Conversation Settings

You can change most settings by asking the design agent. For example: "Make the welcome message shorter" or "Use our brand colors" or "Require participants to sign in with Google."

You can also access settings directly by switching from the Outline view to the Settings view in the sidebar. Settings are organized into five sections.

Title and Welcome Message#

The first thing participants see.

Conversation Title — The external title shown to participants. This is different from the internal name (see Advanced).

Welcome Message — The intro text participants see before the conversation starts. Explain what the conversation is about and roughly how long it takes. Up to 500 characters. You can reset to the auto-generated default at any time.

Branding#

Control how your conversation's identity appears to participants.

Interview Identity — Choose between two modes:

• Personal — Participants see your name, avatar, and an optional signature/byline (e.g., "Sarah, Product Manager at Acme"). Use this when the conversation should feel like it is coming from a real person.
• Brand — Participants see your company logo and name. Use this for product-level conversations where the individual behind it does not matter.

Custom Link — Set a clean, branded URL slug for your conversation link (lowercase, numbers, and hyphens only).

Custom Logo — Upload separate logos for light and dark mode. When enabled, your logo is used as the conversation logo. Use Remove Perspective Branding to hide "Powered by Perspective" labels.

Custom Theme Colors — Set four colors for both light and dark mode:

You can set these manually or let Perspective discover your brand colors automatically from your website.

Voice & Tone#

Voice for Conversations — Select the voice used in voice-enabled conversations. Preview each voice before choosing. The selected voice affects how the agent sounds when participants use voice mode.

Voice Instructions — Describe the voice personality, accent, dialect, or speaking style for voice conversations.

Conversations support both text and voice channels. You can enable or disable each channel independently (see Advanced).

Automations#

Control what happens automatically during and after conversations. Automations are organized into two tabs:

Per Interview — Triggered after terminal conversation events. Participant-facing email automations run only after completed, non-manual conversations:

• Generate interview summary — Always on. A summary is automatically created after each conversation.
• Generate trust assessment — Always on. Flags suspicious patterns and gives you a trust score to review.
• Thank-you email to participants — Optionally send an automatic thank-you email after completion. Recommended.
• Email summary when interviews complete — Get an email summary each time a participant finishes so you can stay on top of incoming responses.
• Custom automations — Add your own per-interview triggers for webhooks, Slack messages, CRM updates, or email sequences. See Automations for the design workflow and Automation Recipes for implementation patterns.

Scheduled — Recurring reports on a daily or weekly cadence:

• Set up digest reports that summarize new conversations over a time period.
• Useful for staying informed without checking the dashboard.

Advanced#

Internal Name — A name visible only to you and your team, separate from the title participants see. Use this to organize conversations internally (e.g., "Q1 Churn Research - Enterprise").

Require Sign-In — By default, conversations are anonymous. Toggle this on to require authentication. When enabled, you choose which sign-in providers to accept:

• Google, LinkedIn, Microsoft, Email

Voice and Text — Enable or disable voice and text channels independently. At least one must be enabled. When both are on, participants can switch between them during the conversation.

Participant Limit — Set a maximum number of participants. Once reached, the conversation link stops accepting new people. Leave blank for unlimited.

No-Activity Timeout — Minutes of inactivity before the conversation is automatically reviewed and wrapped up based on the participant's progress. Default: 30 minutes. Minimum: 10 minutes. Leave blank to use the default.

Allow Multiple Conversations — When on, each participant can complete the conversation multiple times. Useful for longitudinal studies or recurring feedback. When off, each participant can only complete it once.

Quick Reply Options — When on, participants can see tappable response options for eligible closed questions such as yes/no, ratings, and pick-from-list prompts. Helps participants respond faster, especially on mobile.

Remove Perspective Branding — Hide supported "Powered by Perspective" labels from participant-facing surfaces.