Claude Research Skill

These are the Claude prompts and process that we use to reduce analysis, synthesis, reporting, and other time down from days or weeks to often a day or less.

This skill doesn’t replace Researchers. It speeds up the work that AI does decently so that you can deliver more value faster. It doesn’t teach companies that they didn’t need you; it shows why the foundation of everything must be Excellent Research.

The skill acts as a critical thinking partner, doing work with you, and not for you. It guides you through:

• Research preparation and data quality assessment.
• Behavioral segmentation. Optional refining or redefining of segments, personas, ICPs, customer profiles, or behavioral typologies.
• Problem statements for both explicit and implicit problems from your data. The skill will also pull participant quotes for evidence.
• Root cause analysis using Five Whys and service design lenses.
• Strategy development at the project/problem-solving level.
• Possible solution directions with built-in guardrails.

The problem statements and strategies are based on the templates from our Atomic Product-Market Fit book (2026).

Install directly into Claude (Anthropic) as a native skill for the smoothest experience, or copy the skill files and paste them manually into any AI assistant or large language model you already use, including ChatGPT (OpenAI), Gemini (Google), Copilot (Microsoft), Perplexity, and more. The skill and its sub-skills are all in plain markdown text (.md files). Use with any LLM you prefer.

Based on experience, refined over time.

We’ve been doing UX, CX, and other types of research for decades. We’ve been helping companies with AI, ML, automated, and algorithm-based solutions and transformations for decades.

We’ve been using Claude heavily in our work since mid-2024. You get the benefits of that experience.