> For the complete documentation index, see [llms.txt](https://boundaryai.gitbook.io/boundaryai-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://boundaryai.gitbook.io/boundaryai-docs/bringing-in-your-feedback/surveys/managing-surveys.md). # Managing Surveys BAI Analytics provides a structured approach to managing both individual surveys and survey series, allowing you to organise, filter, and delete surveys efficiently. BAI Analytics organises every survey inside a **Feedback Group**: the central hub for a project, programme, or topic. Feedback Groups are what make long-term tracking possible: every survey, upload, connector, segmentation field, and analysis lives within a group, so you can compare cohorts, track trends across cycles, and keep historical data clean. This page covers how to navigate the management surfaces, close and reopen surveys, reach a known audience, and send invitations. For building a survey, see [Creating a Survey](/boundaryai-docs/bringing-in-your-feedback/surveys/editor.md). For look, language, and security settings, see [Customizing a Survey](/boundaryai-docs/bringing-in-your-feedback/surveys/images-and-media.md). *** ### How surveys are structured When you create a new survey, you have two options: * **Create the survey within a specific Feedback Group.** The survey is grouped with other surveys, uploads, and connectors for that project, enabling tracking, comparisons, and long-term trend analysis. * **Create it as an Individual Survey.** The survey is placed in the default **Individual Surveys** group, which acts as a catch-all for one-off or unfiled surveys. This group is shown in your interface language (for example *Sondages individuels* in French), cannot be renamed, and its name is reserved so no other group can use it. Every survey is therefore *always* part of a group, either a custom one you create or the default *Individual Surveys* group. {% hint style="success" %} **Tip.** Use a custom Feedback Group whenever you expect to run a survey more than once, compare results across audiences, or correlate survey data with uploaded datasets. Reserve *Individual Surveys* for true one-offs. {% endhint %} *** ### Navigating the Feedback Groups list The **Feedback Groups** page lists every group in your organisation. #### Search Use the search bar at the top of the list to filter groups by name. Search is live. Results update as you type. #### Sort The sort dropdown supports seven orderings, grouped by axis: * **By name**: Name A → Z, or Name Z → A. * **By date**: Newest first, or Oldest first. * **By data sources**: Most data sources, or Fewest data sources. * **Recent (24h)**: surfaces only Feedback Groups created in the last 24 hours, useful right after a workshop or onboarding session. The current sort is shown on the trigger button so you can see which ordering is in effect at a glance. #### Per-group actions From the menu on each group: * **Rename** the group. * **Delete** the group (and all its sources, if you confirm). Click anywhere on a group card to open it and view the sources inside. *** ### Navigating a Feedback Group's sources Each Feedback Group page lists its sources under tabs: **All**, **Surveys**, **File Uploads**, **Connectors**, **Web & Social**, and **AI Visibility**. Each survey shows a status chip: | Status | What it means | | ---------- | ------------------------------------------------------------- | | **Draft** | Not yet published. Not visible to respondents. | | **Live** | Published and actively collecting responses. | | **Closed** | No longer accepting responses; existing data stays available. | Click **Manage sources** to open the sources drawer, which separates **Active sources** (live surveys, connectors, and agents continuously feeding the group) from **One-time data** (file imports and closed surveys). Each list has its own search box, and this is also where you **Create survey**, upload a file, or add a connector. *** ### Survey lifecycle A survey moves through three states during its life: 1. **Draft**: created and editable. You can add or remove questions, change branding, set a passcode, and translate. Respondents cannot reach the survey yet. Draft surveys appear in the sources drawer as **Not yet published**. 2. **Live**: published. The survey is reachable via its share link, QR codes, or embed code, and responses flow into the Feedback Group's analytics. You can still edit copy and add questions while the survey is live; changes take effect on the respondent's next page load. 3. **Closed**: the share link shows a "survey closed" message; existing responses remain available for analysis. A survey closes when you close it by hand, when its **Close survey on** date arrives, or when it reaches the response count you set under **Closing Options** (see [Survey settings](/boundaryai-docs/bringing-in-your-feedback/surveys/editor.md#survey-settings)). Publish a draft from the builder with **Publish Survey**. Close a live survey with **Close survey** from its row menu, and bring a closed survey back with **Reopen**; it accepts responses again immediately, on the same link. Closed surveys are listed under **One-time data** in the sources drawer until reopened. *** ### Per-survey actions Each survey row in the sources drawer offers: * **Share**: copy the survey link, download a QR code, or **Open Share & Send** for invitations and tracked QR codes. * **Manage** / **Edit**: open the builder and adjust the survey. * **Rename**: change the title without affecting share links or response history. * **Close survey** *(Live only)* or **Reopen** *(Closed only)*. * **Download responses (Excel)**: export the raw responses. * **View analyses by period** *(groups tracked over time)*: jump to the survey's results for each period. * **Advanced options**: see below. * **Delete source**: permanently remove the survey, its responses, and its contribution to every analysis in the group. {% hint style="danger" %} **Heads up.** Deleting a survey also deletes its responses and any analyses derived from them, and cannot be undone. Where possible, prefer **Close survey** over **Delete source** so historical data stays available. If you only want to clear the answers while keeping the survey, questions, and links intact, use **Clear all responses** in **Settings → Danger zone** instead. {% endhint %} *** ### Advanced options Each source has an **Advanced options** panel (reachable from its row menu) with two things: * **Sources included**: whether this source's feedback contributes to the Feedback Group's grouped theme analysis. See [Choosing which sources feed your grouped themes](/boundaryai-docs/feedback-groups/feedback-groups.md#choosing-which-sources-feed-your-grouped-themes). * **API Endpoint** details, for sources fed programmatically: the endpoint and the identifiers you would use to push data into this source. {% hint style="info" %} **Looking for recurring surveys?** BAI Analytics does not clone a survey on a schedule. For an ongoing programme, track it over time with [Evolution](/boundaryai-docs/analysing-your-feedback/tracking-feedback-over-time-evolution.md), which buckets one continuously-collecting survey into weekly, monthly, or quarterly periods automatically. For a fresh cycle by hand, create a new survey in the same group. {% endhint %} *** ### Identified respondent lists For surveys you send to a known audience, you can upload a list of identified respondents so responses are tied to the right person and carry their metadata. Open the survey's **Settings**, go to the **Respondents** tab (also available as **Respondent list** in the builder's Settings panel), and upload a **CSV or Excel** file with an email column plus any metadata columns you want to segment by (department, office, role, and so on). Click **Upload and generate links**. * BAI Analytics auto-detects the email column. If it cannot tell which column holds the email address, it asks you to pick from the likely candidates. Rows without a valid email are skipped and reported. * Every respondent gets a **unique survey URL**. Their metadata is attached to their response automatically and becomes available for segmentation, personalised invitations, and metadata-based conditional questions. * If you plan to send SMS invitations or reminders, tag the **Phone column (for SMS reminders)** on the list. BAI Analytics suggests a phone column automatically when it can. * You can **Download the enriched file** back out as Excel or CSV, either **Full** (all columns plus each person's survey URL) or **Minimal** (email plus survey URL), to send through your own tools. * Deleting a list keeps the metadata already attached to existing responses; new responses simply stop carrying respondent metadata. *** ### Sending invites Once you have a respondent list, you can send invitations directly from BAI Analytics instead of exporting them to your own system. Open a survey's **Share** button, then **Open Share & Send**. Under **Send to people**, build a distribution workflow in three steps: 1. **Source**: upload or pick the **Contact list** to send to (CSV or Excel). Triggering sends from your own application through the API is also available where API access is enabled for your organisation. 2. **Template**: compose the subject and body with the email template editor (below). The survey link must be included somewhere in the message; BAI Analytics warns you if it is missing. 3. **Channel**: review the sender your emails go out from, then click **Send to N** to open the final confirmation. In the confirmation dialog you choose how the campaign runs. Nothing is sent until you confirm here. * **Send a test to yourself first**: enter your address and click **Test** to receive the exact email a recipient would get. * **Channel**: **Email**, **SMS**, or **Email + SMS**. SMS requires a phone column tagged on the list; Email requires a subject or body in the template. Every SMS carries a stop instruction so recipients can opt out. * **When to send**: **Send now**, or **Schedule** a date and time. * **Remind people who have not responded**: tick the box, then set the cadence, for example *Every 3 day(s), up to 1 time(s)*. Reminders go every 1 to 60 days, up to 10 times, and only reach people on the list who have not yet submitted a response, on the channel(s) you chose. Scheduled campaigns and reminder campaigns appear as **Campaign scheduled** on the Share & Send page, where you can **Cancel scheduled campaign** at any time before it runs. A list can have one live campaign at a time. #### Sender address and volume Before your first send, set up how outgoing mail identifies itself in **Settings → Email sending**: * **Sending from BAI Analytics (managed)** is the default: invites go out from a BAI Analytics address, shown as your organisation, with **10,000 included sends per month**. Going past the included volume never blocks a send; usage above it is billed and the BAI Analytics team follows up. A separate hard monthly limit protects against abuse; if a campaign would exceed it, the send is blocked and you can contact support to raise it. * **Use your own domain**: add your sending domain, from-address, and from-name, add the CNAME records BAI Analytics gives you to your DNS, then click **Verify domain**. Sends then go out branded as you, with no monthly cap. * **Your AWS account**: for high-volume programmes, connect your own Amazon SES account by deploying the provided template, pasting the resulting Role ARN, testing the connection, and verifying your sending domain. Volume is billed by AWS directly to you. {% hint style="info" %} **Managed-sending disclaimer.** Invites sent from the managed BAI Analytics address carry a small footer note explaining that the message was sent on behalf of the organisation that created the survey. Sending from your own verified domain or your own AWS account never shows this note. {% endhint %} #### Email template editor The built-in **Email template** editor lets you compose the invitation without leaving BAI Analytics. * **Subject**, **Reply-To** (optional), and a rich-text body with headings, lists, and links. * **Personalisation tokens**: click-to-insert chips for every column on the respondent list (for example first name or account), plus reserved **Email** and **Survey link** tokens. Each token is filled in per recipient, so everyone gets their own personalised URL. * **Conditional content**: insert blocks that branch on a metadata column, mapping each value to its own snippet with a fallback for missing or unmatched values. * **Preview**: see the message as a real recipient (or a sample one) would receive it, with a note on any tokens that have no value. * **Export emails**: if you would rather send through your own system, export the generated, personalised emails as a **ZIP of .eml files** or a **CSV** ready for your email tool. *** ### Tracked QR codes For in-person collection (events, stores, branches), Share & Send can generate a separate QR code per location, table, or campaign. Under **Share a link**, open **Survey link & QR**, then **Track where scans come from**: 1. Define what differs between your QR codes (for example *Zone* and *Booth*). 2. Add one QR code per spot and download it. Every scan lands on the same survey, and the answers are tagged with the QR code's details automatically. Respondents never see the tags. Archived QR codes keep working but stop adding tags. QR codes only work once the survey is live. *** ### Best practices * **Use Feedback Groups to track long-term trends.** Avoid leaving surveys in *Individual Surveys* if you expect to re-run them, compare cycles, or correlate them with other data. * **Let the survey close itself.** For a fixed-quota study, set **Automatically close survey after response count reaches** so you never over-collect; for a campaign window, set a close date. * **Test before you send.** Always send a test to yourself from the confirmation dialog, and check the preview against a real recipient so tokens resolve as expected. * **Close, don't delete, finished surveys.** Closing keeps the historical responses available for analysis; deleting removes them permanently. * **Use Evolution for ongoing programmes.** A weekly NPS or monthly satisfaction pulse tracks more cleanly as one survey with Evolution turned on than as separate surveys you manage by hand. * **Tidy up periodically.** Delete genuine throwaways and rename Feedback Groups whose scope has drifted, so the list stays useful for everyone in the organisation. By structuring your work around Feedback Groups and using the lifecycle states deliberately, you keep BAI Analytics organised, efficient, and data-driven, even as your survey programme scales.