16Fold

01Executive Summary

In one breath: 16Fold is a separate brand built on top of the agentic system Robert already runs at Crabtree Marketing — licensed to businesses as one of three products, supported by a standardized channel partner network, with a hard rule that every client gets their own instance, not a shared piece of Robert's.

02The Story Behind the Brand

16Fold did not come from a naming exercise. It came from a sentence Robert said on the Proforma call:

"I duplicated myself 16 times." — Robert Crabtree, on the Proforma Hess-Crisci call

The line landed because it was true. The agency had gone from sixteen people to six because eighteen agents were running 24/7 in Robert's stead. The product writes itself: the system that lets one operator run like sixteen, in their own voice, without the headcount.

Last night's research session — eight hours of brainstorm with Hermes over Telegram — pressure-tested every other naming direction. Tree and nature themes were attempted (Overstory, Heartwood, Taproot, Canopy, Watershed, Treeline). Every single one had a Class 35 or Class 42 conflict, an active competitor, or a venture-backed satellite. The naming exercise eliminated names; Robert's own voice produced one.

What 16Fold means

• 16 — the number of agents currently running on the system. The number isn't a marketing fiction; it is what Robert literally describes when he tells the story.
• Fold — multiplication, layering, depth. Not "automation," not "software." The connotation that matters: amplifying a person, not replacing one.
• You, multiplied. — the positioning line. Frames the agents as experts in your corner, not "automation tools."

Plan-of-record assumes a relaunch after the first cohort of clients is in market. The launch name is not the forever name; it is the name Robert can ship with this week.

03Brand & Positioning

Brand voice rule

Frame agents as experts in your corner, never "automation tools." A 16Fold customer hires Patel, Schwartz, Ogilvy and a dozen others — they don't "buy software." This phrasing is non-negotiable in every piece of customer-facing copy.

What 16Fold sounds like in a conversation

"It's like a small team of marketing experts who already know your voice, your offers, and your customer — running 24/7. You'd hire seven people for what one of them does. We give you sixteen."

04The Three Product Models

Never conflate these three. They have different price points, different margins, different customer expectations, and different operational risk. A clean line between them is the difference between a clean P&L and a chaos engine.

05Pricing Architecture

Finalized May 4, refined May 5–6. This is the pricing draft pending your final approval.

Model B — Managed Service Tiers

Setup Fee Contract Incentive

The setup fee is the lever for contract length. Industry research shows this approach grows average contract length 40–60%.

Add-On Menu

Available across all Model B tiers unless noted. Add-ons can be turned on / off per-client without re-papering the contract.

Quality Gap — The Real Margin Driver

This is the part that does not fit on a pricing page but defines the unit economics:

• Clients running the system themselves (Model A) get roughly 60% of the output Robert produces.
• Robert's 20+ years of creative and strategic judgment is the actual product on Models B and C — not the agents.
• Plan: encode Robert's review criteria into agent skill files → Matt runs QA against those criteria on Tier 2 → Robert spot-checks Tier 2+, owns Tier 3.
• Never let this gap be a reason not to sell. It's the reason clients pay the retainer.

06Channel Partner Program

Same agreement for every partner. Same tier structure. No bespoke deals, no handshake percentages, no exceptions. The biggest failure mode in many-partner programs is treating each partner like a one-off; that creates resentment, accounting chaos, and legal exposure.

Three Tiers — Codified

Mechanics

Payment cadence

15th of the following month. $100 minimum threshold. Activation key: first commission paid within 5 business days of client signing — not in monthly batch — to make the program feel real on Day 1.

Clawback

Client cancels within 90 days → commission reversed. Standard SaaS clause; partners expect it.

Tracking

Tapfiliate (10+ partners) → PartnerStack (30+). Spreadsheets are not an option past 5 partners.

Agreement

Single Partner Program Agreement. Every partner signs the same document. No exceptions.

CPA compliance note

AICPA ET §1.510 permits referral fees with client disclosure, provided the CPA does not perform attest services (audits/reviews) for the same client. Compliance obligation lives with the partner via PPA language — not 16Fold's burden to police.

Partners Currently Warm

07Technical Architecture

A written brief — not code. The right developer can build against this; the wrong one will tell us why it's "more complicated than that." Robert's correction on May 4 set the load-bearing principle: "Hermes on my machine is mine. They need to have their own Hermes."

The Hard Rule — Client Owns Their Instance

• Client's Hermes: runs on their machine, their Anthropic account, their tokens. Robert never shares his.
• Client's APIs and MCPs: their credentials, their connectors. Their costs, billed to them.
• Robert's cloud layer: n8n, Supabase, content viewer, R2. This is what gets licensed.
• On/off switch: Robert revokes cloud access at the Supabase row / config flag level. Client's local Hermes still runs but loses the pipeline magic.
• Team accounts: Ken, Matt, Madison each get their own Anthropic account. Robert sets per-account budget caps. Nobody ever uses Robert's primary key.

What Goes In the Architecture Brief (T3-A)

1. How client's local Hermes connects to Robert's cloud infrastructure.
2. How client isolation works in Supabase (per-client rows, Robert-controlled access flags).
3. How n8n workflows are organized per client (parameterized — not one workflow per client).
4. How the content viewer becomes multi-tenant (per-client auth, Robert controls accounts).
5. How token usage is logged per client per pipeline run.
6. How Robert's on/off switch works at each layer.
7. Seat model — what counts as a "seat," how multi-user access works per client.
8. What client installs locally vs. what lives entirely in cloud.

Token + API Cost Policy

• n8n logs token usage per client → Supabase per pipeline run → reportable monthly.
• Each team member has their own Anthropic account; Robert sets the budget cap.
• Pass-through vs. absorb: defined per-tier in the contract; the client's heavy usage is their bill, not ours.
• Escalation: if a client's pipeline burns 3× expected tokens, automatic alert + conversation before next billing cycle.

08Pilot Clients & Industry Targets

Pilot Cohort — Free 90 Days for Testimonial

Industry Targets — Where 16Fold Has Native Edge

Industries where Crabtree Marketing already has deep operating experience and pre-built brand voice tooling. Lead with these; expand later.

• Golf cart industry — Icon Golf Cars relationship, multi-location experience, calendar template tested.
• Automotive dealerships — Ford, Chevy, Chrysler, Jeep, etc. The ICP looks similar to golf cars: multi-location operators, weekly calendars, direct-response copy needs.
• Trucking and logistics — recurring content + driver recruiting both fit the engine.
• Marine — Terry's Marine site is a live proof point.
• Doctors / dentists / insurance / financial services — adjacent to GreenPath's referral base; CPA-introduced.

Where 16Fold Does Not Lead

• Enterprise (they "already know it all" — Proforma's read, not ours).
• Pure e-commerce DTC at scale (not the engine's strength).
• Industries where compliance approval slows content to a crawl (regulated finance, pharma) — defer until a specialized template exists.

09Legal Entity & IP

Recommendation: form a separate Ohio LLC for the product brand. The reasoning is liability isolation, channel-partner equity flexibility, and an independently saleable asset later.

Why Not DBA

• An Ohio Trade Name registration ($39, Form 534A) creates no liability isolation.
• A product-liability claim or contract dispute on the AI product reaches the agency.
• Cannot issue separate equity — any "partnership" lives inside Crabtree Marketing LLC.
• P&Ls commingle on the same tax return. Future sale or valuation gets harder, not easier.

Why Separate Ohio LLC

• Full liability isolation from CM agency.
• If GreenPath equity ever happens, it lives in the product LLC only — never in Crabtree Marketing.
• Independently saleable.
• Multi-member operating agreement supports the partner program properly.
• S-Corp election available when product LLC nets $40K+ profit.
• Holding-company structure: revisit at $1M product ARR.

IP Posture

• Client owns content; 16Fold owns the system. This is the IP clause that goes in every client agreement.
• USPTO trademark: Class 35 (advertising, business services) + Class 42 (software, SaaS). File before going public.
• NDAs first: team members AND channel partners sign before any briefing — including before the first technical architecture conversation.

10Six-Phase Launch Sequence

From the master plan written 2026-05-04, adjusted for what's already done. Phase 0 is complete; Phase 1 is in flight; Phases 2–5 sequence Robert's bandwidth, not the calendar.

11Open Decisions for Robert

All eight require research-first / RC-decides discipline. Nothing should be guessed.

12Risks & Mitigations

13Thought Process & Research

The reasoning behind each major decision, in case anything needs to be re-examined later.

Why a separate brand at all (vs. just selling under CM)

Three reasons. One: the buyer for a managed agentic service is not the same buyer as a full-service agency engagement; CM's positioning would muddy the pitch. Two: separate brand makes the product independently saleable down the road. Three: partner economics — paying GreenPath 10% of an agency engagement feels different than paying them 10% of a productized service. The product brand is what the partners refer; the agency stays the agency.

Why "powered by Crabtree Marketing"

Trust transfer. New brand, no track record. CM has a 20-year reputation locally. The "powered by" tag inherits that without forcing 16Fold to live inside CM's brand voice. Same pattern as Stripe Atlas, Square Capital, etc.

Why three product models, not one

Three buyer profiles. Some clients want to own the system on their machine and walk; that's Model A. Most want recurring delivery they don't have to think about; that's Model B. A small number need bespoke; that's Model C. Pricing them as one bundle would force every client into "managed forever or nothing," and that excludes a meaningful fraction of the addressable market.

Why standardized partner tiers (no custom deals)

Research showed every successful many-partner program — agency, SaaS, professional services — uses a published tier structure. Custom deals at Partner #2 become a fairness problem at Partner #20. The PPA + tier addendum + automated tracking are the architecture; the tiers are the only knob.

Why setup fees with contract incentives

Industry data: setup fees with length-of-commit reductions grow average contract length 40–60%. The full setup fee is high enough to feel like skin in the game. The 12-month waiver is high enough to convert hesitant prospects into anchor clients. The 6-month half-off is the middle path that matches the buyer who wants commitment without 12-month exposure.

Why client owns their Hermes

Robert's correction May 4: "Hermes on my machine is mine. They need to have their own Hermes." The architecture decision matters because it answers: who pays for tokens, who controls upgrades, who can be cut off, who carries liability for their own runtime. Bundling into Robert's account creates exactly the failure mode Anthropic credit auditing has already shown: one bad pipeline week destroys a budget that was supposed to last a month. Sandboxing per client is the cleaner model and the easier story.

Why pilot clients are free for 90 days

Testimonials and case studies are worth more than three months of $1,500. Hunter Castle and Dylan Mullins are both choices that compound — Hunter through BNI introductions, Dylan through social reach. The trade is the right trade for a brand pre-launch. Stop offering the pilot tier the moment we have three live testimonials.

Why GoDaddy lied about the domains

GoDaddy showed both 16fold.com and 16fold.ai as taken; dig +short showed neither had DNS records. Reason: GoDaddy inflates "taken" status on keyword domains to drive premium-domain sales. Cloudflare Registrar showed honest at-cost availability. Lesson encoded into the trademark-domain-availability-research skill: always verify with DNS first, always buy through Cloudflare, never trust GoDaddy availability reports.

Why "16Fold" beat the tree-and-nature direction

Robert wanted names that evoked roots, growth, durability — connected to "Crabtree" as tree signature. The direction generated good concept territory (Ironwood, Groundswell, Deepgrain, Arborist) but every brand-tier candidate had hard conflicts. The breakthrough was looking for a name in Robert's own words on the Proforma call. The pattern: when mining for a brand name, search the session transcript for phrases the founder said in their own voice first. That is now a documented research lesson.

14File & Skill Index

Every artifact created during the May 4–6 work, where it lives, and what it covers.

Vault Files (Google Drive · CLAUDEVAULT)

• business/ventures/machine-product/MACHINE-LAUNCH-PLAN.html + .pdf — the original launch deck. This is the file Robert flagged as having broken layout. This document supersedes it.
• business/ventures/machine-product/MASTER-PLAN.md — full plan in markdown form (agent-readable).
• business/ventures/machine-product/PRICING-RESEARCH.md — competitive pricing landscape that informed the Tier 1–3 numbers.
• business/ventures/machine-product/PARTNER-PROGRAM-RESEARCH.md — the channel partner architecture research.
• business/ventures/machine-product/LEGAL-ENTITY-RESEARCH.md — Ohio LLC vs. DBA vs. holding-company analysis.
• business/ventures/machine-product/BRAND-NAME-RESEARCH.md — full naming research log including the eliminated list.
• business/ventures/MACHINE-INDUSTRY-LICENSING.md — industry-by-industry targeting cut.
• _system/COWORK-HANDOFF-2026-05-06.md — the 596-line master handoff doc Hermes wrote 2026-05-06 14:47.
• business/ai-saas/CONTENT-ENGINE-LICENSING-PLAN.md — earlier licensing-plan precursor.

Hermes Skills (Mac Studio · ~/.hermes/skills/crabtree/)

• machine-product-planning — canonical product plan, brand decision, pricing, partner tiers. Loaded automatically on any "The Machine" / "GreenPath" / "channel partner" trigger.
• trademark-domain-availability-research — GoDaddy false-positive lesson + Cloudflare Registrar pattern + DNS verification flow.
• idea-dump-intake — Notion pagination pitfall, file-based parsing pattern.
• rc-ai-install-product — Model A pricing and scope.
• plan — the planning-first discipline that gates all execution.

Hermes Plan File

• ~/.hermes/plans/2026-05-04_120000-machine-product-launch.md — the full task graph with phase ordering.

Live Surfaces

• 16Fold coming-soon page — built, deployed at sixteenfold.pages.dev. Custom domain attachment for 16fold.ai + sixteenfold.com is configured in Cloudflare Pages and pending DNS resolution. (Last action item below.)
• This document — the unified review surface; lives at the URL Robert is reading from.