strategy

Design Machines Strategy Skill

Business strategy and product intelligence for Design Machines OÜ. Company details (registration, location, tax structure) are stored in ai-memory — search for "Design Machines OÜ".

Mission

Design Machines exists to democratize the workplace. We use design, systems thinking, and operations to make it easy to start and run democratic organizations like worker-owned cooperatives.

Elevator Pitch

Governance is the part of running a co-op that nobody wants to do. It's boring. It's confusing. And if you mess it up, you're not really a co-op anymore. I build systems that make governance simple. Even enjoyable. So people can focus on the work they actually started the co-op to do.

Positioning Statement

For small worker cooperatives struggling with governance, Assembly is a bespoke operations system that makes running a co-op simple and even enjoyable. Unlike spreadsheets, generic business tools, or expensive enterprise solutions, I build purpose-built systems informed by years of actually running a worker co-op.

The survival reframe

Worker co-ops outlast conventional businesses.

UK: 76% five-year survival vs ~42% for all new companies (Co-operatives UK). France: 80–90% three-year worker co-op survival vs ~66% conventional (CG Scop / Review of Industrial Organisation 2015). Italy: 87% three-year survival for worker-buyout co-ops vs 48% all Italian businesses (CECOP/CICOPA).

The standard pitch to co-ops assumes fragility. The data inverts the frame. Assembly is infrastructure that compounds an existing competitive advantage, not a life-support system.

Full citations live at references/survival-reframe.md. Verified URLs for external use are at plugins/design-machines/skills/audience/references/survival-reframe-citations.md.

The two moats

Design Machines' moat is two things working together.

First, integration across the six governance fixtures (decisions, meetings, members, equity, compliance, documentation) in one system of record. No competitor does all six.

Second, bylaws stop being paperwork. Assembly enforces statutory requirements (the AGM 15-month rule, the 14-day director-change filing) and bylaw-or-process requirements (the 2/3-of-votes-cast special-resolution threshold encoded in your rules; quorum thresholds set per the Rules of Association; blocks in consent decision-making require a written reason; new-member approval per each co-op's bylaws). Every other tool treats bylaws as reference docs.

Landing lines: "Your bylaws stop being paperwork." · "The rules run the system." · "Governance you can't accidentally break." Full positioning at references/two-moats.md. Audience-specific pitch versions at plugins/design-machines/skills/audience/references/developer-federation-pitch.md and coop-pitch.md.

---

Product Family

Code	Name	Type	Status	What It Does
DM-005	Assembly	Works	Prototype	Bespoke governance OS for worker co-ops
DM-003	Live Wires	Works	Beta	Typography-first CSS framework for editorial web
—	Nimber	Works	Legacy/Future	Rate setting & estimation for agencies

Assembly is the flagship. Live Wires is secondary but strategically important as the top of the conversion funnel. Nimber is future consideration.

Assembly Architecture

Assembly (product line)
├── Baseplate — core application starter (every co-op starts here)
├── Fixtures — modules that bolt on
│   ├── Governance (meetings, resolutions, proposals, decisions)
│   ├── Members (directory, roles, onboarding, lifecycle)
│   ├── Equity (shares, ICAs, patronage, redemption)
│   ├── Compensation (factor-based salary framework)
│   └── Documentation (SOPs, manuals, knowledge base)
└── [Client Name] — whatever the co-op calls theirs
    └── Each co-op names their own install

Naming logic:

• Assembly = factory assembly (making things) + democratic assembly (deciding things)
• Baseplate = machinery mounting plate. Everything bolts to it.
• Fixtures = manufacturing jigs that configure machines for specific work
• Client naming: Each co-op names their own install. It's theirs, not ours.

Tech stack: Go backend + Datastar (real-time SSE) + Live Wires CSS + SQLite (modular) + Templ templates

Live Wires Positioning

"The CSS framework built for content-heavy websites, where typography comes first, class names make sense to editors, and prototypes evolve into production code."

Key differentiators vs Tailwind: semantic classes, typography-first, no utility soup, CMS-friendly naming.

The Conversion Funnel

Designer uses Live Wires → encounters labor/AI content → learns about co-ops → becomes Assembly client. Products catch at craft level, co-op work catches at values level.

---

DM Catalog System

Inspired by Factory Records' FAC numbering. Everything is a release. A CSS framework sits next to a conference talk sits next to a governance system.

Current Catalog

Code	Name	Category	Status
DM-000	"Design Machines" essay	/PRESS	Origin · 2015
DM-001	Design Machines OÜ	Company	Active · 2024
DM-002	"AI or Bust(ed)"	/FLOOR	Dot All 2024
DM-003	Live Wires	/WORKS	Beta
DM-004	The Northern Star	/PLATE	Demo site
DM-005	Assembly	/WORKS	Prototype

Category Markers

Marker	Meaning	Covers	Source
/WORKS	Products & tools	Frameworks, applications, calculators	Factory works (ironworks, steelworks)
/PLATE	Templates & starters	Starter kits, themes, components	Printing plates (reusable forms)
/PRESS	Publications	Essays, guides, books, courses	The printing press; "the press"
/FLOOR	Talks & events	Conference talks, workshops, podcasts	The shop floor; "you have the floor"

Markers are optional. DM-003 works alone. DM-003/WORKS adds clarity. Number is canonical.

Rules

• Sequential numbering, chronological by catalog entry
• Codes always public — on product pages, slides, invoices
• Printed small, monospaced, consistent position
• Reserved numbers for milestones (DM-051, DM-100)

---

Target Market

Primary: Small Worker Cooperatives (5-50 members)

People who:

• Are maturing past startup phase, hitting governance growing pains
• Started without deep cooperative experience
• Care about doing it right — values-driven
• Have access to grants or can match funding

Market Segments

Segment	Priority	Entry Point
Existing worker co-ops	Primary	Co-op developers, federations
New co-ops	High	Co-op incubation pipelines
Platform cooperatives	Underserved	Tech-literate, better funded
Conversion market	Long-term	Live Wires → labor content → co-ops

Not the Customer (for now)

Large established co-ops with own systems. Co-ops with no budget and no path to funding. Traditional businesses wanting a website.

Pain point: admin debt

Admin debt functions like technical debt for governance. Bylaws drift from how the co-op actually works. Equity spreadsheets only one person understands. Decisions get relitigated because nobody can find the last one. The Tuesday-morning member can't act without "asking Sarah." Admin debt sits squarely inside the DM factory/labor/publishing triangle because debt is materialist. Assembly's diagnosis: admin debt is a measurable, fixable thing — the same way you wouldn't run a production system on a spreadsheet, you shouldn't run a co-op on one.

---

Revenue Model

Project-Based Engagements

Assembly projects cover Discovery + Design, Implementation, and Training phases with optional maintenance retainers. Pricing details are stored in ai-memory — search for "Assembly" or check ${CLAUDE_SKILL_DIR}/references/revenue.md.

Funding Sources

Co-op direct payment, federation grants, credit union community development funds, co-op matching funds. Partner network knows the funding landscape.

Financial Targets

Current financial targets, runway, and burn rate are stored in ai-memory — search for "Design Machines OÜ".

---

Partnership Ecosystem

The Emerging Stack

Layer	What
Financial Foundation	Bookkeeping, cash flow, pricing tools for co-ops
Governance Interface	Decisions, members, equity, democracy UX (Assembly)
Real-time Collaboration	Live state sync, SSE for voting and dashboards
Design System	Typography-first CSS, editorial design (Live Wires)

Key Partners

Current partner details and pipeline status are stored in ai-memory. Search for partner names or "Design Machines" relationships. See ${CLAUDE_SKILL_DIR}/references/partnerships.md for partner archetypes and ecosystem structure.

Current Pipeline

Active pipeline details (specific clients, meeting dates, status) are stored in ai-memory. See ${CLAUDE_SKILL_DIR}/references/pipeline.md for distribution channels and pilot criteria.

---

Operating Principles

1. Products, not services. Build once, sell many. No trading hours for dollars.
2. Slow and intentional. Compound over years, not months.
3. Charge from day one. Validate with revenue, not vanity metrics.
4. Build for yourself. The customer is someone like you.
5. Stay small, stay profitable. Basecamp model, not VC model.
6. Sectoral density, not scale. Design Machines does not chase scale in the VC sense. It builds for sectoral density — more co-ops per region, more co-op-to-co-op trade, more federation infrastructure. Co-op developers fluently distinguish scale out (replicate the model in many places) from scale up (grow one org huge). DM scales out, never up.
7. Decolonize governance language. Most worker co-op governance terminology was written by lawyers in the 1970s and imported wholesale by software developers. Per Chris Galloway (April 21, 2026), Assembly ships opinionated plain-language defaults — "how we're splitting this year's surplus" instead of "patronage allocation subject to qualified written notice of allocation." Bylaws and statutory filings keep the legal terms; everything members read uses the plain-language version. Source of truth: plugins/council/skills/governance/references/plain-language-glossary.md.

---

Brand Language Territory

Every name and piece of marketing must feel like it belongs to at least one of three worlds:

Factories (how things are made) + Labor (who makes them and under what conditions) + Publishing (how ideas spread)

The Naming Test

Before naming anything: Can I picture this in a factory, a union hall, or a print shop? If yes, it fits. If it sounds like a SaaS landing page, start over.

Word Banks

Factory & Industrial: Anvil · Bellows · Forge · Gauge · Jig · Lathe · Loom · Press · Proof · Spindle · Voltage · Workbench

Labor & Democracy: Ballot · Charter · Dispatch · Guild · Ledger · Local · Mandate · Muster · Quorum · Roll Call · Steward · Union Hall

Publishing & Editorial: Broadsheet · Byline · Chapbook · Colophon · Folio · Galley · Imprint · Leaflet · Masthead · Newsprint · Typeset · Woodcut

---

Go-to-Market: The Trojan Horse

Primary: channel-first via co-op developers, federations, and incubators

Primary audience is co-op developers, federations, and incubators (USFWC, CWCF, DAWI / Democracy at Work Institute, Cooperation Works!, Cooperatives Europe). They recommend Assembly to dozens of member co-ops each. The federation is the channel; the co-ops are the members. Assembly is what their consulting engagement leaves behind.

What this audience cares about: pipeline economics (compressing TA work from months to days), curriculum embedding, client retention, federation-recognized co-branded artifacts, sectoral density. What they fear: software that competes with their consulting revenue, per-seat pricing that punishes adding member co-ops, tools whose politics are illegible. Pitch Assembly as augmenting TA work, not replacing it; pitch federation-tier pricing that gets cheaper-per-co-op; lead with the politics.

Full pitch material lives in the audience skill (plugins/design-machines/skills/audience/references/developer-federation-pitch.md).

Also primary: existing worker co-ops directly

The Tuesday-morning member; the small-to-mid co-op (5-30 members) that knows what governance is supposed to feel like and is currently making it work despite their tools, not because of them. Pitch material at plugins/design-machines/skills/audience/references/coop-pitch.md.

Secondary / long-term: designer-developer funnel via Live Wires

Lead with internal operations, not public-facing design.

Why:

• Co-ops can get funding for operations infrastructure more easily than marketing websites
• Focuses clients on functionality, bypasses subjective aesthetic debates
• Design language gets established as side effect → public website extension later is cheap
• Internal system itself is propaganda — modern operations make co-ops less scary to outsiders

Channels:

• Co-op incubation pipelines (new co-ops twice yearly)
• Co-op developers and meta co-ops
• Provincial and national co-op federations — grant money and referral networks
• Credit unions as funders AND clients
• Content: blog, YouTube, resources
• Speaking: co-op conferences, design conferences

This Live Wires-led funnel is deprioritized for now in favor of the channel-first federation play above. Live Wires stays in the catalog and stays useful as the top-of-funnel for designers who eventually become Assembly clients; it just is not the lead motion any more.

---

Design System

Design Machines uses a unified visual identity across all products: GT Standard typeface (variable, Grilli Type), a 7-family color palette (Purple, Red, Orange, Gold, Green, Blue, Iron) built in OKLCH, and a three-layer token architecture (primitives > semantic tokens > scheme classes). Brand anchors: Purple-800 #220d46, Gold-400 #ffcb09, Red-500 #ed1d26.

Full palette, token maps, scheme inventory, product assignments, and accessibility guidelines are in references/design-system.md.

---

Competitive Landscape

What co-ops actually use today: Spreadsheets, Google Docs, Notion (if technical), generic tools (Monday, Asana), paper, nothing at all, expensive custom solutions ($50K+).

The real competition: Cobbled-together free tools plus hope.

---

Systems Dynamics — Feedback Loop Map

Design Machines operates as a system of interdependent feedback loops. When making strategic decisions — pricing, prioritization, partnerships, product scope, time allocation — check which loops a decision feeds or starves.

The Loop Map

Four reinforcing loops (compound growth — feed these):

Loop	Name	Mechanism	Current State
R1	Product-Adoption	Assembly improves → co-ops run better → success stories → more co-ops adopt → revenue + feedback → Assembly improves	Not yet turning. TACO prototype is the first turn of the crank.
R2	Solidarity Ecosystem	Solid State incubates → creates demand for governance tools → Assembly gets pilots → governance health validates incubation → pipeline grows	Active but pre-revenue. Relationships strong, no money flowing yet.
R3	Propaganda	Co-ops succeed using Assembly → visible proof co-ops work → more people consider conversion → larger sector → more demand	Dormant. Needs R1's first turn before it can start.
R4	Knowledge	Systems thinking library → informs governance design patterns → better loops encoded in Assembly → pilot data → deeper understanding	Active. RAG library indexed, research informing product design.

Two balancing loops (constraints — manage these):

Loop	Name	Mechanism	Current State
B1	Investment Window	Personal savings depleting monthly. No external income. Window determines how many turns R1 gets. Check ai-memory entity "Design Machines OÜ" for current burn rate, runway, and grant status.	Amber. Controlled investment, not crisis.
B2	Complexity vs Adoption	Governance requirements demand features → complexity increases → adoption friction. Countered by design principles: compliance as byproduct, let formality emerge, be easier than the workaround.	Managed. Design principles are the lever.

Archetype Warnings

Three systems archetypes from the DM library (Senge, Meadows, Kelly) that apply directly. Flag these during strategic decisions:

Eroding Goals (Meadows): Financial pressure gradually lowers ambition. The danger isn't stopping Assembly work — it's "just taking one freelance project" that becomes the new normal, and Assembly development slows to a crawl. The goal erodes so gradually you don't notice. Test: "Am I making this decision because it serves the mission, or because it reduces financial anxiety?"

Growth and Underinvestment (Senge): Investing in infrastructure before the growth signal is real. Don't over-engineer the baseplate before TACO is live and giving real feedback. Extract abstractions from what you've learned, not from what you assume. Test: "Has a real pilot validated this need, or am I building for a future that doesn't exist yet?"

Shifting the Burden (Senge/Meadows): The symptomatic solution (freelance income) undermines the fundamental solution (Assembly revenue). Every hour on freelance is an hour not shipping the prototype that generates its own revenue. Test: "Does this income source compete with or complement Assembly development time?"

Priority Sequencing (from systems analysis)

All loops wait on R1's first complete turn. The sequencing:

1. TACO prototype ships → proves the model
2. TACO pays something (grant-sized $2-4K CAD) → first revenue
3. Second pilot (Solid State or another co-op from pipeline) → doubles feedback + revenue
4. Cooperators grant lands (timing unknown) → extends runway
5. R1 starts compounding → feeds R2 (ecosystem), R3 (propaganda), R4 (knowledge)

Sprint time allocation during this phase: 80% TACO prototype, 10% pipeline warmth, 10% build-process content.

Leverage Points (Meadows' hierarchy, applied to DM)

From highest to lowest leverage:

1. The goal of the system (Level 3): "Make co-ops a no-brainer." This is set. Don't dilute it.
2. The structure of information flows (Level 6): Assembly changes how governance information reaches members. This is the product thesis — inhabited interfaces, not archived snapshots.
3. The rules of the system (Level 5): Grant-sized pricing makes co-op funding infrastructure subsidize R1's spin-up. The pricing structure is a systems intervention.
4. Feedback loops (Level 7-8): The solidarity ecosystem (R2) is a reinforcing loop that capitalist competitors cannot replicate because their incentive structures fight against coordination. This is DM's structural moat.

Source Library (indexed in RAG)

Key texts informing this analysis — searchable via rag_search:

Book	Author	Key Contribution
Thinking in Systems	Donella Meadows	Leverage points hierarchy, stocks and flows, archetype recognition
The Fifth Discipline	Peter Senge	Systems archetypes (Growth and Underinvestment, Shifting the Burden, Eroding Goals), mental models, team learning
Owning Our Future	Marjorie Kelly	Generative vs extractive ownership, feedback loops in cooperative design, Mission-Controlled Governance
Introduction to General Systems Thinking	Gerald Weinberg	Foundational systems theory, observation and partitioning
Systems Thinking Tools	Daniel Kim / Pegasus	Practical toolkit — causal loop diagrams, behavior over time, management flight simulators. Includes Forrester's "System Dynamics and the Lessons of 35 Years"

When making a strategic decision, search the RAG with the relevant archetype or concept name. The library has deep material on all of these.

---

Data Source Convention

Reference files contain stable structural knowledge: pricing tiers, engagement phases, brand language rules, catalog architecture, partner archetypes. These rarely change.

ai-memory contains dynamic state: current pipeline status, active financial runway, specific meeting dates, recent decisions, relationship updates. Search for entity names like "Design Machines OÜ", "Assembly", partner names, or client names.

When in doubt: check the reference file first (faster, always available), then ai-memory for current state.

Companion Skills

Skill	Plugin	When to Load
audience	design-machines	External communication, pitching, federation outreach, competitive analysis (this skill is internal business decisions; audience is everything aimed outward)
voice	ghostwriter	Writing any DM content, copy, or communications
social-media	ghostwriter	Platform-specific content distribution
governance	council	Co-op domain knowledge for client conversations
decolonial-language	council	Terminology positions (patronage refunds not dividends; surpluses not profits; member capital not equity) referenced from the strategy skill's positioning sections
development	assembly	Technical Assembly architecture discussions
livewires	live-wires	Live Wires framework positioning and technical details

Reference Files

Load specific reference files based on needs:

Topic	Reference File	When to Load
DM Catalog system	${CLAUDE_SKILL_DIR}/references/catalog.md	Naming new products, catalog entries
Assembly architecture	${CLAUDE_SKILL_DIR}/references/assembly.md	Technical discussions, scoping
Partnerships	${CLAUDE_SKILL_DIR}/references/partnerships.md	Partner coordination, pipeline
Revenue model	${CLAUDE_SKILL_DIR}/references/revenue.md	Pricing, proposals, financial planning
Brand language	${CLAUDE_SKILL_DIR}/references/brand-language.md	Naming, copy, marketing
Design system	${CLAUDE_SKILL_DIR}/references/design-system.md	Colors, typography, tokens, schemes, product assignments
Pipeline & pilots	${CLAUDE_SKILL_DIR}/references/pipeline.md	Client conversations, preparation
Survival reframe	${CLAUDE_SKILL_DIR}/references/survival-reframe.md	Citable survival statistics with sources; positioning from strength, not rescue
Two moats	${CLAUDE_SKILL_DIR}/references/two-moats.md	Integration + enforcement positioning with concrete BC Act enforcement examples