Voice of the Customer
Voice of the Customer becomes decision material only when corpus coverage, privacy, themes, disagreement, evidence strength, and follow-through remain connected.
Anecdotes become dangerous when a small or biased sample is retold as what customers want. The loudest note can hide missing channels, opposing evidence, and the difference between a pain point and a feature request.
This playbook defines the corpus, protects identifying data, codes each item, preserves disagreement, and assigns a bounded disposition. It reports what the supplied feedback says without pretending the corpus represents the entire customer base.
The readout represents the corpus you have, not every customer you serve.
CORPUS Period: July Items: 6 Sources: support 3, interviews 2, sales 1 Customer scope: supplied anonymized items only Missing coverage: segments not supplied
THEMES Export controls hard to find: pain point, 2 support items New search praised: praise, 1 support item Saved views: request, 2 interview items Approval history: request, 1 sales item
SETUP EVIDENCE Confusing: 2 supplied comments Easy: 1 supplied comment Conclusion: mixed evidence in this corpus Prevalence across customers: not established
DISPOSITION Saved views: Investigate Reason: two interview requests, no broader coverage supplied Action owner: Owner needed Due: Date needed Loop-close owner: Owner needed
voice-of-the-customer/
├── SKILL.md ← synthesize and audit modes
├── meta.yaml ← version, sources, test prompts, changelog
├── LICENSE.md ← MIT
├── assets/
│ ├── voc-readout-template.md ← corpus, themes, and dispositions
│ └── voc-audit-scorecard.md ← evidence and coverage audit
└── references/
├── coding-and-signal-guide.md ← feedback coding and confidence
└── decision-and-loop-guide.md ← dispositions and loop closure"Synthesize this customer feedback into a VOC readout."