What sources should I collect feedback from?

From multiple channels: support tickets, sales calls, user interviews, surveys, social media, and app reviews; gather across channels to avoid bias.

How do I translate feedback into product decisions?

Turn I want X into I am struggling with Y, cluster themes by underlying problem, weigh by target segment and recency, check against usage data, then map to actions: build, investigate, or deprioritize.

What if feedback contradicts usage data?

Treat data as evidence; investigate discrepancies, separate pain from desires, and re-prioritize insights with context before roadmapping.

feedback-synthesis

Scanned

npx machina-cli add skill Dragoon0x/Product-Skills/feedback-synthesis --openclaw

Files (1)

SKILL.md

2.9 KB

Feedback Synthesis (Customer Empathy)

Turn messy feedback into clear priorities the team can act on.

How to use

/feedback-synthesis Apply feedback synthesis constraints to this conversation.
/feedback-synthesis <feedback data> Synthesize the provided feedback into actionable insights.

Constraints

Collection

MUST gather feedback from multiple channels: support tickets, sales calls, user interviews, surveys, social media, app reviews
MUST capture the user's actual words, not just the PM's interpretation
SHOULD tag feedback with: user segment, frequency, severity, and source
NEVER rely on a single channel — each has its own bias
MUST distinguish between who gave the feedback and how representative they are

Pattern Recognition

MUST look for themes, not individual requests. One user wanting X is anecdote. Twenty is a pattern.
SHOULD cluster feedback by underlying problem, not by proposed solution
MUST weight feedback by user segment relevance — feedback from your target segment matters more
SHOULD track frequency and recency — a problem mentioned 50 times this month beats one mentioned 5 times last year
NEVER treat loud feedback as representative. Vocal users are a minority.

Signal vs. Noise

Feature requests are solutions, not problems. MUST translate "I want X" into "I'm struggling with Y"
MUST separate pain (something broken or missing) from desire (something nice to have)
SHOULD deprioritize feedback that only applies to a segment you're not targeting
MUST check if feedback contradicts usage data — sometimes what users say and do are different
NEVER let one large customer's request override patterns from many smaller customers without deliberation

Actionability

Every synthesis MUST end with a prioritized list of insights, not just a summary
Each insight MUST connect to a product decision: build, investigate further, or deprioritize
SHOULD present insights to stakeholders with evidence (quotes, volume, impact)
MUST close the loop: tell feedback sources what you did with their input
NEVER synthesize feedback and leave it in a doc. It MUST reach roadmap discussions.

Anti-Patterns

The Echo Chamber: only hearing from power users who love the product
Request-Driven Building: building exactly what users ask for instead of solving underlying problems
Feedback Fatigue: collecting so much that none of it gets processed
The Spreadsheet Graveyard: meticulously logging feedback that nobody reads
Recency Bias: overweighting whatever came in this week over long-term patterns

Source

git clone https://github.com/Dragoon0x/Product-Skills/blob/main/skills/customer-empathy/feedback-synthesis/SKILL.mdView on GitHub

Overview

This skill helps teams convert noisy feedback from multiple sources into actionable product insights. It emphasizes capturing actual user words, clustering feedback by underlying problems, and separating pain from desires so decisions are evidence-based and roadmap-ready. It also weighs input by target segment and recency to surface patterns that truly matter.

How This Skill Works

Gather feedback from multiple channels and capture exact user words along with segment, frequency, severity, and source. Cluster items by underlying problems rather than proposed solutions, and weigh patterns by target segment and recency. Translate pain points and desires into actionable insights with recommended actions (build, investigate, or deprioritize) and supporting evidence.

When to Use It

Feedback feels overwhelming or contradictory, making it hard to act.
The team can't agree on what users actually want.
Feedback from one group seems louder but doesn't reflect broader usage.
You need to prioritize the roadmap with limited resources.
You want to separate real pain from nice-to-have features and align with data.

Quick Start

Step 1: Collect feedback from support tickets, sales calls, user interviews, surveys, social media, and app reviews.
Step 2: Tag feedback with user segment, frequency, severity, and source; cluster by underlying problem.
Step 3: Produce a prioritized list of insights with evidence and map each to build, investigate further, or deprioritize; share in roadmap discussions and close the loop.

Best Practices

Collect from multiple channels (support, sales, interviews, surveys, social, reviews) and capture exact words.
Tag each item with user segment, frequency, severity, and source.
Cluster by underlying problem, not proposed solutions, and weigh by target segment; track frequency and recency.
Differentiate pain from desire and check against usage data; deprioritize non-target segments.
End with a prioritized list of insights linked to decisions and close the loop with sources.

Example Use Cases

A SaaS team collects 200 support tickets, 15 customer calls, and 3 surveys; discovers onboarding friction is the top pain point; builds guided onboarding.
Sales and interviews reveal pricing confusion; synthesis weighs by target segment and leads to revised pricing tiers and messaging.
Users complain about search quality; pattern yields UX revamp of search results and filters.
App reviews show stability issues; team prioritizes reliability fixes and QA improvements.
Power users push for a new feature; synthesis shows more users want core flow improvements and the roadmap shifts to stabilization before new features.

Frequently Asked Questions

Add this skill to your agents