Marketing-psychology collects decision and behavior models that can help investigate hesitation, confusion, or incomplete actions. Use it on a defined problem. It cannot read minds, and naming a psychological principle does not prove that a copy change will improve conversion.
Install and invoke
Run this in your project terminal to install only this skill for Claude Code:
npx skills add coreyhaines31/marketingskills --skill marketing-psychology -a claude-code
Node.js and npx must be available. For Codex, replace -a claude-code with -a codex. In the agent conversation, explicitly ask it to use marketing-psychology and supply your material. The installation command belongs in the terminal, not in your content brief.
Name one observable action
Instead of asking how to make people buy, identify a page and action, such as sending the first invitation after reading its explanation. Supply the page, feedback, and available funnel evidence before requesting an interpretation.
Example brief:
Use marketing-psychology on onboarding.html and feedback.md.
The target is creating a first review project. Many users stop at the instructions field.
Choose only two relevant models. For each, give page evidence, a possible explanation, and a minimal change.
Do not add countdowns, invented customer counts, or preselected paid options. Explain how to check whether the change helps.
First inspect ordinary problems: a broken form, unfamiliar terminology, or unavailable input material may explain the behavior more directly. Fogg’s own description centers on motivation, ability, and a prompt converging at the same moment. Use that lens to ask whether the person wants to act, can do it easily, and receives a suitable cue.
Change one important barrier
A demanding form may need a genuine example or fewer unnecessary fields. An unclear next step may need a button that explains its result. Concern about irreversible consequences may need an accurate explanation of existing undo or exit options. Recommendations must reflect actual capabilities, not invented refund or migration promises.
The library includes anchoring, social proof, and scarcity. Their use requires real conditions. Price reframing should retain the actual billing period; reviews need provenance; availability limits and deadlines must exist. False information undermines the understanding and trust the design should create.
Preserve the original and revised versions, state which barrier the change should address, and verify through observation or an appropriate data comparison. With limited samples, report usability feedback rather than a statistically meaningful increase. If the change fails, revisit the action and evidence instead of layering more terminology or interpreting every outcome as confirmation.
Apply one model to one observable obstacle
Describe behavior before selecting a psychological label. If many visitors open pricing but do not continue, possible causes include too many plans, unclear costs, payment risk, or simply the wrong audience. For each explanation, ask the skill for supporting observations, contradictory observations, and the smallest testable change. A model helps form a hypothesis; it does not prove why an individual left.
A useful brief is: “Session recordings and support questions show people pausing at the plan comparison. Give at least two plausible causes. Use choice overload and perceived risk to propose one page change each, with an observation that would distinguish them. Do not claim either mechanism has been proven.” Select one main obstacle for the first iteration. Changing prices, layout, CTA, and trial terms simultaneously makes the result hard to interpret.
Review the ethical boundary too. Social proof must come from real material you may quote. Scarcity needs an actual capacity or time constraint. Defaults should not quietly add paid items users did not choose. Assess task completion, understanding, and later feedback together; a short-lived click increase is not evidence of durable trust.
Example: separate price concern from missing information
Suppose trial users spend a long time on the pricing page but rarely choose a plan. Provide a screenshot, the approved plan table, and presales questions. Ask: “Use marketing-psychology to propose two competing explanations: too many choices or uncertainty about total cost. For each, suggest one small page change and the behavior we should observe. Do not assume long dwell time means the price is too high.”
The proposed change must follow the evidence. If billing units are unclear, make units and add-ons visible. If choices are hard to compare, test clearer comparison criteria without quietly preselecting a paid tier. Reject fake countdowns or invented claims that “most users choose this plan.” Define a task before testing: have target users calculate their actual cost and select a suitable plan. Button clicks alone do not show that they understood pricing.
An external brand plan supplies hypotheses
An independent brand plan proposed authority evidence, honest failure stories, and participatory exercises as alternative persuasion paths for a founder page. These are design proposals, not evidence that any path improved conversion. On your own site, select one evidence-backed barrier, preserve the previous page, and test one change at a time. Do not present a planning list as observed customer behavior.