Survey authoring

Surveys are the Retain module's way of asking members questions at scale — exit interviews, retention check-ins, satisfaction polls. National admins draft a survey, deploy it to the organization, and watch responses roll in on the same page.

This guide covers who can author surveys, the four question types, the lifecycle from draft to closed, how target-role filtering works, and why Retain responses are attributed by design (so you don't accidentally promise anonymity you can't deliver).

Who can use this

Only national / organization admins can create, edit, deploy, or delete surveys. Chapter officers cannot author surveys — on the chapter side they only see the retention alerts that come from snapshots. Members are the audience: they fill surveys out, they don't build them.

Getting to surveys

From the org sidebar, open Retain → Surveys. You'll see a list of every survey ever created for this organization, with a status badge on each one (Draft, Active, or Closed). New surveys start from the Create Survey button.

Surveys list with status badges and the Create Survey button.

Author a new survey

The survey builder has two sections: top-level details, and a list of questions.

Survey details

• Title — required. Shows up in member notifications and on the response page.
• Description — optional. A short blurb explaining why you're sending this survey. Members see it before they start answering.
• Target role — optional. Filters who can see the survey when it's deployed. Choices are: All members, Officers, Presidents, Members. If you leave this empty, every member of every chapter gets the survey.
• Close date — optional. After this date, the survey stops accepting responses. Leave it blank and the survey stays open until you manually close it.

Question types

Add as many questions as you need. Each question has a type that determines how members answer and how results aggregate.

Type	What members see	What you see in results
Text	A free-form text field	All responses listed verbatim
Rating (1-5)	A 1-to-5 numeric scale	Average rating + response count
Multiple choice	A radio-button list of choices you defined	Count per choice
Yes / No	Two buttons	Yes count + No count

For multiple choice, you type the options one per line. The order you enter them is the order members see them.

Each question can be marked Required (members can't submit without answering) or optional. Drag-handle reordering exists in the UI; the order you arrange questions in is the order members see them.

the survey builder with two questions added showing the type dropdown

Deploy a survey

A survey stays in Draft until you explicitly deploy it. While in draft, you can edit anything — title, target role, questions, close date — without consequence.

When you're ready, open the survey and click Deploy. Two things happen:

1. Status flips to Active. The survey is now visible to eligible members.
2. Notifications fire. Every member who matches the target role gets an in-app notification ("New survey: {title}") and an email with the survey title and description.

The notification goes only to active members of chapters in your org. If you set a target role, members whose membership role doesn't match are skipped. Members are not notified about surveys they're not eligible for.

Validation before deploy

The platform refuses to deploy a survey with zero questions. Add at least one before clicking Deploy. You also can't re-deploy a survey that's already Active or Closed — it's a one-way transition from Draft.

If you didn't set a Start Date, the deploy timestamp becomes the start date automatically.

Watch responses come in

Once a survey is Active, the survey detail page shows live response counts and aggregated results.

• Rating questions show the average across all complete responses.
• Multiple choice shows a count per choice.
• Yes / No shows the Yes count and No count.
• Text questions show every response verbatim in a scrollable list.

Only complete responses are counted in aggregates. Partial responses (currently impossible to create from the member side — members submit answers in one shot — but the schema allows for it) are excluded.

Each member can only respond once per survey. The platform enforces this on submit, returning an error if they try again.

Close a survey

A survey closes one of two ways:

1. Close date passes. Members who try to submit after the close timestamp get an error. The survey's status doesn't auto-flip to Closed — it stays Active in the list but the response endpoint rejects new submissions.
2. You delete the survey. From the survey detail page, click Delete. This removes the survey and all responses permanently. There is no recovery — if you need the data, export results before deleting.

There's no explicit "Close now" button at the moment. If you want to stop responses early without deleting the survey, the workaround is to edit the close date to a past timestamp.

About attribution

The Retain module is designed for attributed member feedback. Every response is recorded against the user who submitted it — the respondent field on a survey response is non-nullable. This is intentional, not an oversight: officer follow-up, individual coaching, attendance-tied check-ins, and longitudinal retention scoring all depend on knowing who said what.

What you see in the UI:

• The survey detail page aggregates results at the organization level — averages, choice counts, and verbatim text — without showing names on the dashboard.
• Individual response rows include the respondent's identity and are visible to org admins and to platform admins.

What this means for your survey copy:

• Do not tell members a survey is "anonymous." It isn't, by design.
• Honest framings that match the schema: "your answers are reviewed only in aggregate on the dashboard," "your response is visible to your org admin," or simply "we want to know how you're feeling — please respond."
• For genuinely sensitive topics (mental health check-ins, anonymous tip lines, harassment reports) Retain is not the right tool. Use a third-party anonymous survey service (Typeform, Google Forms, SurveyMonkey) and link to it from a bulletin or share in a chapter meeting. The platform deliberately does not offer an "anonymous survey" mode — attribution is the contract.

Deploying to specific roles

Target role filtering is role-name match against active membership. The choices in the dropdown — officer, president, member — correspond to the role field on the membership record, not to broader categories.

A few practical notes:

• All members means everyone in the org with any platform-access membership status (new member, active, intent-to-renew). Alumni statuses are excluded from notifications because the survey is delivered via active-member notification pipelines.
• Officers means just the role officer — NOT presidents.
• Presidents means just the role president — NOT all officers.
• If you want both officers and presidents, you have to either deploy two surveys or use "All members" and accept the broader audience.

There is no chapter-level targeting. Surveys are organization-wide; you cannot scope them to a specific chapter or region. If you need chapter-specific feedback, you'll need to filter responses by chapter in your own analysis after they come in (each response carries a chapter ID).

Editing a deployed survey

Once a survey is Active, you cannot edit its questions, target role, or core configuration. Members are already responding to a specific version of the survey, and mutating it mid-flight would invalidate their responses.

If you find a typo or need to reword a question after deploy, the practical options are:

1. Delete the survey and start over. You'll lose existing responses.
2. Live with the typo. Note the correction in a follow-up bulletin or chapter announcement.

This is intentional. Survey integrity matters more than fixing a typo on the fly.

Related

• Retention snapshot configuration — how the membership-declining alerts are generated alongside surveys
• Module enablement — the Retain module must be on for an org before surveys appear
• Permissions matrix

Last verified against v0.62.1 (2026-05-11).