Open Conduct

OPEN DRAFT STANDARD

The open standard behind the Conduct Quotient.

Open Conduct is a draft standard for asking a company's own people one question about a customer, combining that answer with what business records show about what the customer agreed to do, and turning the result into a private score the customer holds. The standard requires that ordinary service and legal rights never depend on it, that the worker's answer stay private from their manager, and that nothing reach a business except one approved eligibility result. Everything is published in draft so the right people can examine it before anything goes live.

CURRENT STATUS

Draft 0.5.1, open for review.

Every decision, every open question, and every test result is documented on this site. No one can receive a Conduct Quotient yet, and the work between here and there is the part that needs other people in it.

See what has been completed

START HERE

02

Inspect the protections

Review baseline service, sharing control, worker rights, correction, and prohibited uses.

Read protections
03

Browse the standard

Start with plain-language summaries, then open the detailed requirements and decisions.

Enter the specification

THE EXPERIENCE THE STANDARD MUST GOVERN

Three roles, three different views.

The calculation is only one part of the standard. It must also define what each person can see and do at each step.

See the larger interface examples
01

PRIVATE RESPONSE

Based only on how this customer interacted with you and others, would you want to serve them again?Definitely yesProbably yesProbably notDefinitely notNot enough interaction to say

The employee answers one question about the interaction.

The standard requires that managers never see an individual response and that abstaining carries no consequence. Neither protection has been tested with real workers.

02

YOUR CONDUCT QUOTIENT · ILLUSTRATIVE

4.81 TRANSPORTATION
How your CQ was formedHuman experience + approved business records
Using your CQYou choose when to ask for one additional benefit

The customer sees one score for one clearly named context.

They can inspect both inputs, question the result, and move between the company and industry scores available to them.

03

AIRLINE B · BENEFIT SYSTEM

EligibleWelcome benefit added for this booking

The receiving business gets one approved result.

The benefit can be added automatically. Frontline employees do not see scores, source history, or a failed attempt.

Illustrative research interfaces only. They are examples of the proposed information boundaries, not final product designs.

CURRENT DESIGN RULES

Five rules that will not bend.

Govern the two inputs separately

NCS records eligible human experience and VCI uses approved operational evidence. Each input must clear its own gates before synthesis begins, and those input gates are what stop one input from compensating for the other. The synthesis formula then caps how far a strong input can carry a weak one, and a minimum below which no score appears at all remains unspecified.

Lead with one contextual Conduct Quotient

When both inputs qualify, CQ presents one private score from 1.00 to 5.00. The two inputs remain available beneath it for explanation, correction, and challenge.

Name the scope and prove every transfer

A company CQ, an industry CQ, and an overall CQ are different claims. Any transfer needs an approved source, destination, and purpose that explain why the history is relevant.

Share the smallest useful result

The consumer starts the request. A receiving business learns only whether the customer qualifies for one approved benefit. Any wider disclosure would be a separate use requiring its own approval, and none is approved today.

Ask a person, never infer

A response has to be a deliberate act by a worker who was there. No system may generate, infer, pre-fill, rank, or suggest one, and facial, voice, tone, sentiment, affect, and biometric signals are prohibited as inputs. There is no comment box and no written account of a person. This is the mechanism the proposal exists to refuse.

BENEFIT DELIVERY

Conduct qualifies. Loyalty platforms deliver.

A Conduct implementation should produce the minimum approved result. A separate, replaceable system can turn that result into a real benefit through the business's existing channels.

Conduct governs eligibility

The Conduct layer determines whether a customer qualifies for one named, additional benefit under one approved policy. It governs the score, evidence, consent, proof, protections, correction, and challenge rules.

An incentive platform can deliver the benefit

Loyalty and promotion infrastructure can apply points, a perk, an upgrade, or another approved reward after receiving the limited result. Existing incentive-orchestration platforms are the obvious category, and no provider is named here because none has been chosen, approached, or endorsed.

The delivery platform sees four fields

The delivery platform receives the eligibility result, purpose, policy version, and expiry, and nothing else. It does not receive CQ, NCS, VCI, source companies, employee responses, or underlying events.

Any delivery provider stays replaceable

Conduct should work with multiple incentive, loyalty, commerce, and customer-experience systems. A delivery provider cannot recalculate CQ, redefine conduct, turn eligibility into a reusable audience, or make ordinary service depend on the result.

THE PUBLIC RECORD

Everything is on the record.

CONTRIBUTE

Your expertise can change this.

Consumers, workers, businesses, lawyers, technologists, disability advocates, labor experts, privacy experts, and measurement researchers all see something the specification currently misses. There are two ways in: a short form for a focused point, and an open discussion thread for the wider argument.

See both ways to contribute