Discovery & Research Plan: Curator
Executive Summary
The product thesis (operational pain is real and paid-for; reproducible methodology is the moat) is currently backed by one founder conversation and the ICF Global Coaching Study cited on the page. This plan defines the minimum research to move the core hypotheses from T5/T6 (assertion/inference) toward T1/T2 (behavioral evidence) before automation investment.
Research Questions (ranked)
- How much time do coaches actually spend on between-session ops, per session and per week? (decides if pain is real)
- What do coaches do today, tool-by-tool, and where does it break? (workaround map)
- Would they pay to remove it, and how much? (willingness-to-pay signal)
- Does a structured analysis they'd hand a client require reproducibility they don't get from raw ChatGPT? (moat test)
- Who decides/pays — coach, client, or sponsor? (business-model unknown, see plan page)
Method
- Primary: 8–12 fact-based coach interviews (not concept pitches): "walk me through your last session's aftermath," timed. Recruit via ICF chapters + coaching schools.
- Behavioral: ask 3–5 coaches to log ops minutes for 2 weeks (diary study) → baseline.
- Artifact test: run the same recorded (consented, depersonalised) session through the analysis 3× → show coaches; ask "trustworthy enough to send a client?"
- Anti-bias: recruit at least 3 "no pain" coaches to seek disconfirming evidence.
Evidence Tiers Targeted
| Question | Now | Target |
|---|---|---|
| Ops time magnitude | T5 | T1 (diary) |
| Workaround map | T6 | T2 (observed) |
| WTP | none | T3 (stated, triangulated) |
| Reproducibility need | T6 | T2 (artifact reaction) |
Synthesis Plan
Code interviews for: pain frequency, current tools, breakpoints, trust thresholds, payer signals. Feed into JTBD canvas (06), opportunity tree (07), prioritization (08), and the reproducibility metric (10).
Research Gaps / Risks
- Recruiting bias toward tool-enthusiast coaches (over-states pain). Mitigate with mixed recruiting.
- Stated WTP ≠ real WTP; treat as directional only.
- Small N at concept stage — findings are signal, not proof.