Product.ai / Join / Projects / Skincare First-Verdict Conversion Funnel Instrumentation
Project

Skincare First-Verdict Conversion Funnel Instrumentation

Product.ai's first GTM beat ("Your Health, Verified" — skincare, sunscreen, supplements) targets a May 31 launch, and the FULL-JOURNEY outcome (1,100 impact points, the highest in the consumer-experience pillar) requires ≥25 tracked purchase completions per month sustained for 30 days. Today: zero. The Truth Card foundation is locked, Card 1 is in drafting, and the team is shipping content. What is not shipping is the instrumentation that tells us where in the funnel real shoppers drop. The Trust Paradox (74% rate verified-truth assistants 4-5/5 on trust; 93% verify before transacting) means stated intent will not predict purchase — only behavioral telemetry will. If we ship the launch without funnel instrumentation, we will get a vanity-traffic number, no diagnostic on why conversions did or did not happen, and no path to a calibrated second beat. The window is six weeks.
Project Overview
Discipline
growth-architect
Duration
2 weeks
Compensation
Your stated freelance rate
Surface
Product.ai · Revenue
Kernels
productai · revenue
Outcomes
full-journey · personalize
Tier
Consequential
Tooling
Claude Code or Co-work

Why we want this done

This project derives from a direct collision of three load-bearing pieces: the Product.ai kernel A-4 ("Adjudication Ends Searches") which states the verdict is the product, not the search; the Revenue kernel A-2 (Sovereign Transaction / Nexus S2S) which makes server-side attribution the substrate for every Surface Fleet endpoint; and the FULL-JOURNEY outcome which says the first end-to-end research-to-purchase journey is the market-validation milestone. Brand and content can ship the verdict; instrumentation determines whether we know what happens after a shopper reads it. Without this layer, the GTM Beat 1 launch is a content release with no telemetry — the same posture that produced 16 years of SimplyCodes data that lived in the team's heads instead of in the model. Shipping this also unlocks the personalization flywheel (P-Axiom Conversion Lift, Revenue kernel S-2) because the funnel instrumentation IS the substrate against which P-Axiom variants get measured.

Scope

Surfaces: productai-web (the Truth Card landing pages and verdict surfaces — the project ships the instrumentation layer there), external (Nexus S2S postback wiring for skincare merchants in the first beat). Repos: productai-web. Coordinates with: Elena Madrigal (FULL-JOURNEY measurer, GTM Beat 1 owner), Tess (GTM execution), Bri Stanback (multi-surface technical infra, AIOS-pulse for measurement reliability), Cansu Kaya (analytics architecture for the Project Unity baseline comparison). The operator owns the funnel instrumentation contract end-to-end: event taxonomy, Transaction ID lifecycle, S2S postback handlers, dashboard consumption, the diagnostic protocol. Does NOT own: Truth Card content authoring (Tess), brand voice (out of scope per recruiting-driver §5.1 distinction). The project lives at the seam between marketing and engineering and the operator must coordinate with both without owning either.

What success looks like

A robust solution looks like this: a stranger lands on a Product.ai skincare verdict from any source (organic, social, agent referral). Their session generates a Transaction ID at recommendation time per Revenue kernel A-2. Every behavioral event in their session — scroll-depth on the verdict, dwell time on evidence, outbound click, eventual conversion via Nexus S2S postback from the merchant — joins to that Transaction ID server-side. The dashboard the operator builds shows daily funnel drop-off by stage, by Truth Card format, by traffic source, by user segment. When the May 31 launch happens, the team can answer "did the launch work" with stage-resolved telemetry, not vibes. When Card 2 ships, the operator can A/B test format variants with statistical power because the instrumentation supports it. The work ships into production and the next operator who joins inherits a calibrated stack instead of having to reinvent measurement. The Test: a stranger to the team can pick one weekday in the first month after launch, query the dashboard, and produce a one-page funnel diagnostic that names the bottleneck with confidence intervals.

References

references.md
Product.ai Enterprise Kernel — A-2 Multiplicative Intelligence, A-4 Adjudication, A-6 Radical Explainability
Revenue Engine Kernel — A-2 Sovereign Transaction (Nexus S2S architecture)
FULL-JOURNEY outcome — evidence criterion, current state
PERSONALIZE outcome — downstream P-Axiom flywheel this work enables
Anti-reference: Honey/Capital One Shopping conversion-tracking pattern (Brand kernel S-1, SimplyCodes S-1) — surveillance attribution under ad-driven incentives. We are doing the structural opposite (server-side, single-customer, transparent).
Anti-reference: classical multi-touch attribution (brand-and-growth State of Practice §4 — "multi-touch attribution has collapsed"). The operator should not rebuild MTA.

Constraints

  • Claude Code or Co-work as primary AI substrate (recruiting-driver §8). Hybrid Cursor/V0 allowed for visual work.
  • IP boundary: trial contractor scope. No aios-methods access. No production keys to merchant-side postback systems until verified by Bri.
  • Time: 2 weeks default. Skincare Beat 1 launches May 31 — instrumentation must ship before launch, not after.
  • Privacy boundary: Revenue kernel A-7 ("We will never be an ad company"). The instrumentation cannot leak user signal to merchant-side beyond the conversion postback. Closed-loop attribution only.
  • Signal Separation (Revenue kernel A-3, Product.ai kernel A-5): instrumentation cannot feed ranking. Funnel data flows upstream to calibrate; never downstream to influence which products surface.

```

---

```yaml

Apply
01

Read the Codex (10 min)

The operating principles we work by. If they resonate, the rest of this will land. Open the Codex →

02

12-minute video screen

Hireflix, async. Questions are calibrated to this project specifically.

03

Chemistry call (30-60 min)

Direct call with the CEO. Strategic alignment and mutual fit. No problem-solving exercise.

04

Project begins within 2-3 weeks

1099 contractor agreement, NDA, paid at your stated rate. Day 1 in Santa Monica.