Editing the Brief Manually

While the AI Consultant is excellent at drafting and iterating on your research brief through conversation, sometimes you know exactly what change you want to make and you'd rather just do it yourself. Koji's brief editor lets you directly modify every part of your study design through a structured interface.

When to Edit Manually vs. Chat

Both approaches update the same research brief, so choose whichever feels faster for the change you're making:

The general rule: if you know the exact words you want, edit manually. If you want to think through implications or get suggestions, chat with the Consultant.

Accessing the Brief Editor

The research brief is displayed in the artifact panel alongside your conversation with the AI Consultant. The editor is organized into four tabs:

• Problem — edit the problem statement, decision to inform, hypothesis, success criteria, and problem cost
• Participant — define required experience, behavior of interest, relationship to problem, and screening question
• Approach — select from available methodologies and configure the interview mode (structured, exploratory, or hybrid)
• Questions — add, edit, configure, and reorder your interview questions using the structured question editor

To edit any section, click the corresponding tab and modify the fields directly. Your changes are saved automatically.

Editing the Problem Context

The Problem tab contains the fields that anchor your entire study. Changes here can have cascading effects.

Tips for editing the problem context:

• Keep the problem statement to one to three sentences
• Focus on what you want to learn, not what you plan to build
• Use the hypothesis field to state what you currently believe — the study will test it
• The decision to inform field should name a specific business or product decision
• After editing, scan the rest of the brief to make sure everything still aligns

If you significantly change the problem statement, consider asking the AI Consultant to review the rest of the brief for consistency. You can say something like: "I updated the problem statement — does the rest of the brief still make sense?"

Changing the Methodology

You can switch the methodology from the Approach tab. Select from the available methodologies — each is described in detail in our methodology guide.

Important: Changing the methodology affects how the AI interviewer conducts conversations. Different methodologies use different questioning styles, follow-up strategies, and conversational approaches. If you switch methodologies, review your interview questions to make sure they still fit the new framework.

For significant methodology changes, chatting with the AI Consultant is often more efficient, since it can regenerate appropriate questions for the new approach.

Working with the Structured Question Editor

The Questions tab is where most manual editing happens. Unlike a plain text editor, the question editor provides a structured interface for each question:

Adding Questions

Click the "Add Question" button to create a new question. You'll set:

1. Question text — the actual question the AI will ask
2. Question type — select from the dropdown: open-ended, scale, single choice, multiple choice, ranking, or yes/no
3. Type-specific configuration — depending on the type you chose:
   • Scale: set min/max values and endpoint labels (e.g., "Very Unlikely" to "Very Likely")
   • Single/Multiple Choice: add your option list, optionally enable "Allow Other" for free-text responses
   • Ranking: add items participants will order by preference
4. Probing configuration — set the maximum follow-up depth (0–3) and add specific probing instructions

For a full guide to question types and their configurations, see Structured Questions in AI Interviews.

Editing Existing Questions

Click on any question to expand its editing panel. You can modify the text, change the type, adjust configuration, or update probing settings. Changing a question's type will reset its type-specific configuration, so set the type first before configuring options.

Rewording Questions

Small wording changes can make a big difference in how participants respond.

Before: "What challenges do you face with the product?" After: "Walk me through a recent situation where you ran into difficulty while using the product."

The second version is more specific, grounded in real experience, and harder to answer with a vague generalization.

Reordering Questions

Drag and drop questions into the sequence that makes sense for the conversation flow. A well-ordered interview:

1. Starts with easy, broad questions that get the participant talking comfortably
2. Moves into the core topic once rapport is established
3. Goes deeper with specific, potentially sensitive questions in the middle
4. Places structured questions (scales, choices) after open-ended discovery, once the participant has context
5. Ends with reflective or forward-looking questions that leave the participant feeling valued

After reordering, read through the whole plan to check that transitions between questions still feel natural.

Configuring Probing

Each question has a probing configuration that controls how the AI follows up:

• Max follow-ups (0–3) — how many probing questions the AI can ask after the initial answer. Set to 0 for no probing, or up to 3 for deep exploration.
• Probing instructions — specific guidance for the AI, like "If the participant gives a low score, ask what would need to change."
• Anchor probing (scale questions only) — after a rating, the AI asks something like "You said 7 — what would need to change to make it a 9?" This is powerful for understanding the gap between current and ideal experience.

Adding or Removing Questions

Things to consider when adding questions:

• Will there be enough time to cover this in the interview?
• Does it contribute directly to answering your research question?
• Is it distinct from other questions, or does it overlap?
• Would a structured type (scale, choice) or open-ended approach work better here?

Things to consider when removing questions:

• Is the question truly unnecessary, or is it just phrased poorly? Sometimes rewording is better than removing.
• Does removing it leave a gap in the conversation flow?

Adjusting the Target Participant

The Participant tab describes who your study is designed for. When editing this section:

• Focus on behavioral criteria — what participants have experienced or done — rather than demographic categories
• Required Experience should describe specific past actions (e.g., "has evaluated at least two CRM tools in the past year")
• Behavior of Interest should name the specific behavior you're studying
• Screening Question helps verify participant fit at the start of the interview
• Consider whether your participant description is specific enough to recruit against

After Editing: A Quick Review Checklist

After any manual edits, run through this checklist before publishing your study:

• Does the problem statement still match the rest of the brief?
• Are all questions open-ended and non-leading (unless intentionally structured)?
• Do questions flow naturally from one to the next?
• Is the total number of questions realistic for your target interview length?
• Do probes add depth rather than just asking for "more"?
• Are structured questions configured correctly (scales have labels, choices have options)?
• Is the target participant specific enough to recruit?
• Does the methodology match the conversational style of your questions?

Switching Between Manual Editing and Chat

Remember, you can freely switch between editing manually and chatting with the AI Consultant. They operate on the same document. A common workflow is:

1. Use the AI Consultant for the initial draft and major structural decisions
2. Switch to manual editing for precise wording adjustments and question configuration
3. Return to the Consultant if you're unsure about a change and want a second opinion

For a full walkthrough of what each section contains, see Understanding the Research Brief.

Related Articles

• Understanding the Research Brief — what each section of the brief contains
• Structured Questions in AI Interviews — the full guide to question types, configuration, and probing
• Working with the AI Consultant — collaborating with the AI on study design
• Publishing Your Study — going live with your study
• Choosing a Methodology — understanding available research frameworks

Further reading on the blog

• Customer Interview Questions: 50+ Templates for Discovery, Churn, and Win/Loss (2026) — The template is not the bottleneck — conducting the interview at scale is. Here are 50+ customer interview questions organized by use case,
• How to Run AI-Powered Customer Interviews at Scale — Learn how to conduct effective customer interviews at scale using AI. This comprehensive guide covers everything from planning and question
• The Mom Test for Customer Interviews: How to Ask Questions That Get Real Answers (2026) — The Mom Test is the founder's guide to customer interviews that actually work. This complete guide covers Rob Fitzpatrick's three rules, rea

<!-- further-reading:blog -->