skills/diagnose-marketing-structure/references/core-principles.md

# Marketing Compass Core Principles

## Contents

1. Purpose and definition
2. Shared operating principles
3. Evidence levels
4. Shared guardrails

## 1. Purpose and definition

Marketing Compass is a decision system, not a tactics catalog. It decomposes the structure that produces business outcomes, identifies the maximum bottleneck, and defines what to do, what not to do, what evidence would change the judgment, and where to stop.

Marketing seeks the business outcomes of being chosen and chosen again. Its mechanism is intervention in the customer's future probability of choice through demand recognition, memory and recall, meaning, comparison, availability, friction, experience, continuation, and relationship.

Marketing communications are a means within marketing. Their direct responsibility normally ends at contact, attitude change, recall, trust, and the path to action. Attribute sales directly only when a defensible counterfactual and control of material non-communication variables exist.

## 2. Shared operating principles

1. **Revenue is a combined result.** Do not assign it to advertising, sales, product, or another single cause without decomposition.
2. **Find the bottleneck before selecting tactics.** A channel name in the request is not a diagnosis.
3. **Use named growth states.** Prefer `demand → recall/acquisition → choice/value realization → continuation/relationship`; do not rely on ambiguous layer numbers.
4. **Separate measurable from important.** Clicks and conversions are visible; recognition, understanding, preference, recall, and trust may not be.
5. **Design a decision boundary.** Use analysis for direction, anomalies, hypotheses, and ranges; connect it to signals and actions.
6. **Separate optimization from threshold defense.** Smooth, observable response can be optimized. Delayed or irreversible loss requires a provisional floor, small changes, anomaly detection, and restoration rules.
7. **Include implementation reality.** Operating routines, evaluation systems, incentives, politics, and reversibility affect whether a recommendation is correct.
8. **Frameworks are subordinate to reality.** Use 3C, 4P, funnels, and similar models as tools, never as proof that the diagnosis is complete.

## 3. Evidence levels

| Level | Basis | Permitted claim |
|---|---|---|
| A | Controlled experiment or strong counterfactual | State conditional incremental or causal effect |
| B | Multiple independent sources point the same way | State a strong hypothesis |
| C | Observational data, correlation, or a single source | State consistency, indication, or hypothesis |
| D | Experience, analogy, or insufficient evidence | State an exploratory hypothesis and limit action |

Always separate facts, interpretations, hypotheses, and unknowns.

## 4. Shared guardrails

- Do not call observational association causal proof.
- Do not use a meaning equation or hypothesis tree as an accounting formula.
- Do not treat a proxy as the underlying construct.
- Do not treat proprietary Marketing Compass terminology as an industry standard.
- Do not start large execution before identifying the primary bottleneck.
- Do not approve investment without a return definition, review period, and stopping rule.
- Do not increase MA when demand, ICP, or value realization is structurally absent.
- Do not make a large irreversible brand reduction to discover a hidden threshold.
- Do not use customer bias through deception, false scarcity, or misleading framing.
- Do not improve a KPI by damaging customer experience or trust.
- Do not let organizational feasibility become an excuse to preserve a known failure.