{"site":{"name":"Koji","description":"AI-native customer research platform that helps teams conduct, analyze, and synthesize customer interviews at scale.","url":"https://www.koji.so","contentTypes":["blog","documentation"],"lastUpdated":"2026-05-18T13:34:18.954Z"},"content":[{"type":"documentation","id":"4fa534e5-6c26-4da4-b3fb-dffdadd31ddc","slug":"mcp-workflow-researchers","title":"MCP Workflow Guide for UX Researchers","url":"https://www.koji.so/docs/mcp-workflow-researchers","summary":"Workflow guide for UX researchers using Koji MCP with Claude. Covers designing studies with structured question types (scales, choices, rankings), transcript analysis with per-question structured answers, cross-study quantitative comparison, multi-study management, report generation with charts, data export for research repositories, and methodology-specific tips including best practices for structured question design.","content":"## Why Researchers Love MCP\n\nUX researchers already know how to do research. The bottleneck is not skill — it is **operational overhead**. Scheduling, transcribing, coding, synthesizing, reporting. Koji MCP eliminates the overhead so you can focus on what matters: understanding users.\n\nHere is what changes with MCP:\n- **Study setup**: Minutes instead of hours (Claude generates methodology-aligned briefs with structured questions)\n- **Transcript review**: Ask Claude to summarize and highlight key moments, including structured answer data\n- **Cross-study analysis**: Compare themes and quantitative metrics across projects in one conversation\n- **Reporting**: Generate and publish stakeholder reports conversationally\n- **Data export**: Structured JSON for your insight repository\n\n---\n\n## Multi-Study Management\n\nResearchers typically run 3-5 studies simultaneously. MCP makes this manageable:\n\n> \"List all my active studies with their interview counts\"\n\nGet a quick dashboard of everything in flight. Then drill into any study:\n\n> \"Show me the latest interviews from my onboarding study — any negative sentiment?\"\n\n### Comparing Across Studies\n\nThis is where MCP shines for researchers. You can ask Claude to synthesize across your entire research portfolio:\n\n> \"Get the study data from my onboarding study and my retention study. Are there overlapping themes? What shows up in both?\"\n\nClaude pulls structured data from both studies and identifies patterns that cross project boundaries — something that is incredibly time-consuming to do manually.\n\n---\n\n## Designing Studies with Structured Questions\n\nResearchers can create methodologically rigorous studies with mixed question types through MCP:\n\n> \"Create a usability study with the discovery methodology. Include a System Usability Scale question (1-5 scale, labeled 'Strongly disagree' to 'Strongly agree'), a task completion satisfaction rating (1-7 Likert scale), a multiple-choice question about which features they used (with an 'Other' option), and 4 open-ended questions with probing depth of 2.\"\n\nClaude creates a study using `koji_create_study` with fully configured structured questions. Each question type captures data differently:\n\n- **Scale questions** generate quantitative benchmarks you can track over time (SUS scores, CSAT, NPS)\n- **Choice questions** segment respondents and reveal usage patterns\n- **Ranking questions** show relative priority across a set of options\n- **Open-ended questions** with probing depth capture the qualitative context behind the numbers\n\n### Adjusting Question Design\n\nAfter reviewing the initial brief:\n\n> \"Show me the structured questions in my study. Change the SUS question to a 1-7 scale and add anchor probing so the AI asks respondents to explain their rating.\"\n\nClaude uses `koji_update_brief` to modify the question configuration. The anchor probing feature is particularly useful — after a respondent gives a rating, the AI interviewer asks \"You said X — what would need to change for that to be higher?\" — capturing the reasoning behind quantitative scores.\n\n---\n\n## Transcript Analysis Workflow\n\n### Rapid Screening\n\nInstead of reading every transcript start to finish:\n\n> \"Show me interviews from my study that have themes related to 'pricing' or 'value'\"\n\nClaude filters by cached theme data and surfaces the relevant interviews.\n\n### Deep-Dive Reading with Structured Answers\n\nWhen you find an interview worth studying:\n\n> \"Show me the full transcript for this interview. What are the key moments where the respondent expressed strong emotion? What were their structured answers — NPS score, feature selections, and ranking?\"\n\nClaude reads the transcript and highlights pivotal moments while also surfacing the per-question structured answers with their qualitative context and confidence levels.\n\n### Coding and Theming\n\nWhile Koji auto-generates themes, researchers often need to recode:\n\n> \"Get the study data. What themes are you seeing that might not be captured in the automatic analysis? Look at the summaries and suggest additional themes.\"\n\nClaude can suggest themes based on its reading of the summaries — a starting point for your codebook.\n\n### Quantitative Analysis Across Interviews\n\nFor structured questions, you can get aggregate statistics:\n\n> \"What is the average SUS score across all completed interviews? Show me the distribution. Are there differences between respondents who rated task completion above 5 versus below 5?\"\n\nClaude pulls per-question aggregations from `koji_get_study_data` — averages, medians, distributions for scale questions, and frequency counts for choice questions.\n\n---\n\n## Research Report Workflow\n\n### Generate from Evidence\n\n> \"Generate a report from my usability study. I want it to include recommendations we can present to the design team.\"\n\nThe report includes citations linking every finding back to specific interviews, charts showing scale distributions and choice breakdowns, and recommendations grounded in both qualitative themes and quantitative data.\n\n### Section-by-Section Review\n\n> \"Show me just the executive summary and key takeaways from the report\"\n\nReview the high-level narrative first. Then:\n\n> \"Now show me the charts and question coverage\"\n\nWork through the report methodically without loading the entire document.\n\n### Publish for Stakeholders\n\n> \"Publish the report so I can share it with the product team\"\n\nGet a public link that non-Koji users can access — no account required for viewing.\n\n---\n\n## Building a Research Repository\n\n### Export for Your Insight Hub\n\nResearchers often maintain research repositories in Notion, Dovetail, or custom databases:\n\n> \"Export the full data from my study — brief, respondents, transcripts, and report summary\"\n\nGet structured JSON that maps to your repository schema. Transcripts are paginated (max 10 per request) to handle large studies:\n\n> \"Export transcripts 11-20 from my study\"\n\n### Cross-Referencing Over Time\n\nWhen a stakeholder asks \"what do we know about onboarding?\", you can pull insights from multiple studies:\n\n> \"List all my completed studies. Which ones have themes related to 'onboarding' or 'first-time experience'?\"\n\n> \"Get the study data from those three studies and synthesize the key findings about onboarding. Compare the NPS scores across studies.\"\n\nThis turns your research history into a searchable knowledge base through natural conversation, with both qualitative patterns and quantitative benchmarks.\n\n---\n\n## Research-Specific Tips\n\n### Methodology Matters for AI Interviewers\n\nThe methodology you choose directly shapes how the AI conducts interviews:\n\n- **Mom Test**: The AI avoids hypotheticals and focuses on past behavior. It will not ask \"would you use feature X?\" — instead it asks about current workflows and pain points.\n- **JTBD**: The AI explores the \"switching moment\" — what triggered the user to seek a solution and what alternatives they considered.\n- **Discovery**: Broad exploration with minimal assumptions. Good for new problem spaces.\n\nChoose your methodology based on your research question, and the AI interviewer follows the guardrails automatically.\n\n### Designing Effective Structured Questions\n\nBest practices for structured questions in research studies:\n\n- **Use established scales** — NPS (0-10), CSAT (1-5), SUS (1-5 or 1-7). Consistent scales enable cross-study benchmarking.\n- **Enable anchor probing on scale questions** — This captures the \"why\" behind the number, which is where the real insight lives.\n- **Keep choice options to 3-7** — Too many options create decision fatigue. Include \"Other\" with free text for unexpected answers.\n- **Set probing depth based on question importance** — Use 2-3 follow-ups for core research questions, 0-1 for screening or demographic questions.\n- **Group questions into sections** — Use the section parameter (e.g., \"Usability\", \"Satisfaction\", \"Demographics\") for organized data collection.\n\n### Quality Over Quantity\n\nDo not aim for 100 interviews just because you can. Qualitative research reaches saturation at 12-20 interviews for most topics. Monitor theme stability:\n\n> \"Show me the top themes from my study. Have they changed since interview 15?\"\n\nIf themes stop changing, you have likely reached saturation.\n\n### Voice vs. Text\n\nVoice interviews tend to produce:\n- Longer, more detailed responses\n- More emotional expression\n- More natural language patterns\n\nText interviews tend to produce:\n- More precise, considered responses\n- Easier to search and quote\n- Lower respondent friction\n\nMost researchers use voice for exploratory research and text for validation.\n\n---\n\n## Next Steps\n\n- **[Tool Reference](/docs/mcp-tool-reference)** — Detailed parameter docs for all 15 tools\n- **[Structured Questions Guide](/docs/structured-questions-guide)** — Deep dive into question types and probing\n- **[Best Practices](/docs/mcp-best-practices)** — General MCP usage tips\n- **[MCP Overview](/docs/mcp-overview)** — Full integration overview\n- **[How Many Interviews Are Enough?](/docs/how-many-interviews-enough)** — Sample size guidance\n\n## Further reading on the blog\n\n- [Why AI Interviewers Are the Future of Customer Research](/blog/why-ai-interviewers-are-the-future-of-customer-research) — AI interviewers are transforming how product teams conduct customer research, enabling conversations at scale without sacrificing depth or q\n\n<!-- further-reading:blog -->\n","category":"Claude & MCP Integration","lastModified":"2026-05-13T00:25:38.788654+00:00","metaTitle":"MCP Workflow for UX Researchers — Scale Qualitative Research with AI | Koji","metaDescription":"How UX researchers use Koji MCP with Claude to manage multiple studies, analyze transcripts, compare themes across projects, and generate stakeholder reports. Includes research-specific tips.","keywords":["UX researcher AI tools","qualitative research automation","research operations AI","UX research MCP","multi-study management AI","transcript analysis AI","research repository AI","scale qualitative research"],"aiSummary":"Workflow guide for UX researchers using Koji MCP with Claude. Covers designing studies with structured question types (scales, choices, rankings), transcript analysis with per-question structured answers, cross-study quantitative comparison, multi-study management, report generation with charts, data export for research repositories, and methodology-specific tips including best practices for structured question design.","aiPrerequisites":["Koji MCP connected to Claude","UX research experience"],"aiLearningOutcomes":["Manage multiple research studies via Claude","Analyze and compare themes across studies","Generate and publish stakeholder reports","Export structured data for research repositories","Choose between voice and text interviews based on research goals"],"aiDifficulty":"intermediate","aiEstimatedTime":"10 minutes"}],"pagination":{"total":1,"returned":1,"offset":0}}