# Marketing Psychology
Use this skill to explain customer behavior and apply psychological principles ethically to marketing decisions.
The goal is not to sprinkle bias names into copy. The goal is to identify what behavior needs to change, why the current path fails, and which mental model suggests a practical improvement.
Context Loading
Before asking questions, check whether one of these files exists and read it:
- `.agents/product-marketing-context.md`
- `.claude/product-marketing-context.md`
- `./references/dependencies/conversint-current/references/org/positioning.md` and `./references/dependencies/conversint-current/references/org/value-signals.md` when advising Conversint
Use existing product, audience, and offer context before asking for more.
Intake
If not clear, ask:
- What behavior should change?
- Who is the buyer or user?
- Where in the journey is this: awareness, consideration, decision, activation, retention, referral?
- What currently blocks action?
- What asset is being improved: ad, landing page, pricing, email, onboarding, product flow, sales deck?
- What constraints or compliance rules apply?
Workflow
- **Define the behavior.** State the desired action and current blocker.
- **Pick relevant models.** Choose 3-6 models, not a long encyclopedia.
- **Explain the mechanism.** Say why the model applies in plain language.
- **Translate to marketing moves.** Give concrete edits, tests, prompts, page changes, or offer changes.
- **Add ethics and risk.** Flag manipulative uses, false scarcity, unsupported claims, or dark patterns.
- **Recommend a test.** State what to measure.
- **Route the learning.** If the recommendation changes messaging, offer, page, ad, or content strategy, define the hypothesis for `/marketing-learning-loop` and the metric that may later feed `/marketing-roi`.
Model Map
| Challenge | Useful models | |---|---| | Low conversions | Hick's Law, activation energy, BJ Fogg, EAST, friction cost | | Price objections | Anchoring, contrast, mental accounting, loss aversion, good-better-best | | Weak trust | Authority, social proof, reciprocity, pratfall effect, familiarity | | No urgency | Loss aversion, Zeigarnik effect, genuine scarcity, present bias | | Decision paralysis | Paradox of choice, defaults, choice architecture, recommendation framing | | Poor onboarding | Goal-gradient effect, IKEA effect, commitment and consistency, quick wins | | Churn | Endowment effect, switching costs, status-quo bias, habit loops | | Growth stall | Theory of constraints, 80/20, local vs global optima, compounding |
Core Models
Buyer Understanding
- **Jobs to Be Done:** Customers hire the product for an outcome.
- **Status-quo bias:** Change feels risky, even when the current process is painful.
- **Curse of knowledge:** Experts underestimate how confusing their own category feels.
- **Confirmation bias:** People trust messages that fit what they already believe.
- **Availability heuristic:** Vivid examples and recent stories feel more likely.
- **Mere exposure:** Repeated, consistent visibility builds comfort.
Persuasion
- **Reciprocity:** Give useful value before asking.
- **Commitment and consistency:** Small commitments increase later follow-through.
- **Authority:** Credible expertise lowers perceived risk.
- **Social proof:** Similar others create confidence.
- **Scarcity:** Works only when real; fake urgency damages trust.
- **Loss aversion:** Avoiding loss often motivates more than gaining upside.
Pricing and Offers
- **Anchoring:** The first number shapes perception.
- **Contrast effect:** Comparisons make value easier to judge.
- **Mental accounting:** Framing affects how the same price feels.
- **Good-better-best:** Three options can guide decisions.
- **Rule of 100:** Under $100, percentage discounts often feel larger; over $100, dollar discounts often feel larger.
Behavior Design
- **BJ Fogg:** Behavior requires motivation, ability, and a prompt.
- **EAST:** Make the desired behavior easy, attractive, social, and timely.
- **Hick's Law:** More choices slow decisions.
- **Goal-gradient effect:** Visible progress pulls users forward.
- **Activation energy:** The first step must feel easy.
Output Format
```markdown ## Behavior to Influence [desired action and blocker]
## Best-Fit Models 1. **[Model]** - [why it applies]
## Recommended Changes | Change | Model | Implementation | Metric |
## Ethical Guardrails - [what not to do]
## Test Plan - Hypothesis: - Variant: - Metric: ```
Quality Rules
- Prefer specific behavioral diagnosis over generic persuasion advice.
- Do not recommend deceptive urgency, fake scarcity, hidden fees, or dark patterns.
- Tie every model to a concrete action.
- Separate what is likely from what must be tested.
- In regulated categories, avoid psychological pressure that could be interpreted as coercive or misleading.
- Do not treat persuasion as success by itself. The useful proof is whether the changed framing improves qualified response, pipeline, revenue, or a clearly named learning.