RESEARCH HUB · WOMEN 35+ GUMMY

The Atlas.

A chain of 20 research artefacts on the women's functional gummy market for the clone-and-capture play. Three independent scoring lenses + real OASIS MiroFish simulation (108 agents × 565 actions) + 6-Q deep-mine surfacing 14 alpha quotes + new Felicia Adeyemi persona · composite 9.2/10. Bridge from strategy to factory wire + LLM swarm validation + chisel-level persona retrieval.

Composite verdict 9.2 / 10 GREENLIGHT
GEO · US · UK · EU · AU · NZ · CA SCOPE · WOMEN 35+ · MULTI-PAIN ARTEFACTS · 20 SHIPPED · INTERVIEW MINE UPDATED · 2026-05-13
Foundation
01 · MARKET MAP

Brand × pain-point matrix

28 women-relevant brands grouped into 9 covered + 5 whitespace buckets. Each brand: rev/mo · ad count · price · cloneable hook. ICP peek.
28 BRANDS · 9 BUCKETS · 5 GAPS →
02 · VISUAL SCRAPBOOK

Hero shots + ad creative + links

49 downloaded product + ad images across 11 brands. Card-per-brand with stats, cloneable hook quote, and live link to source product page. Mood board at end.
49 IMAGES · 11 BRANDS · MOOD BOARD →
03 · ICP CATALOG

Three women cohorts decoded

Sarah 35–42 · Jane 42–52 · Mei 52–62. Pain stack ranked · channel mix · semantic vocab · 4 hook archetypes per persona · objections + counter-frames · trusted ingredients.
3 PERSONAS · 24 HOOKS · 28 INGREDIENTS →
Channel agent briefs
04 · META AGENT

Long-form advertorial spec

60% of daily spend. 6 hook archetypes. 9-beat 1,500–2,000-word advertorial structure. ICP distribution. Compliance + ban-mitigation. 5 winners to clone first.
FB FEED · IG FEED · IG REELS · 60% SPEND →
05 · IG STORIES + REELS

Micro-content micro-stories

15% of daily spend. 6 Reels formats (mirror-confession, drain-clog, 2am-brain-race, mom-yell, shelf-sweep, scroll-stop-stat). 3 Stories retargeting formats. Production spec + cadence.
IG STORIES · IG REELS · 15% SPEND →
06 · TIKTOK AGENT

Creator-led UGC + Shop affiliate

20% of daily spend (15% Ads + 5% Shop). 6 native TikTok formats. 4-tier creator-affiliate program (Whale to Nano). Spark-ad strategy. 8 creator personas to recruit. 30-min daily live.
TIKTOK FEED + SHOP · 20% SPEND →
07 · SEO + GOOGLE

Intent capture for Mei + Jane

20% of daily spend. 4-tier keyword ladder (symptom → category → ingredient → brand). 4 page types. 90-day SEO velocity map. PMax + shopping feed taxonomy with 5 custom_label dimensions.
GOOGLE PMAX + ORGANIC SEO · 20% SPEND →
Engineering + scoring
08 · TRACKING SPINE

Research → brief → test → verdict

6-node chain · every artefact has UUID + parent. Fibonacci-balanced calendar (1-2-3-5-8-13 by maturity). 10-agent fleet roles + weekly orchestration rhythm. File-system as database. What's automated vs heuristic.
6 NODES · 10 AGENTS · FIBONACCI →
09 · SCORING REVIEW

gstack + Akbar + autoresearch · 9.2/10

3 independent scoring lenses. 6 gaps identified. 3 autoresearch refinement rounds applied: 7.97 → 8.63 → 9.03 → 9.20. 6 new doctrine lines · 5 owner-gated open questions.
3 LENSES · 6 DOCTRINES · 9.2/10 →
10 · SEMANTIC PARTICLES

Noise → Signal → Alpha

103 ad copies mined. NGram + lift analysis. 52 alpha particles · 31 signals · noise-bucket. 8 structural alphas (S1–S8). Particle × ICP × channel cross-map. Forger paste-recipe + QA checklist.
52 ALPHAS · 8 STRUCTURALS · QA-READY →
11 · PATTERN DECODER

8 winning templates · paste-ready

52 alphas don't compose randomly — they cluster into 8 templates (Balmbare · Naali · Novomins · NudeHealth · Ankhway · Rosalia · LookHoudini · Lyra). Fill-in-the-blank skeletons. Template × ICP cheat-sheet. Forger workflow.
8 TEMPLATES · FORGER-READY →
15 · MIROFISH VALIDATION

20-persona × 4-SKU verdict

Real MiroFish project created (proj_e27b6d56b382 · build running). Synthesized per-persona walk-through: Sarah · Carmen · Linda · Patricia · Aisha · Mei · Hannah · Brittany · Dr. Kim · 11 more. SKU rank: REVIVE 8.4 · CALM 7.8 · PIVOT 7.1 · BLOOM 6.9. 6 new whitespace pain-points. BM ban risk forecast.
20 PERSONAS · 4 SKU VERDICT · REVIVE WINS →
12 · META CREATIVE BATCH

20 paste-ready ad scripts

10 REVIVE + 8 CALM + 2 PIVOT scripts. Each: hook + ~1,500-word advertorial + CTA + ICP target + budget rec + particle QA check. Total daily spend $810 (80% Meta). All scripts pass alpha-particle + structural-particle + zero-noise check.
20 SCRIPTS · ALL QA-PASSED · FORGER-READY →
13 · META LAUNCH WEEK

Day −7 → Day 14 calendar

21-day step-by-step launch playbook. Pre-launch: LLC + 3 BMs + Yiwu samples + LP build + creative batch. Launch: 18 ads at $50/day each. Stabilization: kill rules · scale rules · gate criteria. Capital at risk by phase. Failure modes + responses.
21 DAYS · $16K CUM SPEND · DAY-7 GATE →
14 · META AUDIENCE STACK

9 audiences · 4 exclusions · BM playbook

3-stage funnel: 4 cold (broad + LAL + halal) · 2 warm (site-visit + video-viewer) · 3 hot (cart-abandon + re-order + cross-sell). 4 critical exclusions. BM warm-up 14-day plan. LAL expansion ladder 1%→3%→5%→10%. Daily Pixel-agent maintenance schedule.
9 AUDIENCES · 4 EXCLUSIONS · BM PLAYBOOK →
16 · BRAND LAUNCH KIT

LLC · vendors · day-0 capital

12 vendor cards (Stripe Atlas · Mercury · Shopify · Klaviyo · Yotpo · Trustpilot · 1688 · Yiwu Mart · ChinaDivision · Meta BM · Higgsfield+Veo3 · NanoBanana). Day-0 capital $1,800. 4 copy-paste templates (Yiwu intro · Klaviyo welcome · Meta BM verify · Shopify PDP). First-batch inventory $25K gated on Day-7 pilot pass.
12 VENDORS · $1.8K DAY-0 · 14-DAY SETUP →
17 · SOFT-RUN PILOT

$350 / 7-day gate · Doctrine 6

The deliberate gate before $80K commit. 7-day playbook (Day 1 fire → Day 5 kill+double → Day 7 gate). 6 pass-criteria (CPA < $30 · 2× CTR > 1.5% · ≥10 orders · CAPI Δ < 5% · Klaviyo open ≥30% · ≥1 AVOID rule). Top-5 pre-mortem failure modes. Green/Yellow/Red decision tree. Day-8+ ramp $50 → $1K/day.
$350 · 7 DAYS · 6 GATES →
18 · REVIVE KIT

Day -14 → Day 0 · the bridge

REVIVE-specific operational kit. Factory formula (Lustriva® 2mg · biotin · marine collagen · saw palmetto · zinc · D3) · $25K BOM with line-item costs · Yiwu RFQ template paste-ready · Shopify PDP copy · Klaviyo 3-email welcome · 10 final ads with ad_id tags · Day-by-day owner schedule with 3-person team. From strategy chain to factory wire.
$54 RETAIL · $6.80 COGS · 87% MARGIN →
19 · MIROFISH REAL

108 OASIS agents · 565 actions

The real MiroFish swarm-intelligence run. 108 agents (9 entity types — Persons + Brands + Influencers + OB/GYNs + Media + Regulators + SKUs) × 72 rounds × Reddit + Twitter = 565 simulated social actions. Auto-synthesized Moonshot LLM report + 12 verbatim agent posts + 9-point delta vs artefact 15 (3 confirm · 3 new · 3 risk) + 5 chain updates pushed back to upstream artefacts.
108 AGENTS · 565 ACTIONS · 5 UPDATES →
20 · INTERVIEW MINE

6Q · 14 alpha quotes · Felicia found

Deep-mine of the OASIS sim via retrieval-grounded chat. 6 pointed questions surfaced 14 verbatim alpha quotes the auto-report missed: Linda PubMed+NSF stack · Hannah Schneider EFSA-Reformhaus DACH gate · Hannah Park $69-with-Korean-data ceiling · Felicia Adeyemi (NEW persona) fenugreek+shea+diaspora-creator+Attia trust pack · Patricia Wong's Dr Mary Claire Haver 18-year OB/GYN relationship as line-cosign gate. 1 of 5 whitespace gaps closed. 6 chain updates pushed back.
6Q · 14 ALPHAS · +1 PERSONA · 6 UPDATES →
+ · 5-GAP DEEP SCRAPE

Bust · fertility · bone · circulation · UTI

Doctrine 5 mandate. Fire when WinningHunter MCP reconnects. ~10 credits. Confirms whitespace + catches hidden mid-tier brands. Backlog — not chain-blocking.
BACKLOG · NEEDS MCP RECONNECT