# Customer Case Study

> Create an evidence-safe customer case study and reusable proof assets from verified outcomes, approved quotations, and publication permissions. Use when turning a customer success story into a full case study, one-page summary, sales proof, or social excerpt.

- Skill: `majesticlabs-dev/customer-case-study` (Agent Skill)
- Install (CLI): `npx skillmds@latest add majesticlabs-dev/customer-case-study`
- Raw SKILL.md: https://api.skillmd.com/api/skills/majesticlabs-dev/customer-case-study/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Marketing & Growth
- Author: majesticlabs-dev (https://skillmd.com/u/majesticlabs-dev)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/majesticlabs-dev/customer-case-study

---


# Customer Case Study

## Boundary

Own the approved customer-success narrative and its derivatives. Do not invent results, strengthen causality beyond the evidence, alter quotations, or imply permission to identify a customer.

## Required Inputs

- Customer identity or anonymization requirement and publication permission status
- Problem, prior approach, selection context, implementation, and timeline
- Verified before-and-after measures with definitions, baselines, and time windows
- Exact quotations with speaker attribution and approval status
- Intended audience, channel, call to action, and review owners

## Workflow

1. Record what may be published, what must be anonymized, and who must approve the final asset.
2. Build an evidence table separating customer statements, measured outcomes, company claims, and unresolved gaps.
3. Verify metric definitions, baselines, comparison periods, attribution limits, and material confounders.
4. Frame the story around customer context, challenge, decision, implementation, result, and remaining limitation.
5. Select a headline result only when its evidence and permission support prominent use.
6. Preserve quotations exactly. Mark paraphrases as paraphrases and never manufacture a customer voice.
7. Draft the requested full, one-page, sales-proof, or social formats from the same evidence register.
8. Return the asset for customer, legal, brand, and factual approval as applicable before publication.

## Output

1. **Permission and evidence register**
2. **Full case study or requested primary format**
3. **Approved derivative snippets**
4. **Claims, attribution, and approval gaps**
5. **Publication checklist**

## Quality Gate

- Every result traces to supplied evidence and a defined period.
- Causal language matches what the evidence can support.
- Quotations remain exact and approval status is visible.
- Customer identity and sensitive details follow explicit permission.
- No asset is presented as publication-ready while required approval is missing.