# Business & Product Requirements (from Bruce's Knowledge Base)

> Source: the **Company Knowledge Base → Super System** docs that Berner & Bruce
> already discussed (`Claude CoWork/.../Company Knowledge Base/super-system/`).
> Our earlier research was infrastructure-led; this captures the *business* and
> *product* requirements it missed. Reconciliation notes:
> [reconciliation-with-bruce-kb.md](../03-architecture/reconciliation-with-bruce-kb.md).

---

## The core thesis (the "why")

The current stack is a **"Frankenstein": ~8 platforms held together by webhooks,
with NO single source of truth.** The same customer lives in ActiveCampaign,
WordPress, a **39,000-row Google class-list Sheet**, and ThriveCart — kept in
sync by fragile automations. That causes recurring bugs, hours of weekly manual
work, inaccurate tracking, and ~$7k/year in tool fees.

**The Super System collapses all of it into one platform where data lives once
and everything reads from it.** Every feature below serves that thesis.

---

## What it replaces (the kill list)

| Today | Becomes |
|---|---|
| **ActiveCampaign** (CRM + email + course timing) | Built-in CRM + email via **Amazon SES** + a real scheduler |
| **Pabbly / n8n** (automation glue) | Internal code — no inter-system webhooks |
| **Hookdeck** (webhook retry) | Obsolete — traffic is internal |
| **Google class-list Sheet** (39k rows, backup truth) | The system DB *is* the source of truth |
| **ThriveCart** (checkout) | Custom cart → Stripe/PayPal direct |
| **Thrive Apprentice / WP drips** | Custom course delivery + proper access control |
| Manual Missive lookups | A real CRM screen |

**Kept (connected via API):** Missive (support), WebinarJam/EverWebinar (initially).

---

## New product requirements our infra plan missed

### 1. One product, many pricing options 🔴 (reshapes the data model)
Today: ~**18 Conflict Cure products** in Thrive + ~**8 in ThriveCart** —
duplicates created *only* because the tools couldn't model one course at
different prices/drips. Super System = **ONE course** + a **pricing structure**:
- Full $697 · Masterclass $277 · Follow-up $397 · Valentine's $197 · annual/lifetime add-ons.
- Each pricing option → its own checkout, **one product underneath.**
> **Data-model impact:** `Product` (the course) ↔ `PricingOption` (many) ↔
> `Checkout`. Access is granted to the *product*, not per price variant. This is
> the single biggest schema change vs a naive port of `tva_orders`.

### 2. Time-zone-aware everything 🔴 (cross-cutting)
Detect zone (JS location → IP → billing-country fallback), user can override,
then **open courses and send emails in the customer's LOCAL time.** Retires the
"Phoenix-time" standard and Eastern-hemisphere email workarounds. Caveat: Apple
privacy can hide IP/email → hence auto-detect-with-override.
> Touches Courses (drip), Messaging (send time), and the scheduler.

### 3. Gamification (Duolingo-style) — a *revenue* feature
Progress bars/badges, smaller lesson chunks, **cohort visibility**
(first-names-only "others in your group are here too"), and **couple/partner
progress** ("how the two of you are doing"). Goal: higher completion → retention
→ upsells. New module territory; not in the original 8-module list.

### 4. CRM with per-user "3-dimensional" column views
Replaces ActiveCampaign-as-CRM **and** the class-list Sheet:
- Leads / Students (with progress) / Partner details.
- Per-customer: payments (incl. **double-payment detection**), hover-notes, tags,
  course progress, time-on-video, masterclass watch %.
- **Per-user show/hide columns** — Bruce hides emails; Linka shows status; a coach
  gets a filtered "coaching call" view with quick note-taking. Sensitive columns
  **permission-gated**.

### 5. Real scheduler (not "waiting" automations)
A cron runs **every minute**: "what needs to happen now?" → course unlocks,
emails, reminders fire (in local time). Replaces AC's "contact waiting in an
automation" model.

### 6. Editable marketing vs hard-coded logic 🔴 (architectural principle)
A firm split:
- **Hard-coded & locked:** access control, course timing, syncing, back-end logic
  — can't be changed by a click, so it can't break.
- **Editable by the team:** email sequences, split tests, page content,
  course/pricing setup — marketing (John) ships funnels without engineering.
> Preserves WordPress's one real advantage (non-coders can edit) without its
> fragility. Affects how we build the admin app + content models.

### 7. Funnel builder + custom checkout
GoHighLevel-style funnel building (what John wants) without the limitations.
Stripe/PayPal direct (Apple Pay, Google Pay, Klarna via Stripe). Refunds from
within the system. Time-zone-aware signup.

### 8. Revenue, tracking & attribution
A sales/revenue screen (replacing distrusted ThriveCart reports) + end-to-end
**attribution under one domain** — directly addresses the ad-cost/tracking crisis.
Could absorb **ClickMagick** (~$3,500/yr).

### 9. Accounting sync
Push orders straight into **QuickBooks/Wave** via API (Berner has done ~10k
orders/mo before, no manual entry).

### 10. Real-time (sockets) + AI-native
- **Sockets:** live notifications, instant CRM updates, live cohort progress.
- **AI-native APIs:** so **Claude** can query data, generate reports, assist ops
  (same pattern John uses with Meta). A deliberate goal, not an afterthought.

### 11. Per-user views & permissions
Every screen permission-gated; finances hidden from most; per-user column views.

---

## Security posture (from Bruce's KB)
- Secrets in env files (encrypted, 256-bit), never in code/public HTML
  (fixes the current exposed-Missive-token problem).
- API keys **rotated monthly**, scoped by sensitivity.
- Bearer-token auth on inbound calls.
- **Bugsnag / "Bear"** error monitoring → instant email/Slack alerts.
- **YubiKey** for staff logins under consideration; password manager TBD
  (RoboForm / 1Password / Bitwarden / Proton / custom vault).
- Backups: platform DB + Berner's local RAID/NVMe + **international** copy.

---

## Commercials & timing 🔴
- **~$7k/year tool spend** is the target to retire (exact figure: Berner to total).
- **Backstop deadline: ActiveCampaign renews 4 June 2027 (~$5,581/yr).** Natural
  "be live by" date so we don't pay another AC year. (AC partial-refunds unused
  time, so it's soft, not a hard wall.)
- **First win in motion:** the **dispute-evidence tool** Berner already built —
  standalone, high-value, folds into Commerce.

### Savings targets (annual, rough)
| Tool | ~$/yr | Fate |
|---|---|---|
| ActiveCampaign | ~$5,581 | Replaced (SES + built-in CRM) |
| ClickMagick | ~$3,500 | Largely absorbed (own tracking) |
| AWS (was video backups) | ~$2,800 → ~$0 | Already moved to **BunnyStream** |
| Hookdeck | ~$470 | Obsolete |
| Pabbly | ~$240–360 | Obsolete |
| Uncanny Automator / Thrive licences | licences | Obsolete |

> Bigger wins than tool cost: **accurate data**, **time saved**, and **new
> revenue** from gamification/retention.

---

## Phasing (Bruce/Linka's framing — complements our Phase 0–5)
- **Stabilise first, build in parallel.** Current stack is stable enough; Bruce
  maintains it (training with Berner) while Berner builds the replacement.
- **Eliminate one system at a time** — each phase retires ≥1 tool → early savings.
- **Don't rush the funnel/website** — highest value is back-end / course delivery
  + the single source of truth.
- **Open phasing question:** Linka's idea = pilot with the simpler **Trust Again**
  program first; Bruce's lean = build **course-delivery/back-end (CRM + delivery)**
  first, leave marketing site, tackle tracking later. *(Decide with Linka.)*