Publishing & sharing

Draft, published, archived

A survey is a stable identity with a history of versions behind it. You always edit a draft; respondents always get the latest published version; older published versions become archived but are never deleted.

• Opening the editor on a published survey starts a new draft, seeded from what is currently live. The live version keeps serving respondents while you work.
• Publishing marks that draft as live and archives the version it replaced.
• Every response records the exact version it answered. A response collected last month is still tied to the questions as they were worded last month — changing the survey never rewrites history.

The public link

After publishing, copy the link from the bar at the top of the builder, or from the ⋯ menu on the Surveys list. It looks like https://…/s/istimara-k7mq3xrp and is safe to paste into WhatsApp, an email or a poster QR code.

The address does not contain the survey's name. It is generated when you create the survey: a fixed word and a random code, the same length every time. Your survey's name is internal — it is what your team scans in a list, and it has no business being published to every respondent and every forwarded message. The random part is also what keeps the link unguessable, which matters because anyone with the link can open the survey. Renaming the survey never changes the address, which is what keeps printed posters working.

• Sharing the link on WhatsApp or social media shows a preview card with the survey's own title and description — so write them for a reader, not for your filing system.
• Survey links are unlisted: they are deliberately kept out of search engines. Anyone with the link can open it, but nobody will stumble on it through Google.
• The address never changes, so a printed QR code stays valid across every future redesign of the survey.

QR codes

Every survey link comes with a QR code, generated on the spot. Open it from the QR button beside the link on the Surveys list, from the ⋯ menu, from the survey's own page, or straight from the dialog that appears the moment you publish.

1. 1

   Pick a size

   1024 px is the right answer almost always: it survives WhatsApp's re-compression and prints cleanly at the size a poster code usually gets. Go to 2048 for large-format print, or 512 for an on-screen overlay.

2. 2

   Or take the vector

   Download SVG gives a vector file that stays sharp at any size — hand this one to a designer, or drop it into a banner that will be printed metres wide.

3. 3

   Or print the poster

   Print lays the code out on a page with the survey's title, the address underneath and the workspace's branding, ready to pin to a noticeboard. Choosing "Save as PDF" in the print dialog gives you a file to send on.

• Share hands the image to your phone's own share sheet — straight into WhatsApp, Mail or AirDrop — on devices that support it. On a desktop browser, use Copy image and paste it wherever you are writing.
• Square or Rounded changes only the look of the code. Both carry the same address and both scan the same; square is the safer choice for anything that will be photocopied.
• The centre mark can be switched off with the Logo toggle. Leave it on unless a particular printer or scanner is giving you trouble — the code carries enough redundancy to be read straight through it.

Opening and closing on a schedule

A published survey can open itself on a date and close itself on another, so nobody has to be awake at midnight to press a button. Set it in the builder's Survey settings.

• Neither set — the survey takes answers for as long as it is published. This is the default and nothing changes.
• Opening date only — announce it now, let it go live on the day. Before then, visitors see a page telling them the date it opens.
• Closing date only — it is open now and stops on the day, without anyone remembering to pause it.
• Both — a fixed window, start to finish.

1. 1

   Pick a day

   Open the calendar and choose a date. It stays open afterwards, because a schedule is a day *and* a time.

2. 2

   Pick a time

   The clock is in the same panel, under the calendar. Tap 09:00, 12:00, 17:00, 23:59 or Now, or type any time you like. Then Done.

3. 3

   Read it back

   Under the two fields the window is written out in full — weekday, month and time. Check it: 03/08 and 08/03 are the same eight characters, and a date typed the wrong way round is invisible until the survey opens on the wrong day.

For the common windows there is a shortcut: One week, Two weeks and One month set the closing time from the opening one in a single click. If no opening time is set yet they start from today at 09:00, and they always close at 23:59 so the last day counts in full.

The two dates are read in the workspace time zone, shown above the fields. Outside the window the public page explains itself, and the survey refuses submissions even from a tab that was already open — so a survey left sitting overnight cannot post into a closed survey.

Publishing in one language or two

Surveys are bilingual by default and respondents switch with the toggle in the header. When a survey is only meant for one audience, you can publish it in Arabic only or English only — the toggle disappears and the survey always opens in that language.

Nothing is thrown away. The other language's text stays in the survey and the builder keeps both tabs, so switching back restores it exactly as it was. The choice is about what respondents are offered, not about deleting your work.

A single-language survey also overrides whatever language the visitor's browser is set to — including the link preview shown when the address is shared in a chat.

The published language must be complete. Publishing an Arabic-only survey checks that the title, every question label, every option and every text block exists in Arabic — otherwise a respondent could meet English text inside a survey that promised Arabic. The builder lists exactly which fields still need it, and the language toggle in the builder only flags gaps in the language you are actually publishing.

Public or invite-only

Inviting respondents by email

Switch the survey to Invite only on its page, then paste the email addresses — one per line, or separated by commas — and send. Each person receives their own link, branded in the workspace colours and written in their language.

Every invitation shows its own state and the date and time it reached it, so you always know where a round of outreach stands:

• Resend sends the same personal link again — for someone who lost the email. There is a two-minute pause between sends to one person and a limit of five resends per invitation; past either, copy the link and send it yourself.
• Revoke closes a link immediately, for example when someone leaves a program. Someone who already responded cannot be revoked — their answer is in, and the link is spent.
• Invite again brings a revoked person back with a new link. The one you revoked stays blocked.
• Duplicate addresses and malformed ones are skipped and reported back to you, so pasting a messy list is safe.
• Once the list passes 10 people it gains a search box and a filter by state, and pages 10 at a time — so "who has not answered yet?" stays one question rather than a scroll.

How a language is chosen

An invitation is written in the recipient's language, not in the one you happen to be working in.

The platform takes the first of these it can answer, and works down only when it has to:

1. 1They are on your team. Their Communication language decides it — a setting they chose themselves. See Your account.
2. 2They have answered a survey before. The language they filled it in is used — not a stated preference, but a real choice they made, and the most recent one wins.
3. 3Neither. A survey pinned to one language is invited to in that language, because that is the only language the link will open in. A bilingual survey falls back to the language you are working in.

Controlling an open survey