customer-research

Analyze customer interviews, reviews, and community posts to extract pains, triggers, and personas.

Updated Jul 14, 2026
One-click install
npx skills add https://github.com/usmangurowa/sifter --skill customer-research-usmangurowa
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: customer-research
Source: https://github.com/usmangurowa/sifter/tree/main/.agents/skills/customer-research
Command: npx skills add https://github.com/usmangurowa/sifter --skill customer-research-usmangurowa

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Grounding positioning, messaging, and product decisions in real customer language is hard when research is scattered across transcripts, surveys, support tickets, and online communities. This Skill structures the analysis of existing research assets and guides the gathering of new voice-of-customer data from online sources. ## Core Features & Use Cases - Existing Asset Analysis: Extract jobs-to-be-done, pain points, trigger events, desired outcomes, and verbatim language from interview transcripts, surveys, support tickets, NPS responses, and win/loss notes, with confidence labels and sample-bias checks. - Digital Watering Hole Research: Mine Reddit, G2, Capterra, Hacker News, LinkedIn, app store reviews, YouTube comments, and SparkToro audience data using per-platform playbooks in references/source-guides.md. - Persona & Deliverable Generation: Build evidence-based personas (minimum 5-10 data points), VOC quote banks, JTBD maps, competitive intelligence summaries, and research gap analyses. - Use Case: You have 20 customer interview transcripts and want messaging insights. The Skill extracts themes, scores them by frequency and intensity, pulls money quotes, and produces a synthesis report ready to hand off to copywriting. ## Quick Start Ask the assistant to analyze your customer interview transcripts and produce a themed research synthesis report with verbatim quotes.

Frequently Asked Questions about customer-research

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I analyze customer interview transcripts for insights?

Extract jobs to be done, pain points, trigger events, desired outcomes, exact customer language, and alternatives considered from each transcript. Then cluster themes across interviews, score by frequency and intensity, and label each insight with a confidence level.

How do I do customer research with no existing interviews?

Start with digital watering hole research: mine Reddit communities, G2 and Capterra reviews, app store reviews, and niche forums where your ICP speaks unfiltered. Form hypotheses first, then validate with 5-10 data points before building personas.

What review sites should I mine for competitor research?

G2 and Capterra are the primary B2B sources, with competitor 4-star reviews being highest signal for buried complaints. For consumer products, use app store reviews, Trustpilot, and Amazon reviews, reading 3-star reviews first for honest tradeoffs.

How many data points do I need before building a customer persona?

Build personas only after collecting at least 5-10 independent data points from a consistent segment, such as interviews, reviews, or community posts. Personas built from fewer sources risk being invented rather than evidence-based.

What are the limitations of online review and community research?

Online reviewers skew toward power users and strong opinions, support tickets skew toward problems, and Reddit skews technical and skeptical. Weight sources from the last 12 months and label insights by confidence before drawing conclusions about all customers.