Master Reference

Playbook 0. The doctrine and the 10-slot template every vertical inherits.

Read this first. The five vertical playbooks hang off this doc. Ratified by the 7/12 Fable ruling.

The doctrine

Every vertical is speed-to-lead + follow-up + AEO on one compounding engine that learns from every client. We swap the skin, the anchor source, and the proof. The machine never changes.

Say it in every pitch, enforce it in every build. A new vertical is a configuration exercise, never a product decision.

Why the engine compounds (the actual moat)

The whole engine is one loop: data → opportunities → receipts → AI draft → human approve → send/publish → outcome → attribution. The public brand brain (DH.com) is the same loop with “publish” swapped for “send.”

The portfolio at a glance

Sequencing law: the Bex → FranDevCo snowball. Get paid to prove one flagship. Over-deliver. Let results spark the next partnership. Clone with scarcity pricing. Five playbooks exist so any door that opens finds us ready, only the snowball’s current step gets active push. B2B leads the company; B2C runs when money and relationship make sense (So Bexy = silent volume data, never the spotlight).

The 10-slot playbook template

Every vertical playbook fills exactly these slots. A playbook with an empty slot is unfinished, and an unfinished playbook never reaches a pitch.

The sales funnel play (Bridge Funnel spine)

One page, split left/right, one brand pack, two renderers. Left = StoryBrand promise with a real face. Right = Sales Concierge running 12 canonical scenes like a Typeform: land → 3 quick wins with a live score dial → pre-reveal tease → email gate at the aha → full score + reframe insight → prescribe one path → one testimonial → close (stack, then one routed CTA) → reassure → exit routes with no dead ends.

Routing at close = Trust (ledger) x Heat (session) x Weight (offer price). Buy-now lives in the $47-$1,500 zone. Book-the-call appears only when the qualifier passes vetting, and the calendar gets a full brief. Low heat → nurture. Fail → disqualify-as-value. Only qualified prospects ever see a booking link, and everyone books briefed.

The cold outreach play (constant back half, three flex points)

The engine’s back half (sending, tracking, consent, compliance) is ~80% constant. Only three things flex per vertical: the anchor source, the copy proof-object, and the channel.

• Business-domain inbox → cold email, full engine (Type A reference row: FMCSA/Durafleet).
• Individual professional on IG → IG-led hybrid, email as a light flank (Type C: stylists).
• Consumer Gmail → email is OUT (Google’s 0.3% complaint ceiling, consumers tap). Postal + paid social instead (Type B lane 2).

The 6-question intake routes any new business type to an existing lane in under 30 minutes: who receives the money, where does the inbox live, is the target licensed/registered/filed anywhere, what signal separates a buyer from the census, is the offer itself regulated, what does the copy prove.

Constant copy mechanics, every vertical: 4 touches / 14 days / under 80 words / plain text / no links or images in email one / no pixels / truthful subject and from / postal address + unsubscribe / human approval on copy and first 50 sends.

The partnership play

Work our way in. No big-structure ask upfront. Lead with proof, over-deliver, let their side ask for more (the FranDevCo closing-slide pattern). Fulfillment split: Andrew = engine, Stefi = humans/sales (B2B), Rene + Inah = delivery. Ed Lane = family-discount anchor client, no equity talk.

Paid monetization pattern: one brain, two doors. Build the gated expert brain once; chat door now, MCP door weeks later as a thin wrapper on the same tools.ts. The MCP door is the cheapest door: the member’s own Claude subscription pays for the reasoning, client fuel rounds to zero.

The referral/affiliate play

30% default (30% first payment one-time; 30% recurring for 12 months on subs). 60-day last-click, code beats click. Built on the owned Cloudflare/D1 stack, no new tables. ~80% of revenue comes from the top 5% of partners, so the program is a hunt for 3-5 super-partners, never a mass blast.

The program launches AFTER the bridge funnel is live (clean attribution + the Customer Journey Map gate). Warm 1 super-partner scouting starts now, no links, no mechanics.

The compliance rail (cross-vertical, locked)

• Consumer cold email: permanently OUT (postal lane instead). Cold SMS: legal NO.
• US-only geo-fence excluding CA/EU/UK. CA business-recipient carve requires an explicit ruling with client sign-off.
• IG automation: inbound-triggered only.
• Franchise offers: geo-fenced out of the 14 registration states (CA, HI, IL, IN, MD, MI, MN, NY, ND, RI, SD, VA, WA, WI) pending counsel per client. No earnings claims outside FDD Item 19, ever, disclaimers cure nothing.
• DH never qualifies, scores, steers, or introduces specific franchise prospects (NASAA broker trigger). The franchisor owns the scoring config on their own funnel, DH is the software vendor. Counsel must bless this framing.
• Naming frozen. Honest/defensible ROI only. No new D1 tables (namespaced events + rollup dimensions). Payload keys channel + source_kind ratified once under the Landon rule.

Standing blockers (shared)

1. Domain-attach (THE blocker). Can the engine bolt onto a client’s existing domain without a full rebuild? Subfolder compounds authority; subdomain does not. Until ruled, every vertical is capped by build capacity and no 10x-at-scale number gets pitched.
2. Counsel gate on the scoring framing + franchise law effective dates.
3. So Bexy margin floor at volume (Haiku + $5 cap, cost-curve model in flight).
4. Customer Journey Map (named touchpoints) ratified as the gate before any go-live.

Honesty ledger

All prices are stated targets except: Bex $4,500/mo (live) and the Durafleet ~$5k trial (agreed shape). The “$100M/brand” figures are aspirations, never pitch material. Every ROI claim ships with receipts or ships never.