Creating Surveys

Navigate to Surveys in the sidebar and click the Create Survey button in the top-right corner. A dialog opens with two fields:

Enter a descriptive name that helps your team identify this survey. The name is internal only and not visible to customers. Good naming conventions include:

• By purpose: "Post-Service NPS Q1 2026", "Checkout Experience CSAT"
• By audience: "New Customer Onboarding", "Annual Review - Enterprise Accounts"
• By campaign: "Spring Promo Follow-up", "Website Redesign Feedback"

The name field is required and has a 255-character limit.

Select one of four survey types. This choice determines the primary scoring question that appears first in your survey:

Choosing the right type matters because it determines how demeterrr calculates scores, categorizes respondents (promoters, passives, detractors for NPS), and triggers automations like callbacks for detractors. See Understanding Survey Types for detailed guidance on when to use each type.

Click Create Survey to proceed. The survey is created in Draft status and you're redirected to the survey editor.

The survey editor opens to your new survey with the primary scoring question already in place (except for Custom type surveys, which start empty).

Click Add Question to add follow-up questions after the primary scoring question. demeterrr supports several question types:

• Rating - Numeric scale (stars, numbers, or emoji)
• Multiple Choice - Single or multi-select options
• Open Text - Free-form text area for detailed feedback
• Yes/No - Binary choice questions
• Dropdown - Select from a list of options

Each question has settings for:

• Question text in English and French (for bilingual surveys)
• Required vs. optional - Whether respondents must answer before proceeding
• Conditional logic - Show/hide the question based on previous answers

Drag and drop questions to reorder them. The sequence customers see matches the order in your editor. General best practices for ordering:

1. Start with the score question (NPS/CSAT/CES) - this is your primary metric
2. Follow with a text question asking why they gave that score - this captures the "why" behind the number
3. Add specific topic questions in the middle - service quality, product satisfaction, etc.
4. End with optional demographic or open questions - lower-priority items go last since drop-off increases with survey length

Keep surveys as short as possible. Every additional question reduces your completion rate. For most use cases:

• NPS/CSAT surveys: 2-4 questions total (score + follow-up text + 1-2 specifics)
• CES surveys: 2-3 questions (effort score + what was difficult + suggestions)
• Custom surveys: 5-8 questions maximum for general feedback collection

If you need more than 8 questions, consider splitting into multiple targeted surveys rather than one long questionnaire.

Click the Settings tab in the survey editor to configure how your survey behaves. Key settings include:

Customize the survey's visual appearance to match your brand. You can set colors, add your logo, and adjust the layout. See Branding for details.

Toggle whether responses are linked to the contact who received the survey or collected anonymously. Anonymous surveys may get more honest feedback but prevent you from following up with specific customers.

Enable French/English bilingual mode so respondents see the survey in their preferred language. When enabled, you'll need to provide French translations for all question text, option labels, and any custom content. See Bilingual Surveys for the complete setup guide.

Set up branching rules so certain questions appear only when specific conditions are met. For example, show a detailed complaint form only when the NPS score is 0-6, or display a referral question only for promoters (9-10). See Conditional Logic for advanced configuration.

Before activating your survey, preview it to see exactly what your customers will see.

Use the survey preview to walk through the entire experience as a respondent would. Check that:

• Questions appear in the correct order
• Conditional logic triggers properly (test with different score values)
• Bilingual content displays correctly in both languages
• The survey looks good on both desktop and mobile layouts
• Required question validation works as expected

Send a test survey to yourself or a team member before launching to the full audience. This verifies the complete experience including email delivery, the survey link, and response submission. Go to Sendings and use the test send feature to send to a single email address.

Change the survey status from Draft to Active when you're ready to start collecting responses. Only active surveys can be sent to contacts. You can return a survey to Draft status to make changes, but this pauses any active sendings.

Once active, there are several ways to distribute your survey:

• Manual sending from the Sendings page - select contacts and send immediately
• Bulk sending with CSV upload - send to a large list at once
• Sequences - automate multi-step survey campaigns with delays and reminders
• API - trigger survey sends programmatically from your own systems

See Sending Surveys for detailed instructions on each method.

Your survey moves through three statuses during its lifecycle:

Before activating a new survey, verify:

• All questions have clear, unambiguous wording
• Required questions are marked correctly (don't require optional demographic questions)
• Conditional logic branches work for all score ranges
• French translations are complete and natural (if bilingual)
• The survey renders well on mobile devices
• A test send has been completed and reviewed
• You've identified your target audience and sending method

Too many questions. A 15-question survey will have 3-4x the drop-off rate of a 3-question survey. Every question should earn its place.

Vague question text. "How was your experience?" is less actionable than "How satisfied were you with the speed of service today?" Specific questions produce specific, usable feedback.

Missing conditional logic. Asking a detractor "Would you refer us to a friend?" immediately after they said they wouldn't recommend you creates a frustrating experience. Use branching to show relevant follow-up questions.

Skipping the test send. What looks correct in the editor may not look right in an email inbox. Always send a test to verify the complete flow.

For a guided walkthrough of creating your very first survey, see Creating Your First Survey.

To learn about all available question types and their options, see Question Types.

For detailed survey configuration options, see Survey Settings.