aibook-fun/AiBook_PRD.md

---
title: "AIBook.fun — Product Requirements Document"
type: prd
status: draft
owner: Patrick (Product)
launch-target: 2026-09-26
last-updated: 2026-08-26
---

# 📚 AIBook.fun — Product Requirements Document

> **Status: DRAFT.** Dollar figures, dates past today, and commercial terms in this
> document are planning assumptions, not commitments. Nothing here should be quoted
> to backers or the public until the specific line is signed off (see
> [§10 Open decisions](#10-open-decisions)). This revision rewrites the prior draft
> for plain language and folds in a product review — see
> [Appendix D](#appendix-d--what-changed-in-this-revision) for what moved and why.

---

## TL;DR

AIBook.fun is a **crowd-written, practical field guide to using AI agents responsibly at
work** — sold as an annual subscription, launched on Kickstarter on **September 26, 2026**.
Its unit of content is not a chapter but a **workflow you can actually run**: a documented,
tested recipe for getting a real job done with AI, complete with the guardrails.

The one thing that must work on day one: a reader opens a workflow, **runs it in ~15 minutes,
and shares the result** with other readers. Everything else — the contributor economy, the
cohort program, the agent-onboarding use case — is a layer we build *on top of* that loop
after we've proven the loop works.

- **Who it's for:** busy professionals (real-estate, financial advisors, fractional
  executives, higher-ed, community leaders, technologists) who want to apply an AI workflow
  *this week*, not read AI theory.
- **The ask:** fund the first edition on Kickstarter — **$5K minimum / $35K target**.
- **The bet:** if the "run a workflow and share it" loop is genuinely useful, the same
  people who read become the people who contribute — and the book stays alive.

---

## 1. The core loop and the one metric

Everything in this product either serves this loop or waits its turn:

```
   READ a workflow  →  RUN the Try-it (~15 min)  →  SHARE the result
        ▲                                                  │
        └──────────  see others' results, come back  ──────┘
```

**North-Star metric — Activation:** *a backer reads at least one workflow and completes at
least one Try-it within 14 days of getting access.*

If activation is healthy, the product is alive and the rest (renewal, contribution,
community) becomes possible. If activation is weak, no amount of funding, community
scaffolding, or governance saves it. We report activation **first**, above the fundraising
numbers, from the first day backers have access.

---

## 2. Background & problem

**The problem:** Professionals are drowning in AI hype and have nowhere practical, trustworthy,
and current to learn how to *actually use* AI agents in their real work — safely. Books go
stale in months. Courses are generic. Twitter threads aren't accountable. And nobody's guide
is written by, and kept current by, the practitioners doing the work.

**Why us, why now:** WeOwnNet needs a shared, practical AI-literacy base for the people (and,
later, the AI agents) it works with. A living book of *tested workflows* — contributed by
practitioners, curated for quality — is that base. The Kickstarter both funds the first
edition and assembles the founding community.

**What makes it different (the claim we have to earn):** every workflow is *grounded in
practice* — it shows the real steps and the real tool output, not a plausible-sounding
description. "It works" is not a claim we make; it's something the reader can verify by
running the Try-it themselves.

---

## 3. What the product is — and what it is *not yet*

The prior draft treated five different products as one launch. They are not one product. Here
is the honest layering:

| Layer | What it is | When |
|:--|:--|:--|
| **Core (v1 — launch)** | A subscription book of tested AI workflows + the read→run→share loop | **9/26 launch** |
| Layer 2 (post-launch) | Community contribution — readers submit their own workflows | After core loop proven |
| Layer 3 (post-launch) | Cohort learning — "Book Circles," Workflow-of-the-Week showcase | After first content proves demand |
| Layer 4 (later) | Agent onboarding — the same content, read by AI agents to onboard | Deliberate future use of the format |
| Layer 5 (later) | Funnel into the wider WeOwnNet ecosystem (AgencyPRO) | Ongoing, low-effort |

**This matters because** each layer has a different success metric and a different failure
mode. Trying to build all five for 9/26 is how a small team ships none of them well. v1 is the
**core loop, done excellently.** The other layers are designed-for but not built-for launch.

### Explicit non-goals
- Not a general-purpose AI chatbot or tool. It's a learning product.
- Not regulated professional advice on the book's own authority.
- No crypto / token / Web3 / NFT language on any public or Kickstarter-facing surface.
  *(This is a hard Kickstarter-policy constraint — see [Appendix A](#appendix-a--kickstarter-compliance-the-nft-scrub).)*
- Not building a separate "eBookNFT platform" — that's a future idea; this book is the first product.

---

## 4. Who it's for

Personas are written as **jobs-to-be-done**. Note the buyer's job is *utility* ("help me do a
thing this week"), which is how we should lead the pitch — ownership and community come later,
as reasons to *stay*, not reasons to *buy*.

| Persona | The job they're hiring us for | We've succeeded when… |
|:--|:--|:--|
| **Reader / subscriber** | "Apply one documented AI workflow in my real job this week." | They complete a Try-it within 14 days of getting access. |
| **Contributor** *(Layer 2)* | "Publish a workflow I've proven, earn recognition (and, if we build it, income)." | Their first workflow is accepted within 90 days. |
| **Curator / co-author** | "Keep the quality bar high; own a vertical." | Reviews turned around within ~10 working days. |
| **Affiliate** | "Refer backers, earn commission." | First payout lands within one cycle of a referred pledge. |
| **Ecosystem prospect** | "Learn responsible AI; maybe get more involved." | ≥40% of backers join the community *(draft target)*. |
| **AI agent** *(Layer 4)* | "Onboard by reading the corpus." | *Designed-for, not a launch requirement — see §6.4.* |

**Who it's *not* for:** people who want AI theory with no practice behind it; anyone seeking
regulated professional advice from the book itself; teams wanting private hosted
infrastructure (out of scope).

---

## 5. The content model — where the real risk is

### 5.1 The unit of content: a Workflow Recipe
Every entry follows one template. This template *is* the product's quality guarantee — it's
the structural answer to "how do I know this AI advice isn't hallucinated?"

| Field | What it captures |
|:--|:--|
| **Problem (the job)** | The real task or pain this solves |
| **Actors** | Who runs it — a person, a role, or an AI agent |
| **Method** | Ordered, reproducible steps — **showing the actual tool output**, not a description of it |
| **Safeguards** | The responsible-AI guardrails: data handling, human oversight, failure modes, escalation |
| **Evidence** | The documented result — artifacts or metrics |
| **Try-it** | A ~15-minute exercise so the reader learns by doing |
| **Costs** | Model/tool costs where known — honesty about the economics |
| **Author & sources** | Who wrote it, when, for which vertical, with citations |

**Chapters** are curated collections of these recipes, one set per vertical: Real Estate ·
Financial Advisors · Fractional C-Suite · Higher Ed · Communities · Technology. **Editions**
refresh annually — which is the honest justification for a *subscription* rather than a
one-time purchase: the content genuinely goes stale as AI changes.

### 5.2 The cold-start problem (this is the #1 product risk)
The value proposition is "a rich living library." At launch, the prior plan has us with **2
sample chapters and 3 runnable kits** — a pamphlet, not a library — and a hope that the crowd
fills the rest. **Crowd-sourced libraries reliably die at cold-start.** We must not bet the
launch on strangers showing up to write content for free.

**Requirement — "Minimum Lovable Library":** before we ask anyone to pay, the founding
co-authors and team produce a floor of **15–20 fully-tested workflows across at least 3
verticals** *(recommended number — team to ratify)*. This is a **launch gate**, not a
post-launch aspiration. Community contribution is **Layer 2** — an experiment we run *after*
we've seeded a library worth subscribing to.

### 5.3 Be honest: v1 is an *authored* book, not yet a *crowd* book
The prior draft's "contributor economy" pays co-authors **15% affiliate commission** — which
is a reward for *referring backers*, not for *writing workflows*. Writing a workflow currently
earns a byline and nothing else. The mechanic that would actually pay contributors — royalties
— is unresolved (see [D-2](#10-open-decisions)).

So for v1, we state plainly: **the first edition is written by a founding set of co-authors and
the team.** Open community contribution — with a real incentive attached — is the very next
thing we build, and its incentive design ([D-2](#10-open-decisions)) is a **launch-blocking
decision, not a nice-to-have**, because the entire "living book" thesis depends on it.

---

## 6. Requirements

Priorities: 🔴 must-have for 9/26 · 🟡 should-have (launch window) · 🟢 post-launch.
Requirements are grouped by the layer they serve; **most 🔴 items serve the core loop.**

### 6.1 Campaign (Kickstarter) — 🔴
| ID | Requirement | Pri |
|:--|:--|:--|
| C-1 | Kickstarter page: hero video (2–3 min), problem, the book, co-authors, 6 pledge tiers ($10→$4,888), stretch goals, FAQ (≥10), risks — **passes the NFT-language scrub** and Kickstarter category rules | 🔴 |
| C-2 | Pre-launch landing page on aibook.fun with email capture + a 3-email warm-up sequence. **Target: 500+ emails before Day 1.** | 🔴 |
| C-3 | Affiliate program live (Kickbooster): co-author 15% / community 10% / podcast 10% / press 5%; 30-day cookie, last-click | 🔴 |
| C-4 | Launch playbook: Day-1 hour-by-hour schedule, weekly cadence, mid-campaign tactics, final-48h push | 🔴 |
| C-5 | Campaign analytics dashboard: raised, backers, avg pledge, email list, cost-per-click, affiliate share — reviewed daily | 🔴 |
| C-6 | Post-campaign comms staged before launch: final update, backer sequence, fulfillment announcement | 🟡 |

### 6.2 Content (the core loop) — 🔴
| ID | Requirement | Pri |
|:--|:--|:--|
| K-1 | Workflow Recipe template (§5.1), published, with **3 filled, reviewed examples** | 🔴 |
| K-2 | **Minimum Lovable Library: 15–20 tested workflows, ≥3 verticals** (§5.2) — team/co-author authored | 🔴 |
| K-3 | Chapter outline per vertical, with a named owner for each | 🔴 |
| K-4 | 2 sample chapters delivered pre-launch as the backer preview | 🔴 |
| K-5 | 3 "clone-and-run" kits, each **verified by actually running it** | 🔴 |
| K-6 | Editorial pipeline: submit → co-author review → accept/revise → publish → track updates | 🔴 |
| K-7 | Content stored in a structured, versioned, machine-readable format (frontmatter + provenance + trust tiers) so editions are diffable and, later, agent-readable | 🟡 |

### 6.3 Access & subscription — 🔴 / 🟢
| ID | Requirement | Pri |
|:--|:--|:--|
| A-1 | Post-campaign access: a digital key unlocks PDF + web reader for 12 months, renews annually | 🟢 |
| A-2 | Anchor price **$47/yr** *(draft — needs commercial re-sign)* + hardcopy at-cost upcharge | 🔴 |
| A-3 | Lifetime tier — **decision required before copy freeze** (see [D-1](#10-open-decisions)) | 🟡 |
| A-4 | Fulfillment ledger: backer ↔ key/upcharge tracked (a spreadsheet is fine for v1) | 🟢 |

### 6.4 The other layers — mostly 🟡 / 🟢
| ID | Requirement | Pri |
|:--|:--|:--|
| L-1 | *(Layer 3)* Try-it feedback loop: readers post results in the community channel (pinned template) | 🔴 |
| L-2 | *(Layer 3)* "Book Circles" — 6-week cohort format + facilitator playbook; **1 pilot cohort** post-launch | 🟡 |
| L-3 | *(Layer 3)* Workflow-of-the-Week showcase (site + community + email) | 🟡 |
| L-4 | *(Layer 2)* Public contribution intake + curation checklist — opens **after** the Minimum Lovable Library exists | 🟡 |
| L-5 | *(Layer 4)* Agent-onboarding: the same content, structured for machine ingestion — **a property of the K-7 format, not a separate build for launch** (see decision in [D-5](#10-open-decisions)) | 🟢 |
| L-6 | *(Layer 5)* Ecosystem funnel: backers flow into the wider WeOwnNet programs | 🟢 |

### 6.5 Site, email, ops — 🔴
| ID | Requirement | Pri |
|:--|:--|:--|
| S-1 | aibook.fun production landing (hero, the book, a sample workflow, tiers, affiliate, CTA, FAQ) | 🔴 |
| S-2 | Working HTTPS (current cert fails — verified 08-21); keep existing analytics continuity | 🔴 |
| S-3 | Email/CRM tags for the funnel: `prelaunch`, `backed`, `subscriber`, `affiliate` | 🔴 |
| S-4 | 8-email campaign sequence (3 pre-launch, 1 launch, 2 mid, 2 final), dates aligned to 9/26, NFT-scrubbed | 🔴 |
| S-5 | Instrument **events, not content** — track that a workflow was downloaded/submitted, never the confidential content itself (see [Appendix B](#appendix-b--trust--privacy-posture)) | 🔴 |

---

## 7. Non-functional requirements

| ID | Requirement |
|:--|:--|
| NFR-1 | **Truth in status:** no draft/unconfirmed figure appears in any external asset. Hard gate before publish. |
| NFR-2 | **Kickstarter compliance:** zero crypto/Web3/NFT terms on any campaign surface ([Appendix A](#appendix-a--kickstarter-compliance-the-nft-scrub)). |
| NFR-3 | **Privacy:** no confidential client data in any workflow or campaign asset; track events, not content. |
| NFR-4 | **Availability:** landing page up through the campaign; emails on schedule; affiliate tracking live from Day 1 (the 30-day cookie means every hour of Day 1 counts). |
| NFR-5 | **Content quality:** every workflow passes the Recipe template + at least one co-author review. |
| NFR-6 | **Domain continuity:** aibook.fun must not lapse (it currently expires 09-05 — see [§9](#9-immediate-next-steps)). |
| NFR-7 | **Accessibility:** reader and landing page are mobile-readable, targeting WCAG AA. |
| NFR-8 | **Cost discipline:** pre-revenue burn stays within the ~$6K budget ([Appendix C](#appendix-c--budget)). |

---

## 8. Success metrics

**Product health first, fundraising second.** The prior draft led with fundraising vanity
metrics; those measure the *campaign*, not the *product.* A funded campaign can still hide a
dead product if nobody activates, renews, or contributes.

### Product health (report these first)
| Metric | Floor | Target | Why it matters |
|:--|:--|:--|:--|
| **Activation** — read 1 workflow + do 1 Try-it in 14 days | 20% | 35% | The North-Star. Is the core loop actually useful? |
| **Workflow submissions** during campaign window | 10 | 25 | Early signal that Layer 2 (contribution) is real |
| **Renewal rate** — year-1 cohort | — | 60% | Is a *subscription* the right model at all? |
| **Curation turnaround** ≤ 10 working days | 80% | 95% | Can we keep quality without becoming the bottleneck? |

### Campaign performance (the fundraise)
| Metric | Min | Target | Stretch |
|:--|:--|:--|:--|
| Total raised | $5,000 | $35,000 | $54,000+ |
| Total backers | 300 | 700 | 1,000+ |
| Pre-launch email list | 500 | 1,000 | 2,000+ |
| Days to fund | 7 | 3 | 1 |
| Ad return on spend | 3:1 | 5:1 | 10:1 |
| Affiliate share of total | — | ≥20% | 35% |
| Backers → ecosystem pipeline | 50 | 150 | 300+ |

> ⚠️ **Two bars, not one.** Hitting the **$5K funding floor** (~106 backers) proves people will
> *pre-order*. It does **not** prove they'll *renew* or *contribute* — the two things this
> business actually depends on. Treat the **product-health floors above as a separate,
> equally-weighted validation bar.** "Funded" is not "validated."

*All draft targets must be ratified before being quoted anywhere.*

---

## 9. Immediate next steps

**This week — decisions and de-risking (before any copy or build locks):**
1. **Renew aibook.fun today.** It expires 09-05; launch is 09-26. This is a chore, not a
   strategic decision — do it now and delete it from the risk list. *(Owner: whoever holds the
   registrar login.)*
2. **Resolve the pricing spine together — lifetime ([D-1](#10-open-decisions)) *and* royalties
   ([D-2](#10-open-decisions)).** They're linked, and both block copy freeze.
3. **Ratify the North-Star + the product-health floors** (§8) so we're measuring the right
   thing from Day 1.
4. **Ratify the Minimum Lovable Library number and assign every workflow an author** (§5.2).
5. **Confirm the two unconfirmed co-authors** (Higher Ed, Communities) or re-scope those
   verticals out of v1.

**Build phase (weeks to launch):**
6. Landing page ships as a **hook A/B test** ("run this AI workflow today" vs. "own a
   community-built AI guide") — we should *learn which message converts* before we spend ad
   budget, not just capture emails.
7. Author the Minimum Lovable Library to its floor number.
8. Standard campaign build — hero video, email sequence, Kickstarter page — **after** 1–5.
9. Keep two governance gates sharp and drop the ceremony from the rest ([Appendix B](#appendix-b--trust--privacy-posture)):
   the **NFT-language scrub** (existential — Kickstarter will reject us otherwise) and
   **truth-in-claims** (no draft figures go public).

**Launch → fulfillment:** run the campaign, then fulfill (manuscript → edit → layout → proof →
issue access keys → print/ship hardcopies), then turn on the renewal system and open Layer 2
contribution. Detailed timeline lives in the Implementation Plan.

---

## 10. Open decisions

These are the *genuinely open* strategic decisions. Each carries a **recommendation** so the
decision-maker has a starting point — but the recommendation is not yet a decision.

| # | Decision | Recommendation | Blocks |
|:--|:--|:--|:--|
| **D-1** | **Lifetime tier — offer it, and at what price?** A ~$97–147 lifetime breaks even against $47/yr in ~2–3 years, which *undercuts the recurring-revenue thesis.* | **Pick a spine.** Keep $47/yr as the anchor. If you want a lifetime option, frame it *only* as a bounded early-bird instrument to capitalize the community fund (capped quantity), not the default SKU. Don't ship recurring + lifetime as equals. | KS copy freeze |
| **D-2** | **Contributor payment — how do people earn for *writing workflows* (not just referring)?** Today the only paid mechanic is affiliate referral. | **Resolve before launch.** This is the engine of the "living book." Decide: royalty pool, per-accepted-workflow bounty, or an explicit "v1 is authored, paid contribution ships in Q1" — but decide, don't leave blank. | The core value prop |
| **D-3** | **Co-author confirmations** (Higher Ed, Communities verticals). | Confirm within the week or cut those verticals from v1 and add post-launch. | Chapter outline, scope |
| **D-4** | **Refresh the source-of-truth plan** to the 9/26 dates and re-signed terms (depends on D-1/D-2). | Do after D-1/D-2 land. | Copy freeze |
| **D-5** | **Human-first or agent-first for v1?** Writing for human motivation and for machine ingestion are different jobs. | **Human-first.** Build the content format so agents *can* read it later (K-7), but optimize v1 for human learners. Agent onboarding is Layer 4. | Content design |
| **D-6** | **Does "you own a stake" survive the NFT-scrub as a *true* claim?** If the ownership model was fundamentally token mechanics, stripping them for Kickstarter may leave "ownership" as marketing over an ordinary subscription. | **Audit the claim.** Either define what ownership concretely means without tokens (governance rights? revenue share? credited authorship?) or stop using the word publicly. Add it to the claims register. | Public messaging integrity |

> *Resolved and removed from this list:* the domain "decision" — it's a chore (renew it, §9),
> not a strategic gate.

---

## Appendix A — Kickstarter compliance (the NFT scrub)

Kickstarter bans NFT/crypto/Web3 language. Any internal terms rooted in that model must be
translated on every public surface **before** submission, e.g. an internal "subscription NFT"
becomes, publicly, an **"annual digital subscription."** A single scrub gate reviews *every*
asset (page, emails, landing, video captions) against a translation table before anything
ships. **This is the single most existential compliance requirement** — a rejection here kills
the launch date.

## Appendix B — Trust & privacy posture

Kept deliberately lightweight for this stage; two rules are load-bearing, the rest is
proportionate:
- **Confidential material never appears in public content.** Workflows are illustrative and
  use non-confidential examples.
- **Instrument events, not content.** We record *that* a workflow was downloaded or submitted —
  never the private content of what a user did.
- **Honest data-handling language.** Describe what's actually true ("your documents stay on
  your instance"); never claim a security property we haven't verified is switched on.
- Heavier governance ceremony (formal sign-off registers, isolation-floor audits, corpus
  conformance) scales up *with the product* — it doesn't precede a pre-revenue, unvalidated
  launch.

## Appendix C — Budget

Pre-revenue, ~$6K, matching the source plan:

| Item | Cost |
|:--|:--|
| LinkedIn ads (campaign wks 2–4) | $1,300 |
| Meta ads | $700 |
| Google ads | $280 |
| Cover / video edit | $0–500 |
| Kickstarter fees (~8% on $35K) | ~$2,800 |
| Kickbooster (10% on referred pledges) | ~$700 |
| Domain renewal | ~$15–20 |
| **Pre-revenue total** | **~$6,000** |
| Projected revenue | $35,000–$54,000 |

## Appendix D — What changed in this revision

This revision is a **plain-language rewrite with a product review folded in.** Substantive
changes from the prior draft, and the reasoning:

1. **Named one core loop + one North-Star** (read→run→share; activation). The prior draft
   marked ~half of 40 requirements "must-have" — no real prioritization.
2. **Layered the product** (§3) so v1 = the core loop; contribution, cohorts, agents, and the
   funnel are explicitly later layers.
3. **Elevated the content cold-start** to the #1 risk and added the *Minimum Lovable Library*
   launch gate (§5.2).
4. **Told the truth about incentives** (§5.3): v1 is an authored book; the "contributor
   economy" is currently a referral scheme; paid contribution (D-2) is now launch-blocking.
5. **Forced the subscription-vs-lifetime tension** into a real decision (D-1).
6. **Split the validation bar from the funding floor** (§8) so a funded-but-dead outcome can't
   hide.
7. **Reframed the pitch as utility-first** (§4, §9 hook test) — the buyer wants a workflow, not
   an ownership philosophy.
8. **Demoted governance to proportionate** (Appendix B), keeping only the NFT-scrub and
   truth-in-claims as active gates.
9. **Turned the domain "decision" into a same-day chore** (§9).
10. **Translated the jargon out.** Internal codes, handles, and merge-history are gone from the
    body; a who's-who / glossary can be maintained separately if the team wants traceability
    back to the source-of-truth IDs.

*(A field-by-field mapping from the old FR-x IDs to these can be produced on request.)*