> ## Documentation Index
> Fetch the complete documentation index at: https://help.equip.co/llms.txt
> Use this file to discover all available pages before exploring further.

# Questions, Topics, and Scenarios

> The three content types you can mix in a conversational interview, and how each one shapes what the AI asks.

A conversational interview is built from a list of items in the **Topics & questions** section of the interview's settings page (under **Interview Configuration**). Each item is one of three content types: a **Question**, a **Topic**, or a **Scenario**. You can mix all three in a single interview.

<Tip>
  Use a **Question** when the exact wording matters, a **Topic** when you want the AI to probe an area in its own words, and a **Scenario** when the candidate needs reference material such as code or a case study to react to.
</Tip>

## Choosing a Content Type

The **Add item** menu on the Topics & questions section offers all three types. Here is how they differ:

| Content type | What the AI does                                                   | Follow-ups | Best for                                                                                    |
| ------------ | ------------------------------------------------------------------ | ---------- | ------------------------------------------------------------------------------------------- |
| **Question** | Asks your exact question, word for word                            | None       | Compliance questions, salary or notice-period checks, anything that must be asked precisely |
| **Topic**    | Generates its own questions around your theme, then probes deeper  | 0 to 5     | Exploring skills and experience with natural back-and-forth                                 |
| **Scenario** | Shows the candidate reference material in a panel and discusses it | 0 to 5     | Code reviews, case studies, diagrams, situational judgment                                  |

Every item, regardless of type, shares a few common controls: a type selector, a **Difficulty** level, reorder arrows, and a delete action. Item text is required and can be up to 500 characters.

<Note>
  When you [create a conversational interview](/create-conversational-interview), Equip's AI drafts an initial set of topics from your job description. You can then edit, reorder, delete, or add items of any type before inviting candidates.
</Note>

## Questions

A **Question** is the most predictable content type. The AI reads it to the candidate exactly as you wrote it and never asks follow-ups, so every candidate hears identical wording.

This makes Questions the right choice when consistency across candidates matters more than depth. Typical uses:

* Screening questions where legal or HR teams have approved specific wording
* Logistics checks such as availability, relocation, or work authorization
* A fixed closing question you want every candidate to answer

<Note>
  Because follow-ups are always zero for Questions, each one adds exactly 1 to your interview's question count. Topics and Scenarios can add more, as explained under [How Items Count Toward Limits](#how-items-count-toward-limits).
</Note>

## Topics and Follow-Ups

A **Topic** is a theme rather than a script. You describe the area to evaluate, and the AI generates its own questions around it during the live conversation, adapting to what the candidate says.

Each Topic has a **Follow-ups** setting from 0 to 5. Follow-ups are the extra probing questions the AI may ask after the candidate's initial answer, which is what makes the interview feel like a real conversation rather than a survey.

<Steps>
  <Step title="Add a Topic">
    In the Topics & questions section, click **Add item** and choose the Topic type.
  </Step>

  <Step title="Describe the theme">
    Write what the AI should explore, for example a skill, a project type, or a behavioral trait. Keep it within 500 characters.
  </Step>

  <Step title="Set follow-ups and difficulty">
    Choose 0 to 5 follow-ups and pick a difficulty level. Higher follow-up counts consume more of your total question limit.
  </Step>
</Steps>

<Tip>
  You can also steer how the AI handles every item at once with the **Instructions for AI Interviewer** field (up to 4,000 characters). For example: "Start with easier questions, and then make them harder."
</Tip>

## Scenarios

A **Scenario** presents reference material that the candidate can see in a side panel while they discuss it with the AI. Scenarios appear to the candidate numbered in order as Q1, Q2, and so on.

A Scenario has two parts:

| Field                         | Shown to candidate | Purpose                                                                                      |
| ----------------------------- | ------------------ | -------------------------------------------------------------------------------------------- |
| **Candidate view** (required) | Yes                | The material itself, written in Markdown. Code blocks, image URLs, and tables are supported. |
| **AI context** (optional)     | No                 | Extra guidance for the AI. If left empty, the AI reads the candidate view as-is.             |

Use **AI context** to describe anything the AI cannot parse from the candidate view, such as what an image or diagram shows. The Scenario's title is taken automatically from the first Markdown heading or first line of the candidate view.

Like Topics, Scenarios support 0 to 5 follow-ups, so the AI can dig into the candidate's reasoning about the material.

<Note>
  Scenarios are rolling out gradually and may not yet appear in your Add item menu. Scenarios pair well with the [Interview Workspace](/interview-workspace), where candidates can write code or text answers instead of speaking them.
</Note>

## Difficulty Levels

Every item carries a **Difficulty** setting of Medium, Hard, or Expert. Difficulty determines how much the item weighs in the candidate's score:

| Difficulty | Scoring weight |
| ---------- | -------------- |
| **Medium** | 10 points      |
| **Hard**   | 20 points      |
| **Expert** | 30 points      |

An Expert item therefore influences the overall result three times as much as a Medium item. Topics generated by Equip's AI from your job description arrive with a difficulty already assigned, which you can change at any time.

<Tip>
  Mix difficulties deliberately. A few Medium items early in the list let candidates settle in, while Hard and Expert items later separate strong candidates from average ones.
</Tip>

## How Items Count Toward Limits

Each item counts as **1 topic**, and the total question count is the number of items plus all their follow-ups. Your interview type sets the limits:

| Interview type              | Topics  | Total questions |
| --------------------------- | ------- | --------------- |
| **Quick Interview**         | 3 to 5  | 5 to 10         |
| **Comprehensive Interview** | 5 to 10 | 10 to 25        |

A footer counter shows your running total, for example "4 topics, 9 questions added". When you hit a limit, the **Add item** button is disabled and the follow-up selector blocks values that would exceed the cap. Remove follow-ups or items to free up room.

<Warning>
  Every interview must contain at least one item. You cannot delete the last remaining item, and deleting any item is permanent once confirmed.
</Warning>

## Related Resources

* [Create a Conversational Interview](/create-conversational-interview) - Generate your first draft of topics from a job description
* [Interview Workspace](/interview-workspace) - Let candidates answer in code or written text instead of speaking
* [Interview Settings](/interview-settings) - Everything else you can configure before inviting candidates
* [AI Interview Reports](/interview-reports) - How item difficulty and answers turn into a candidate report