Product.ai / Join / Projects / MLP Conversion Lift A/B Framework — PYO vs Legacy Head-to-Head
Project

MLP Conversion Lift A/B Framework — PYO vs Legacy Head-to-Head

The MLP-CONVERT outcome (600 impact points, revenue-fortress pillar) requires "PYO MLPs show statistically significant conversion lift ≥15% vs. legacy pages in A/B testing across 3+ merchant categories." Current state per the outcome file: PTR Host V1 shipped 2026-03-17; PYO Phase 4 in progress; "no A/B conversion data yet." The simplycodes-kernel makes PYO load-bearing — A-5 says discount behavior is governed by merchant economics not product category, and PYO is the migration to physics-based polymorphic rendering by archetype. The kernel's "What Winning Looks Like" includes "400K+ MLPs with polymorphic rendering matched to merchant discount physics" and the entire Project Unity migration (currently at ~65% of 2025 baseline per SC-TRAFFIC) was justified by the assumption that PYO converts better than legacy. If that assumption is wrong, every revenue forecast collapses and the Project Unity ROI becomes negative. Today nobody knows whether it's true. The team has been shipping PYO components without the conversion-measurement layer that tells us if PYO actually converts.
Project Overview
Discipline
growth-architect
Duration
2 weeks
Compensation
Your stated freelance rate
Surface
SimplyCodes · Revenue
Kernels
simplycodes · revenue
Outcomes
mlp-convert · sc-traffic
Tier
Consequential
Tooling
Claude Code or Co-work

Why we want this done

This project derives from a direct collision: simplycodes-kernel A-1 (Verification beats aggregation through asymmetric trust — Success Rate is the operational north star, retention is a function of accuracy not coverage) plus simplycodes-kernel A-5 (discount behavior is governed by merchant economics not category — PYO key off archetype). The downstream causal chain: PYO conversion lift validates the Project Unity investment, which validates the SC-TRAFFIC recovery curve, which validates the revenue floor as the old domain structure phases out. Every link in the chain is currently uncalibrated. Revenue-kernel A-1 (aligned monetization creates structural alignment — we make money only when the merchant makes a sale) means conversion-lift data is signal at the closest possible point to revenue. The brand-and-growth State of Practice §4 warns that classical multi-touch attribution has collapsed and that A/B testing under low-volume conditions requires statistical-power discipline; the operator must navigate that constraint with rigor.

Scope

Surfaces: simplycodes-web (primary — the MLPs themselves, the assignment layer, the conversion-event capture). Repos: simplycodes-web. Coordinates with: Kelli Schall (MLP-CONVERT measurer, PYO architecture), Dylan Ler (PYO Phase 4 + V3 ML injection), Isaac Sornborger (PTR Host V1 — the templating substrate the framework builds on), Cansu Kaya (analytics architecture, Project Unity baseline). The operator owns: the assignment mechanism, the conversion event taxonomy, the statistical-power methodology, the dashboard, the archetype-segmentation surface. Does NOT own: PYO component design (Kelli/Dylan), merchant ontology evolution (Kelli), the underlying templating engine (Isaac).

What success looks like

A working solution looks like this: every shopper landing on a SimplyCodes MLP is deterministically assigned to PYO or legacy by user-hash. Conversion events (code copy, outbound merchant click, eventual Nexus S2S postback per Revenue kernel A-2) join to the assignment server-side. The dashboard the operator builds shows daily conversion rate by variant, by category, by merchant archetype, with confidence intervals that respect the actual sample size. After 30 days the framework emits a defensible point estimate: "PYO converts at +X.X% vs legacy across these 3 categories, with the largest lift in Exclusion Fortress archetype (+Y.Y%, 95% CI [a, b])." The operator also ships the contract: future MLP variants get tested through this framework with no per-test re-engineering. Kelli Schall and Dylan Ler can ship PYO V3 (ML-Driven Dynamic Injection) on top of the framework. The Test: Kelli or any team engineer can pick a future PYO variant, deploy it through the framework, and read a statistically-powered conversion delta within 14-30 days without consulting the operator.

References

references.md
SimplyCodes Strategy Kernel — A-1 Verification, A-5 Merchant Economics archetype physics, MLP Architecture schema, Surface Strategy schema
Revenue Engine Kernel — A-2 Sovereign Transaction (Nexus S2S substrate)
MLP-CONVERT outcome — evidence criterion + initiatives
SC-TRAFFIC outcome — Project Unity context creating sample-size constraint
Brand and Growth State of Practice 2026 §4 — measurement physics, A/B-under-AI-mediated-discovery
Anti-reference: any A/B testing tool that buries methodology (treatment of p-values, power, multiple-testing) — VWO/Optimizely default reports famously do this and produce false-positive lifts at low power.
Anti-reference: PYO conversion data inferred from non-randomized rollout. The framework must be deterministic-assignment, not "we shipped PYO in March, and conversions went up since" — that is correlation not causation.

Constraints

  • 2 weeks default.
  • Claude Code or Co-work primary substrate.
  • No aios-methods access (trial contractor scope).
  • Assignment must be deterministic-by-user-hash (not random per session).
  • Framework cannot violate signal separation (simplycodes-kernel A-3, revenue-kernel A-3): conversion data flows upstream to calibrate the framework, never downstream to influence MLP ranking.
  • Privacy boundary: no per-user data leaves the closed loop.

```

---

```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.