is not a report — it's a team that has seen the truth together and knows exactly what to do next.
Available Workflows
| # | Workflow | Produces |
|---|
| 01 | 01-define-bullseye-customer.md | Bullseye customer definition (~7 concrete attributes: inclusion, exclusion, triggers) + screener questionnaire |
| 02 | 02-build-comparison-prototypes.md | Three distinct flat prototypes, each a different value-prop "recipe" spread across the variables you need to test |
| 03 | 03-run-research-sprint-day.md | Interview guide + watch-party operating plan + debrief/takeaways system for the one-day sprint |
Quick Reference
- Bullseye customer ≠ ICP. The bullseye is the research-grade subset — comically narrow — that everyone agrees would most want this. The ICP you sell to later is broader.
- Formula: 5 + 3 in 1. Five one-hour interviews (data saturation), three distinct prototypes (comparison shopping), one day (patterns become obvious).
- Criteria come in three types: inclusion (easy), exclusion (dwell here — experts, competitor-locked, disqualifying history), triggers (events that make someone ripe: new CISO, new baby, something broke).
- Learning mode, not selling mode. Humble inquiry — the gentle art of asking instead of telling. Founders default to pitching; the interview requires the opposite.
- Weight past behavior over stated intent. "I would totally use this" that contradicts their history is noise.
- Predict before, capture after. Written predictions beat hindsight bias; a big-takeaways form beats a report nobody reads.
- Can't recruit them? That's a finding. If the bullseye can't be found, ask whether they exist at all.
Execution Prompts (structure-pure v2)
4 deterministic practitioner prompts — each carries an Output Contract, Output Skeleton, and Quality Gate. When a deliverable matches one, Read it and honor its contract instead of improvising the output shape.
- Michael Margolis — Bullseye Customer Definition + Screener —
skills/michael-margolis-user-research/references/prompts-v2/bullseye-customer-definition.md
- Michael Margolis — Comparison Prototype Set —
skills/michael-margolis-user-research/references/prompts-v2/comparison-prototype-set.md
- Michael Margolis — Sprint Interview Guide —
skills/michael-margolis-user-research/references/prompts-v2/sprint-interview-guide.md
- Michael Margolis — Watch-Party & Takeaways System —
skills/michael-margolis-user-research/references/prompts-v2/watch-party-takeaways-system.md
