The Latent Job of the Skincare Verifier — Pre-Beat-1 JTBD Excavation

Why we want this done

This project derives from consumer-experience-kernel A-1 (the conversation is the product, must serve the whole journey not a lookup moment), A-2 (honest depth over false breadth — depth must be in the right places), and A-8 (Trust Architects validate what LLMs cannot). It directly serves the full-journey outcome (currently on-track but launching in 30 days with no behavioral-verification fallback) and chat-expert. The skincare beachhead is the institutional-credibility-establishing launch per truth-graph-kernel A-5 (launch fewer categories with overwhelming depth). If the Beat 1 axiom set is grounded in the wrong job-being-hired-for, the launch fails not because the axioms are wrong but because they are answering the wrong question. The latent-demand layer is the pre-condition on every subsequent forge decision.

Scope

• Beat 1 cohort: the candidate identifies and recruits 15-25 skincare buyers in the target audience (split: ChatGPT shoppers, ingredient-conscious skincare buyers, Reddit-native verifiers — the kernel's three implicit JTBD shapes). Recruitment is part of the work, not a contracted-out step.
• Trust Architect recruitment: coordinate with whoever owns Trust Architect strategy (the Phase 1 program is specified but no TAs are recruited; the candidate may be the first to recruit). Run the kernel A-8 selection gate (3 axioms, 1 wrong, 1 missing context — can they catch it?).
• aios/kernels/consumer-experience-kernel.md: propose Signal updates to S-1 (Phase 1 Trust Architect Validation Pending) based on findings. Route via /strategy-sync if forged into axiom.
• productai-web (chat surface): the latent-job axioms become grounding material. Coordinate with the chat lead on the May launch.
• aios/outcomes/full-journey.md: the project's findings may revise the evidence criterion's category breakdown. Route via /outcomes if six-test gate holds.

What success looks like

A stranger reads the shipped artifact, opens the canonical JTBD list, picks the top latent job, follows the evidence trace to the source interviews / behavioral data / Trust Architect calls, and can verify the claim independently. The artifact is integrated into the active axiom forge: the next 20 skincare axioms forged after the project ends visibly ground in one of the surfaced latent jobs. The chat-expert team can read the document and know what to test for. Elena/Tess can read it and know whether the May launch surfaces should change. The Trust Architects who participated treat the artifact as their own diagnostic substrate.

The robust solution may surface a better cut: maybe the right unit is not "the latent job" but "the latent disqualifier" (what makes them reject a verdict regardless of accuracy). The candidate is invited to propose. What is non-negotiable: production-tie-in evidence (axiom candidates ready for forge, not standalone insights), Trust Architect concordance (external truth anchor — the candidate's own conclusions cannot be the verifier), and a refusal to confuse this with PM scope (no PRD, no roadmap, no feature spec — research-led artifact only).

References

Constraints

• Claude Code or Claude Co-work as primary AI substrate.
• Three-week duration. Day 1 onsite. Optional second onsite mid-trial for Trust Architect alignment.
• Trust Architect recruitment is the project's external truth anchor — the candidate cannot be the verifier of their own JTBD claims (oracle separation per recruiting-driver §5.2 rule 9). Skip this and the work fails the gate.
• Final deliverable: canonical JTBD list + axiom candidates + Trust Architect concordance evidence + integration brief for Elena/Tess/chat lead. NOT a slide deck. NOT a PRD. NOT a feature spec. Research-led artifact ready for forge consumption.
• The work is in the May launch surface; misalignment with the actual launch ships is rejected.

---

```

---

```yaml