Rakazo builds an agentic, vertically-integrated ecosystem for the world's independent optical retailers: a global directory with first-party data authority, a website subscription that pays for itself, and factory-direct own-brand supply — distributed through owned inventory that no multinational can ever buy.
Independent optical retailers — roughly 460,000 storefronts worldwide — are squeezed between two giants: EssilorLuxottica owns the brands and increasingly the retail; Google owns their visibility. Nobody owns their digital ground truth or their supply relationship. Rakazo takes both, and wires them together.
The system runs in four connected layers. A global directory (glassesnearme.org) maps and verifies every optical business on earth at near-zero data cost. A website product (Zumo, getzumo.app) converts retailers into paying subscribers with AI-built sites that outperform what an agency would sell them. A marketplace (Glasses Gateway) turns the verified retailer network into a trade surface. And a sourcing engine — 50 manufacturers, 17,000 SKUs, our own warehouse — lets us launch our own frame brands offered to independents at margins the multinationals can't match.
Every layer feeds the next: the directory finds them, the website product converts them, the marketplace retains them, the supply relationship monetizes them at the highest margin. Distribution for our own brands rides on advertising inventory we own outright — permanently house-only, never sold to external advertisers. That inventory is the structural advantage over Essilor: they rent attention; we own the shelf.
An independent's entire web presence is often a Google Business Profile they don't control — no ownership of their own reviews, no analytics, and directory listings they can't edit. When the profile decays, the business disappears from the map.
EssilorLuxottica owns Ray-Ban, Oakley and LensCrafters — and the labs. Independent retailers buy from the same conglomerate that competes against them at retail, on terms set for them, not by them.
Agency quotes run $1,500–5,000 plus maintenance. Most small opticians go without — in our harvested Malaysian corpus, the majority of strong, well-reviewed shops have no real website at all.
We harvested 1,079 Malaysian optical records and enriched the top 100 with live Google Places data: 100% phone-reachable, 581 real photos, 500 reviews — and found chains whose websites are literally dead while their reviews stay strong (the Hollywood Eye-Trend case: 4 branches, dead site, domain paid through 2028). The pitch writes itself: "Your site is down right now. This one is live — and it's yours."
Anyone can sell websites to opticians. Only we can own the directory they're listed in, the data spine their truth lives on, the ad inventory their customers see, and the factory that makes their frames. Each asset makes the next one cheaper — and the whole system harder to replicate.
Every optical business on earth, mapped at $0 — 459,817 pages of owned shelf space.
The retailer claims their record: RM9.95/mo, owner-authoritative data, "edit once, updates everywhere."
House promos, featured tiers, pay-per-lead — served from inventory we own, to a warm verified pool.
Own-brand frames at factory cost, opt-in stocking for independents — the multiplier stage.
| Property | Role | Status | Live proof |
|---|---|---|---|
| Glasses Near Me glassesnearme.org | The global optical directory — discovery, verification, and claim funnel for every optical business on earth. Malaysia first (1,079 shops, 42 towns live), then global replication. | LIVE | 148 URLs live · JSON-LD · 100 enriched records · claim flow routing to Zumo |
| Get Zumo getzumo.app | The website product — AI-built sites for local business. Flagship demos convert opticians at RM9.95/mo; 340+ prospects in the engine. | LIVE | 11 demos live · 5.0★ widget (126 reviews) · claim flow awaiting Stripe links |
| Glasses Gateway BU-E | The marketplace layer — where the verified retailer network trades, and where GG-customer status derives from real transactions. | PARKED | Charter & CEO in place; reactivates on network traction |
| Sourcing & Own Brands channel #9 | Factory-direct frames from 50 manufacturers / 17k SKUs — offered opt-in to claimed independents as a stocking line, never as a competing consumer brand. | LATE-STAGE | 50-mfr supply chain + Danyang warehouse already held |
The production template system is built and running in a live sandbox: three dormant ad slots per page (top leaderboard, right rail, bottom banner) that sleep invisibly until a JSON data slice activates a house campaign — turning a promotion on across any region, town, or page-type is one data-file edit. Claimed-owner pages carry zero ads unless the owner opts in. Featured placement is capped at 3 per page, ordered by verification seniority, never sold to outsiders. The full demo is browsable at glassesnearme.org/template-demo/.
| # | Channel | Terms | Stage |
|---|---|---|---|
| 1 | Claims core engine | RM9.95/mo subscription — the owner claims and controls their record everywhere. | NOW |
| 2 | Featured listings | RM49.95/mo or RM399/yr · active claims only · max 3 per page · seniority ordering · never pay-for-rank. | NOW |
| 3 | House banners funnel asset, not revenue | 3 dormant slots/page, house promos only, permanent no-external-advertiser lock, Sponsored-labeled, zero ads on claimed pages unless owner opts in. | NOW |
| 4 | Pay-per-lead | Claimed shops only, opt-in, post-revenue. | LATER |
| 5 | B2B supply | Dashboard-only offers to claimed owners. | LATER |
| 6 | Data products | Aggregated clusters only — never raw owner data, never below the aggregation floor. | LAST |
| 9 | Own brands / private label | Our frames, factory cost, offered opt-in to claimed independents as a stocking line that beats multinational margin. Four neutrality guardrails: never as listing influence, opt-in only, owner toggle, firewall from ranking. | MULTIPLIER |
Parked permanently: sponsored pages and external ad-share — the Founder's house-only commitment admits zero exceptions. Rejected: consumer affiliate.
Pricing authority is delegated to the business's CEO layer, governed by the Founder's charter: "We genuinely want to add value to independent optical retailers — find the win/win pricing strategy." Every pricing change must carry a win/win rationale; a machine tripwire halts any change that raises effective price on a cohort whose retention is declining. The company prices to keep independents profitable, not to extract from them.
3 slots × 460k pages, dormant by default, activatable in seconds. Essilor buys attention through reps and co-op programs; we can never be outbid because we never sell. The refusal is the moat.
Claimed owners edit once — phone, hours, photos — and the update propagates across every property. Precedence is machine-enforced: owner > corroboration > free sources. Over years, the most accurate optical dataset on earth, assembled from consented owners.
459,817 stores discovered via free width-scan pipelines (IDs-only discovery, chain locators, open maps) with paid-API spend reserved for corroboration only. Competitors would face a data bill we simply don't have.
50 manufacturers, 17,000 SKUs, our own Danyang warehouse. This is the asset no directory competitor and no ad-network competitor possesses — and it's what turns a media business into a commerce business.
The global optical retail vertical maps to roughly 459,817 individual stores across 197 countries — a fragmented, almost entirely independent-facing market with no dominant owned directory. Our entry market is Malaysia (1,079 shops, fully harvested, directory live), chosen for language coverage, supply-chain adjacency, and PPP-appropriate pricing.
Discipline: width pages stay text-first (no scraped photos, no hotlinking — ToS-clean by design); photos appear only post-claim via owner uploads or sanctioned showcases. Quality decays on clocks; stale records suppress themselves.
The Founder's target is $20,000/week take-home (~$87k/mo). The path is a compounding ladder, not a bet: each stage funds the next, and the final multiplier — own brands — rides entirely on assets the earlier stages built.
| Stage | Mechanism | Indicative math |
|---|---|---|
| 1 · Claims | 460k pages; even 1% claiming at RM9.95/mo | ≈ RM45k/mo (~$10k/mo) recurring |
| 2 · Featured | 3-slot scarcity at RM49.95/mo on the engaged cohort | Meaningful add-on as claiming compounds |
| 3 · PPL + B2B | Leads and supply offers to the claimed pool | Post-revenue, high margin, opt-in |
| 4 · Own brands | Factory-cost frames (landed ~$30 vs $200–800 retail RRP) stocked opt-in by verified independents, promoted through owned inventory | The multiplier — the lane Essilor cannot touch |
Sole approval gate on sends, schema, external money — and the values charter. Sets direction; sees everything; approves by exception.
Scopes every mandate, verifies every deliverable empirically before it reaches the Founder, owns the decision register.
Specialist build agents with scoped deploy tokens. Any permission excess stops at 403 — by design.
Objectives-only owners per business unit: goals, metrics, roadmaps, monetization framing. No code, no outreach — proposals route through the Chief.
An admin-dashboard architecture (9 modules, spec complete) makes the rules executable: 17 tripwires monitor the live system — ads on a claimed page, own-brand logic touching ranking, an unlabeled promo, an external advertiser appearing in any slice — and any violation triggers alert, auto-halt, and an item in the Founder's queue. Every write is versioned in git and audited. Agents hold scoped tokens; a promo agent cannot touch pricing; a pricing owner cannot touch ranking.
| Phase | Scope | Status |
|---|---|---|
| Phase 0 now | Free global pipeline: chain locators + open-map independents + worldwide width discovery. $0 cash, pure agent time. MY directory live; template sandbox live; admin-dashboard spec complete. | AWAITING GREEN-LIGHT |
| Phase 1 | Claims engine live: Stripe programmatic checkout, owner edit panel, first-party data spine, house-promo activation, featured tier. Human web UI arrives only when claimed owners need it. | GATED ON PHASE 0 |
| Global | 197-country shard rollout under {cc} subdirectories; ~460k pages; PPP pricing bands delegated to the CEO layer with A/B testing. | SEQUENCED |
| Multiplier | Own-brand launch to the verified, opted-in retailer network; marketplace (GG) reactivation; sourcing-as-a-service at 5–10% per PO. | SEQUENCED |
This proposal is not just an investor document — it is the company's source of truth. Once signed off, it converts downward:
Strategy, values, guardrails, sequencing. Founder sign-off makes it v1.0 — the reference every decision cites.
Each objective-owner agent derives quarterly OKRs from the charter's stages and success metrics.
Product sequencing (Phase 0 → 1 → Global → Multiplier) broken into verifiable build mandates.
The admin dashboard's approval queue + agent work queues execute the roadmap, with tripwires keeping execution inside the charter.
On Founder approval this document becomes the governing reference: amended only by recorded Founder decisions, inherited automatically by all tripwires and team charters, and reviewed against reality at the weekly scorecard. Until signed, it is a draft — and the teams continue building exactly what it describes, because every element of it was decided with them.
Version: 1.0-draft · Prepared by: Chief of Staff (agentic) · Date: September 2026 · Distribution: Founder, CEO layer, investor conversations