How to Connect Koji User Research to Figma & FigJam

The Short Answer

Koji does not lock your research in a separate silo. You can bring AI interview insights — themes, verbatim quotes, and structured-question results — into Figma and FigJam four ways: export the report and drop quotes onto a FigJam board, pull data through the REST API, generate design-ready artifacts with Claude via Koji's MCP server, or wire an automation with Zapier/n8n. The result is a design surface where every screen is backed by a real customer quote instead of a guess.

The fastest path for most teams: run the study in Koji, let it auto-generate the research report, then export themes and quotes into a FigJam affinity board your whole team can design against.

---

Why Connect Research to Figma at All

Design decisions made without evidence are expensive to unwind. When the insight lives in a separate research tool, designers rarely open it — so it never shapes the work. Pulling Koji's output into the canvas where design actually happens closes that gap:

• Quote-backed designs — pin a real customer sentence next to the screen it justifies
• Affinity mapping in FigJam — cluster Koji's auto-extracted themes into opportunity areas
• Evidence in design reviews — stakeholders see the why beside the what
• Faster discovery-to-design handoff — no re-typing findings from a PDF

---

Method 1 — Export → FigJam (no setup)

The simplest workflow, and the one most teams start with:

1. Run your study in Koji and open the auto-generated report
2. Export the report and structured-question data (export options)
3. In FigJam, create sticky notes from Koji's coded themes and verbatim quotes
4. Group stickies into clusters — this is classic affinity mapping, now grounded in real interviews

Because Koji already codes open-ended answers into themes during analysis, you are pasting organized insight, not raw transcript. A study that produced 40 interviews lands as a tidy set of theme clusters with supporting quotes — minutes of work instead of hours of manual tagging.

---

Method 2 — REST API (live data)

If you want a repeatable pipeline, Koji's REST API exposes study data, transcripts, and report content programmatically. A small script can:

• Fetch the latest report for a study
• Pull the top themes and their supporting quotes
• Format them as text your team imports into FigJam, or push them to a design-ops dashboard

This is ideal for teams running continuous discovery who want the current "voice of the customer" always reflected in their design system documentation.

---

Method 3 — Claude MCP (AI-assisted)

Koji ships a Model Context Protocol (MCP) server so AI clients like Claude can read your studies directly. With MCP connected, you can ask Claude to:

• "Summarize the top 5 themes from my onboarding study as FigJam-ready sticky text"
• "List the three strongest quotes about pricing confusion"
• "Draft a problem statement for the checkout redesign based on this study"

Claude reads the live Koji data through MCP tools and hands you structured text you paste straight onto the canvas. It is the closest thing to a research analyst sitting next to your designer.

---

Method 4 — Zapier / n8n (automation)

For hands-off workflows, connect Koji to Zapier or n8n. Trigger on "study completed" or "report generated," then route the output wherever your design team works — a Slack channel, a Notion research database your FigJam links to, or a Google Doc the team pulls from. You set it once and new insights flow automatically.

---

Pairing Structured Questions with Design Decisions

Koji's six structured question types — open_ended, scale, single_choice, multiple_choice, ranking, and yes_no — give your Figma work quantitative backbone, not just quotes:

• A ranking question ("order these features by importance") tells you what to put first in the layout
• A scale question (satisfaction or NPS) gives you a baseline to redesign against
• A single_choice question ("which entry point did you expect?") settles navigation debates with data
• open_ended answers, auto-themed, supply the verbatim quotes you pin beside mockups

So a single Koji study feeds both the direction (structured results) and the empathy (quotes) of your design — no separate survey tool required.

---

A Practical Discovery-to-Design Loop

1. Define the design question (e.g., "why do users drop at step 3?")
2. Interview in Koji — voice or text, with AI follow-up probing
3. Auto-analyze — Koji codes themes and aggregates structured answers
4. Bring to FigJam — export or MCP the themes and quotes onto a board
5. Affinity map — cluster into opportunities
6. Design in Figma with each decision tied to a quote
7. Validate the new design, then re-interview to confirm it worked

This keeps the canvas honest: every pixel traces back to something a real customer said.

---

Tips for a Clean Workflow

• Keep one FigJam board per study so evidence stays traceable
• Color-code stickies by theme to make clusters obvious
• Always carry the verbatim quote next to your paraphrase — it is what convinces stakeholders
• For recurring research, use the API or Zapier route so the board updates itself

---

Related Resources

• Structured Questions Guide — the 6 question types that give your designs quantitative backbone
• Generating Research Reports — the report you export into FigJam
• Affinity Mapping — cluster Koji themes into opportunity areas
• Understanding Themes & Patterns — how Koji codes open-ended answers
• MCP Overview — connect Claude to your Koji studies
• User Interview Guide — run the interviews that feed your design