CurioQuest

wiki/open-questions.md

Open Questions

Source: Project-Objective.md §12 + Research_Synthesis.md §13/§16. Resolved items are tracked as ADRs.

Strategic / decision-level (still open)

  • Primary customer wedge — parent-direct vs. gifting vs. homeschool. Current lean: parent-direct / summer-led, but left validatable in the pilot rather than hard-locked. Research leans gift-givers and/or homeschoolers first (proven WTP, low CAC).
  • Teacher / classroom B2B timing — deferred, but when? Proven WTP ($95–325/yr [FACT]) and smooths seasonality; different/slower sales motion.
  • Exact pricing numbers — bundle-anchored, volume-rewarded is locked as a shape; the actual tiers are set in the unit-economics model (Lane 2). Research guidance: AOV $40–75 [ASSUMPTION]; single $25 is too thin for hardcover.
  • Shopify ↔ generator-app seam — checkout/fulfillment on Shopify; creation wizard + generation likely a custom app. The boundary is undefined → resolve in the PRD.

Pilot-only unknowns (desk research cannot validate these)

These are the hinge metrics. The concierge pilot exists to measure them; every one is [ASSUMPTION] until then.

Unknown Pass / kill threshold Why only the pilot answers it
Preview→paid conversion ≥6% pass; <3% kill freemium entirely No benchmark exists for "preview→personalized-print-gift"; SaaS 2–3% [FACT] is the wrong proxy for gift-givers.
CAC (low-cost channel exists?) ≥40% non-paid acquisition pass Whether a cheap organic/referral/homeschool channel actually exists is unproven.
Retention / repeat intent ≥40% intend repeat Whether the grade-progression + gift-occasion cadence drives real repeat is unmeasured.
Willingness-to-pay / realized AOV ≥$50 realized AOV pass Real margin at our exact specs (hardcover + activity insert) is unmodeled — may be negative at $25.
AI reject rate qualitative gate (human-reviewed sample to ~20 parents/teachers) Trust/quality acceptability and "AI slop" risk only show up with real buyers.
POD feel passes hand-feel + durability Which POD partner delivers premium kids'-book quality at viable margin needs physical samples.

Rule: the big unknowns (conversion, CAC, retention, WTP, AI reject rate, POD feel) are pilot-only. Do not treat any desk-research figure as bankable for these.

  • The economics these unknowns drive
  • Validation plan & kill-criteria (§13)
  • How we got here