Distribution
Distribution
The distribution step has two independent choices: who receives the survey (Recipient type — a public link anyone with the URL can respond to, or targeted distribution to selected org-tree nodes: everyone, groups, managers, or a single department) and how it reaches them (Channel — see below). You also choose whether it sends immediately or on a schedule, once or recurring (daily/weekly/monthly), and how long it stays open for responses (1–30 days or always open). A custom message can be attached and saved to a reusable library.
Channel — Email, SMS, Share link, QR code, or Kiosk mode. Right below the Recipient-type choice is a row of five icon buttons: Email, SMS, Share link, QR code, and Kiosk mode. Email is selected by default and is the only channel that sends a real, personal link to each recipient today. Channel and Recipient type are independent — you can, for example, target a department AND still use a Share link/QR code/Kiosk mode as an extra, shared way into the same survey. Choosing SMS persists the choice, but the survey CANNOT be published/activated while SMS is the selected channel — a clear error message appears both in the distribution step and if you try clicking “Klar” (“Finish”) anyway, since real SMS sending isn’t wired in yet. Saving as a draft still works as normal regardless of channel.
Provisional groups — self-service filter builder. Under “Target people” → “Saved groups”, you can (if your organization has enabled it) click “Create group (filter)” instead of picking people one by one. Build any number of filters against any stafflist column (the same rule builder as Saved Views/Custom Filter), watch a running “X match” count as you type, and choose between “Everyone matching” (the group stays live — membership is recomputed automatically every time it’s used, so a new employee who satisfies the filter shows up on their own) and “Pick individually” (freezes today’s match into a fixed list — later staff changes don’t affect the group). Groups share the same “Saved groups” library regardless of how they were created.
Editable multilingual dispatch message. If your organization has enabled it, the plain info message is replaced by an editor with a Subject and a Body field per language (SV/EN/NO). Language tabs switch which variant you’re editing; a “Translate” button (shown for EN/NO) machine-translates the Swedish text into the chosen language automatically — the suggestion lands in an editable field, never saved directly. A recipient automatically sees the Swedish or English variant depending on their own language choice on the survey page (the same sv/en toggle the rest of the page uses) — Norwegian is saved and editable/translatable today, but isn’t shown separately yet (the survey page has no Norwegian mode yet) and falls back to Swedish.
Recurring pulses. When a recurring survey closes, the next round is created automatically: a new scheduled survey with the same questions and settings, its next send time computed from the interval. The rounds are linked as a series so trends can be followed over time — you don’t have to hand-create each week’s pulse.
Avoid public holidays for sending. For a recurring survey, an Admin can turn on “Avoid public holidays for sending” in the schedule step. If the next computed send time would land on a Swedish public holiday (e.g. Ascension Day or Christmas Day), it’s automatically shifted forward, one day at a time, until it lands on a non-holiday — the clock time is preserved. Off by default; if left off, a send can still land on a holiday exactly as before. The calendar covers Sweden’s 13 official public holidays (including Easter and Midsummer calculations); “eve” days like Christmas Eve and Midsummer’s Eve are not counted as holidays in this calculation.
Automatic repeating reminders. In addition to sending a reminder by hand (see below), an Admin can set “Send a reminder after X days” and, optionally, “Repeat every Y days” in the schedule step. When set, the system automatically reminds non-responders X days after the send, then again every Y days after that, until everyone has responded or the survey closes. An automatic reminder shares the same double-send guard as a manual one (at least a day — or the shorter gap between two automatic rounds — between any two reminders for the same survey, whether triggered by hand or automatically). Automatic reminders currently go out by email only (plus the usual count-only chat notification) — not via SMS.
Question rotation (fights survey fatigue). For a recurring survey, an Admin can turn on question rotation in the schedule step (requires question rotation to be enabled for your organization — off by default). You choose a set of “core” questions that are always asked every round (e.g. eNPS plus one or two anchor driver questions) and how many of the remaining questions should rotate each round. The next round’s rotating subset is selected deterministically from the round number — the same round always produces the same selection, and over enough rounds every question gets its turn. With rotation off (the default), every configured question is asked every round, exactly as before.
Survey-fatigue cap (shared across every survey type). A recurring pulse, a lifecycle survey, and an ad-hoc campaign can in theory all reach the same employee in the same week. When the “Survey-fatigue cap” feature is enabled for your organization (off by default), before a new invite is written by any of the three dispatch paths, the system checks whether that employee already received ANY survey invite (of any type) within a minimum number of days (an organization setting, default 3 days). If the gap hasn’t passed yet, the invite is deferred to a later run instead of going out immediately. The check is designed to never accidentally block all invites.
Non-responder reminders. You can send a reminder that targets only those who haven’t answered yet — already-submitted responses are never touched (their status isn’t reset). Any chat notification carries a count only, never who has or hasn’t responded, so anonymity is preserved. If your organization has enough responses so far (above the anonymity threshold), the chat message also shows “X of Y responses so far” — below the threshold, no count is ever shown.
SMS reminders (org-wide, separate from the Channel choice above). SMS distribution for REMINDERS can be enabled for the organization (off by default) — each employee’s unique link is then also sent via SMS whenever an admin manually sends a reminder, provided they have a phone number on file (not a requirement — if the number is missing, that employee is simply skipped for the SMS part; email/chat are unaffected). This is a completely separate mechanism from the “SMS” Channel option above: this setting governs REMINDERS for every survey in the organization regardless of which Channel was chosen, while the “SMS” Channel option is a per-survey choice for the INITIAL send that (see above) currently always blocks publishing — the two SMS concepts should not be confused. One SMS failure never stops the reminder to the rest of the batch, and there’s a daily cap on how many SMS an organization can send. The reminder text used for SMS can be customized by an Admin under Settings → Survey Settings → Reminder message (placeholders {{surveyTitle}} and {{link}} are filled in automatically; the text is always shortened as needed since SMS must stay short). The SMS part of a reminder is only sent when an admin manually sends the reminder for an open survey campaign — an automatic/scheduled reminder (see above) currently goes out by email only, not SMS.
Important: chat vs. SMS/email. A Slack/Teams webhook can only post to a WHOLE channel — it can never send a private link directly to one specific person. The chat message is therefore always a shared summary (title, recipient count, optional response count), not the survey’s unique link. If you want each person to receive their own link, use email and/or SMS.
QR code. Choosing the “QR code” Channel shows a “Show QR code” button in the distribution step. It opens a QR code generated right in the browser (no server round-trip) — handy for a poster on a notice board, in a break room, or by a kiosk device so employees can scan straight into the survey from their phone. The panel also shows the link as text for copying, plus a button to download the QR code as a PNG image. QR code shows regardless of which Recipient type you chose (public link or targeted) — see above on how Channel and Recipient type interact.
Kiosk mode. Choosing the “Kiosk mode” Channel shows its own section right in the distribution step — no detour to a separate settings page. Click “Create kiosk device” to create a dedicated kiosk link (separate from the regular public link), which can then be shown as a QR code, opened directly, or copied. A kiosk device is meant for a shared tablet/screen (e.g. in a break room) where several people respond one after another on the same device. “Remove kiosk mode” disables the device’s link/QR code immediately (with a confirmation dialog first). Like QR code, Kiosk mode shows regardless of the chosen Recipient type.
Every recipient gets their own link by email. Both the initial send and reminders deliver a real email to each recipient with their own personal response link — it shouldn’t be forwarded. The email shows your organization’s name (and logo, if you’ve uploaded one) and defaults to Swedish. If you’ve set a reply-to address under Settings → Company logo, replies go there instead of to Novro. If a single email fails to send (e.g. a bad address), it never blocks the send to everyone else — the summary after sending shows how many were sent versus failed.
Large sends go out in the background. For a small or medium organization the emails are sent immediately while you wait, exactly as before. For a large send (more than 100 recipients) the links are created straight away but the emails themselves are placed in a delivery queue and sent in the background by a scheduled job, in batches — so a very large send can never make the action time out partway through. You’ll get a confirmation that the emails are queued; delivery may take a little while to finish. If individual addresses fail, that’s recorded per recipient in the queue, so a partial delivery stays visible afterwards instead of being hidden behind an error message.
Test send (Testutskick). The editor header’s “Testutskick” button (requires the feature to be enabled for your organization — off by default) sends a real, fully-rendered copy of the saved draft to yourself, or a small list of email addresses you type in — exactly as it will look live. A test send’s response is NEVER counted in the survey’s real results/statistics — it’s excluded from every report and the anonymity-threshold-gated aggregate. Testutskick is for checking how the survey looks, not a real pilot group.
Save to library (personal template). The “Spara i bibliotek” button in the editor header (optional feature, off by default) saves the current question(s) as a PERSONAL template — visible only to you, unlike the shared/curated template library every admin can browse. Your personal templates show in their own “Personal” section under the Question Library, and only you can delete them.
Excel export/import of questions (round-trippable). The Content step can export the survey’s questions to an .xlsx template and later re-import it. The file now has a hidden “Fråga-ID” column that makes import round-trippable: download, edit one question’s text, and re-upload — the SAME question updates in place instead of a new one being created (and re-importing an unmodified export is a true no-op). A row with no Fråga-ID (a brand-new question you added in Excel) still gets a fresh id automatically, exactly as before.
Rotate questions smartly (pulse). For pulse-format surveys, the Content step shows a third toggle next to “Require all questions” and “Randomize order”: “Rotera frågor smart (puls)” (requires question rotation to be enabled — the same engine that already drives recurring pulses’ question rotation). It’s a shortcut for the exact same setting as the existing Question rotation card in the Schedule step. With the further, separate “Smart pulse rotation” feature (off by default) enabled, the next round’s rotating question(s) are no longer chosen round-robin — instead they’re picked by the just-closed round’s own real response scores, so the most notable/lowest-scoring questions are the ones re-asked. With no prior-round data yet (e.g. the series’ first rotating round), selection automatically falls back to round-robin.
AI follow-up question (admin config + live during survey-taking). An admin can mark individual questions as “eligible” for a live AI follow-up question (optional “AI follow-up question” feature, off by default) and optionally set a sensitivity threshold (0-10). With the further, separate “Live AI follow-up” feature (also off by default) enabled, the question is actually shown to the respondent immediately under a borderline/notable answer — see the “Live AI follow-up (survey-taking)” page for the full flow.
The anonymity floor can only be raised, never lowered. The minimum group size (Anonymous responses) has always had a server-checked floor of 3. There is now also a ratchet against lowering it again once set — if you’ve set it to, say, 8, you can raise it further but never lower it (not even to a value that would otherwise be legal, ≥3) — anonymity once promised at a given group size must never be walked back.