Design and operate a Voice of Customer program — from NPS/CSAT collection through qualitative analysis to roadmap integration.
| Metric | Question | Scale | When to Use |
|---|---|---|---|
| NPS | "How likely to recommend?" | 0-10 (Detractor 0-6, Passive 7-8, Promoter 9-10) | Relationship health, quarterly+ |
| CSAT | "How satisfied with [interaction]?" | 1-5 stars | Post-transaction, support close |
| CES | "How easy was it to [task]?" | 1-7 (strongly disagree→agree) | Post-task completion |
| PMF Score | "How disappointed if you couldn't use this?" | Very/Somewhat/Not | Product-market fit (target >40% "very") |
Timing triggers (pick ONE per user journey):
Segmentation: Split by plan tier, tenure, geography, and use-case. Compare NPS across segments — the delta tells you more than the absolute score.
Survey rules:
| Channel | Signal Type | Volume | Richness |
|---|---|---|---|
| In-app widget | Feature requests, bugs | High | Medium |
| Post-support CSAT | Service quality | Medium | Low |
| Email surveys (NPS) | Relationship health | Medium | High |
| Support tickets | Pain points | High | High |
| Social/review sites | Brand sentiment | Low | Medium |
| Sales call notes | Objections, gaps | Low | Very High |
| Community/forum | Power user needs | Medium | High |
Score each request: RICE = (Reach × Impact × Confidence) / Effort
| Factor | Definition | Scale |
|---|---|---|
| Reach | Users affected per quarter | Absolute number |
| Impact | Effect per user (Massive=3, High=2, Medium=1, Low=0.5, Minimal=0.25) | 0.25–3 |
| Confidence | Data backing (High=100%, Medium=80%, Low=50%) | 50–100% |
| Effort | Person-months | Absolute number |
# Example RICE calculation
reach = 2000 # users/quarter
impact = 2 # high
confidence = 0.8 # medium — have support tickets but no interviews
effort = 3 # person-months
rice = (reach * impact * confidence) / effort # = 1066
bug, feature-request, ux-friction, praise, pricingTagging rules: Use max 3 tags per item. Maintain a shared taxonomy (see references/feedback-taxonomy.yaml). Review and merge tags monthly.
Respond → Act → Communicate
│ │ │
▼ ▼ ▼
Acknowledge Ship fix/ Notify the person
within 48h feature who requested it
Templates: See references/feedback-response-templates.md
Trigger on cancellation. Keep to 3 questions max:
Analyze monthly. If >20% cite same reason, escalate to product leadership.
| Phase | Audience | Size | Duration | Goal |
|---|---|---|---|---|
| Alpha | Internal + 5 power users | 10-20 | 2 weeks | Find breaking bugs |
| Closed Beta | Opted-in segment | 50-200 | 2-4 weeks | Usability + edge cases |
| Open Beta | Feature-flagged rollout | 5-20% of base | 1-2 weeks | Scale validation |
Recruit from NPS promoters (9-10) first — they're invested and forgiving.
| Tool | Best For | Pricing Model | Key Strength |
|---|---|---|---|
| Canny | Public feature voting boards | Per-tracked-user | Transparent roadmap |
| ProductBoard | Feedback→roadmap workflow | Per-maker seat | Prioritization frameworks |
| Pendo | In-app guides + analytics | Per-MAU | Combines feedback with usage data |
| Hotjar | On-page surveys + heatmaps | Per-session | Visual context |
| Delighted | NPS/CSAT automation | Per-survey-response | Simple, fast setup |
customer-requested labelSee references/feedback-roadmap-workflow.md for detailed integration diagrams.