Decision Vectors by Business Type

Abstract

Your AI coworkers weigh every judgement against the same named set of decision axes, defined in full in the Decision Vectors reference. What changes between a plumbing firm and a credit union is not the axes — it is which ones dominate, which ones are hard gates that are never traded away, and which craft-specific axes come into play.

This page states that, per business type, so a prospective user can check whether the platform’s judgement actually fits how their business works — rather than taking “it understands your industry” on trust.

Why the axes do not change per business type. An install may re-weight the shared set; it may never declare its own axes. If every business had private axes, no two installs’ judgement could be compared, and nothing learned in one could ever generalise to another. Richness lives inside the professions instead — see §7.3 of the reference.

The part you actually fill in

Before the axis detail below, there is a smaller, owner-facing layer that matters more day to day: five business stance questions, pre-answered from your business type and editable in plain English on the business stance surface (/coworker-decisions/stance). Three of them carry a spend ceiling in money, which is what turns a stated value into real delegated authority.

The defaults are genuinely different per business type, not relabelled. On the goodwill question alone:

The other four vectors shift the same way. A clinic’s spend ceiling is $500 because a stock-out delays care; a nonprofit’s is $150 and written around the fact that every dollar carries a donor’s trust; public sector’s pricing stance says fees change by public decision and never as a service gesture. Twelve business types override at least one vector today; the rest inherit the generic set and are edited from there.

One boundary stated plainly: these stance answers are retrieved as your organisation’s own policy and weighed with the evidence discounting described in the reference — they are not yet expressed as weights on the named axes below. See §8.2 of the reference.

How to read an entry

Each business type below states four things.

• Load-bearing stage — the point in that business’s value stream where the money and the risk concentrate. Judgement quality matters most here.
• Dominant axes — what the coworker weighs most heavily in this line of work.
• Hard gates — axes that are not traded off. A high-scoring option that fails one of these is not chosen faster; it is refused, and a human is asked.
• Craft axes in play — the profession-local axes this work activates, each of which rolls up onto a shared axis when the decision leaves that profession.

A note on cost axes: blast_radius, human_cognitive_load, vendor_lock_in, business_disruption and operator_effort are named for the bad thing. “Dominant” for those means the coworker works hard to keep them low.

You go to the customer — field and dispatch

Someone travels to a property or a vehicle. The shared shape of judgement here: the decision is made before anyone can see the site, and it is expensive to reverse once the van has moved.

Across this group, speed_to_value and evidence_density carry the most weight — the fastest routing decision is worthless if the technician arrives without the context to finish the job — and reversibility is structurally low for everything after dispatch, which is why the coworker proposes and a human approves rather than dispatching autonomously.

Trades & home services

• Load-bearing stage: route and dispatch the technician.
• Dominant axes: speed_to_value (an emergency call-out buried under routine quotes is a lost job), evidence_density (property type, equipment, prior visits — the first visit should be the fix), operator_effort kept low across the inbox-to-dispatch handoff.
• Hard gates: public_safety. Licensed work — electrical, gas, refrigerant handling, fall protection — surfaces its readiness gate. The coworker flags readiness and never improvises safety or code advice, at any speed.
• Craft axes in play: capacity_utilization (operations) — crew and truck time already paid for, rolling up onto cost_efficiency.

Automotive services

• Load-bearing stage: schedule, or dispatch right now.
• Dominant axes: speed_to_value — roadside and lockout work is genuinely real-time and must not be forced into a booking calendar — balanced against evidence_density, because a quote without the VIN is a guess about parts and price.
• Hard gates: public_safety and governance_compliance, which here are the same thing: post-install ADAS recalibration and driver hours-of-service are tracked as steps, not afterthoughts.
• Craft axes in play: capacity_utilization across the two dispatch modes.

Moving & logistics

• Load-bearing stage: schedule the pickup and the route.
• Dominant axes: evidence_density (origin, destination, size, condition and disposal notes — the record is what settles a dispute later), cost_efficiency on routes, long_term_maintainability on recurring business accounts that should not be re-quoted every time.
• Hard gates: governance_compliance — hours-of-service limits and chain of custody for medical and legal courier work are limits, not preferences.
• Craft axes in play: capacity_utilization (truck and crew loading).

Security services

• Load-bearing stage: assign officers, or complete the install.
• Dominant axes: blast_radius kept low — each client site is contractually separate, and bleeding one client’s assignments or data into another’s is the failure that ends the business — plus evidence_density on incident records.
• Hard gates: public_safety and governance_compliance: officer licensing and low-voltage certification are verified up front and over time, and incident response is documented.
• Craft axes in play: business_disruption and evidence_confidence (security) — how conclusive an incident assessment is, distinct from how much material exists, and whether the response itself disrupts the client’s operation. They roll up onto blast_radius and evidence_density.

Customers come to you — appointments and booking

The calendar is the product. The shared shape here: decisions are individually reversible — a slot can be moved — but they touch personal records, and the person is often already in the building.

human_cognitive_load and data_privacy dominate across the group, and reversibility is high enough that the coworker can be trusted with more of the routine work than in field dispatch.

Clinics & wellness

• Load-bearing stage: deliver care, informed by the record.
• Dominant axes: data_privacy and evidence_density — the record being present at the appointment is what keeps care safe — with human_cognitive_load weighed hard, because a clinician who has to reconstruct context is a clinician not looking at the patient.
• Hard gates: public_safety above everything. This is the platform’s most safety-critical consumer setting, and the boundary is absolute: the coworker stays in scheduling and operations. It does not diagnose, does not triage, and does not handle a clinical or mental-health crisis — crises route to emergency services. governance_compliance (health-information rules for home health and phlebotomy) is likewise never traded for speed.
• Craft axes in play: none of the demoted craft axes; the weight sits on the spine, which is correct — safety and privacy must stay comparable to everything else in one ledger.

Beauty & personal care

• Load-bearing stage: assign the practitioner and the slot.
• Dominant axes: capacity_utilization in effect — double-bookings and no-shows are the whole economics of the day — with speed_to_value and low operator_effort on self-rescheduling, and legibility_of_consequence so an owner can see what a schedule change will do before approving it.
• Hard gates: none regulatory in most of this group; the effective floor is legibility_of_consequence — the coworker does not move a booking in a way the owner could not have predicted.
• Craft axes in play: capacity_utilization (operations), rolling up onto cost_efficiency.

Pet services

• Load-bearing stage: capture the pet’s context.
• Dominant axes: evidence_density — size, coat, temperament, allergies and medication decide whether the service arrives prepared — and reversibility across the three booking shapes (slots, date ranges, recurring).
• Hard gates: public_safety where an animal’s or a handler’s welfare is involved: vaccination status at intake and controlled-substance handling for mobile veterinary work.
• Craft axes in play: capacity_utilization for kennel and route capacity.

Fitness & recreation

• Load-bearing stage: renewal — not the first sale.
• Dominant axes: long_term_maintainability in its commercial sense: the membership relationship is the asset, so a decision that wins a signup at the cost of the renewal scores badly. cost_efficiency and capacity_utilization on class patterns follow.
• Hard gates: data_privacy on liability intake (age, emergency contact) and customer_consent_state before any retention contact.
• Craft axes in play: customer_consent_state (marketing) rolling up onto governance_compliance; capacity_utilization (operations).

Education & training

• Load-bearing stage: the inquiry or enrolment.
• Dominant axes: evidence_density on the learner-versus-payer distinction — a parent books and a child attends, and an instructor who does not know which is which is a problem — plus long_term_maintainability across terms and packages.
• Hard gates: public_safety and governance_compliance where minors are involved: guardian consent and instructor qualification are gates, never trade-offs.
• Craft axes in play: capacity_utilization (cohort and instructor scheduling).

You sell goods — catalogue and orders

Browse, cart, checkout, fulfil. The shared shape: the moment of commitment is fragile and the decision window is seconds, so speed_to_value and low operator_effort weigh heavily — but reversibility is genuinely high (an order can be amended, a refund issued), which is what makes it safe to weight speed this hard.

Retail & goods

• Load-bearing stage: checkout and the order reference.
• Dominant axes: speed_to_value and operator_effort kept low through the cart; evidence_density on the catalogue itself, because a missing product image or price is a lost sale before anyone talks to anyone.
• Hard gates: data_privacy on delivery and payment details; customer_consent_state before any repeat-buyer contact.
• Craft axes in play: customer_consent_state (marketing) → governance_compliance; capacity_utilization (operations) on stock and reorder points.

Food & hospitality

• Load-bearing stage: reserve, quote, or order — three shapes on one engine.
• Dominant axes: legibility_of_consequence on table and duration blocking (an owner must see what a reservation change does before approving it) and speed_to_value on the booking itself.
• Hard gates: public_safety. Allergen and dietary disclosure is treated as a duty of care, not a free-text field that may or may not reach the kitchen. It is never traded for a faster booking.
• Craft axes in play: capacity_utilization (covers, kitchen and event capacity).

You serve members and community — members, residents and regulated

Donors, members, residents, ratepayers, borrowers. The shared shape is the sharpest inversion on the platform: a trust gate comes before the value stream. Eligibility, disclosure and the duty to serve everyone are settled first, and only then does anything proceed.

Across this group governance_compliance and data_privacy dominate, and speed_to_value is deliberately down-weighted — moving faster through a statutory process is not a benefit.

Nonprofits & community

• Load-bearing stage: the gift or the joining.
• Dominant axes: governance_compliance (a gift produces a receipt, not a bill, and there is no account behind it — getting this wrong is not a cosmetic error) and legibility_of_consequence on member-owned structures and profit sharing.
• Hard gates: governance_compliance on charitable and cooperative obligations — tax-relief framing, patronage allocation, member-democratic governance.
• Craft axes in play: customer_consent_state (marketing) → governance_compliance, on supporter contact.

HOA & property management

• Load-bearing stage: paying dues, or sending a request.
• Dominant axes: legibility_of_consequence — a resident must be able to see what a request or an approval will do — plus evidence_density (the unit or lot captured before a request moves) and low operator_effort for volunteer boards.
• Hard gates: governance_compliance on covenants, reserve disclosure, and landlord-tenant and deposit law.
• Craft axes in play: capacity_utilization for shared amenity booking.

Public sector & civic

• Load-bearing stage: processing and routing the resident’s request.
• Dominant axes: governance_compliance above all — funding is statutory rather than market-priced, and there is a universal obligation to serve every resident — with evidence_density on the archival record and legibility_of_consequence on routing.
• Hard gates: the heaviest set on the platform. Public-records law and its statutory clocks, drinking-water and discharge regulation for utilities, and law-enforcement information handling — where the coworker’s posture is to refuse criminal-justice-information lookups outright and route to intake only. A refusal is a correct outcome here, not a failure to be optimised away.
• Craft axes in play: evidence_confidence (security) → evidence_density, where a determination must be conclusive rather than merely supported.

Banking & credit unions

• Load-bearing stage: the application and its identity check.
• Dominant axes: governance_compliance and data_privacy, with blast_radius held low by architecture: this platform is the engagement layer. It records applications and obligations and never moves money — core banking stays with the institution. That is a deliberate ceiling on how much damage a wrong judgement can do.
• Hard gates: the strictest profile shipped. Deposit-insurance and mortgage-lending disclosure must precede the call to action; identity and anti-money-laundering checks gate intake; rate and fee disclaimers are mandatory. None of these is weighed against conversion.
• Craft axes in play: customer_consent_state (marketing) → governance_compliance; evidence_confidence (security) on identity determinations.

You rent, build and advise — projects, rentals and engagements

Value is delivered over time under a scope or an agreement. The shared shape: the decision that matters was made months before its consequence lands, so durability outweighs speed almost everywhere in this group.

long_term_maintainability and evidence_density dominate, and blast_radius is held low by client isolation.

Equipment & storage rental

• Load-bearing stage: the use period itself — the asset is out of the pool and earning or idle.
• Dominant axes: capacity_utilization above all (a pooled asset double-booked or sitting idle is the entire margin), then evidence_density on handover and return condition, which is what settles a damage dispute.
• Hard gates: governance_compliance on deposit law, which varies by jurisdiction.
• Craft axes in play: capacity_utilization (operations) → cost_efficiency. This is the archetype where that projection is most visible: fleet utilisation is the cost story.
• Note: this is the business shape scoped for the first pilot of an external situational signal — a seasonal demand correlate. Consistent with the reference’s §5 rule, such a signal must be validated against recorded outcomes before it is allowed to influence any live score.

Home building & construction

• Load-bearing stage: the inquiry and the quote — an educated quote decides whether the project is profitable eighteen months later.
• Dominant axes: long_term_maintainability (selections, permits, warranty obligations all outlive the sale), evidence_density on buyer qualification, and reversibility weighed honestly: it collapses once ground is broken.
• Hard gates: public_safety and governance_compliance — building codes, permits and inspections are jurisdiction-specific gates, and design controls and warranty obligations bind.
• Craft axes in play: capacity_utilization on crew and trade scheduling.

Professional services

• Load-bearing stage: the quote and the scope.
• Dominant axes: evidence_density linking the original question to the quote so nothing drifts, and blast_radius kept low through strict per-client isolation — for an IT managed-services firm especially, one client’s files reaching another is the end of the firm.
• Hard gates: governance_compliance and public_safety in the regulated professions: the coworker does not give legal or financial advice, and points to a qualified professional instead. Practitioner licensing and client-data isolation are gates.
• Craft axes in play: operational_independence and vendor_lock_in (devops-platform) for managed-services work — both rolling up onto long_term_maintainability; evidence_confidence (security) on incident determinations.

Software & platforms

• Load-bearing stage: turning an inquiry into a tracked backlog item.
• Dominant axes: long_term_maintainability and schema_grounding — building on substrate that already exists rather than alongside it — plus reusability, since a feature that serves one install and no other is a cost centre.
• Hard gates: data_privacy where the platform touches customer data.
• Craft axes in play: the full software and platform set — schema_grounding, reusability, operational_independence, vendor_lock_in. All four are profession-local, and all four project onto long_term_maintainability (with operational_independence also onto blast_radius). This is the clearest illustration of why the demotion policy exists: this is the profession whose vocabulary would otherwise have colonised the shared space every other business type must reason over.

Where this is going

The per-business-type emphases above are authored and reviewed, not measured. The work to derive them from each profession’s own corpus and from real decision history — and to record how many independent organisations corroborate each one — is designed and under way rather than done. The reference’s §6 and §7.2 state exactly what is measured, what is authored, and what is not yet built.

Full axis definitions, sign conventions, sourcing classes, and the organisation-selection contract are in the Decision Vectors reference.