Skip to main content
Share is where a finished study goes out to respondents by link, or with a person conducting the interview on the respondent’s behalf. From the Share tab’s links table, add a share link. If no link is currently live, this both activates your study and creates the link in one step; if a link is already live, a new one is created inactive. Each link in the table has its own copy button, so you can pull the URL to send over whatever digital channel reaches your respondents (email, WhatsApp, and similar), and its own rename and archive controls, so you can keep a study’s links organised as it evolves. Respondents complete the study on their own device, in their own time, choosing to type, speak, or record a short video for each open-ended question, in their own language.

​
Which version respondents see

Your changes in Plan save as you make them, while a share link keeps serving the version of the interview it is set to until you update it. When you have changed the interview since, a note in Share says so, and Update all share links to the latest save moves every link onto your newest save. Each link’s menu also offers Update to the latest save for that link alone. Plan’s outline shows the same note with Update in Share, and a small dot on the Share tab tells editors that the live link still serves an earlier save. The links table’s Version column names the version each link serves. A version gets a number, such as Interview v2, once respondents first reach it through a share link, and Analyse and the export use that number. To try the interview yourself, open the test link from Preview in Plan. It always shows your latest save and allows five test interviews per saved version, and test interviews never count as responses. See Collection methods for the in-person and phone-call paths voi supports when a respondent lacks a reliable connection, a device, or the digital comfort to answer alone.

​
Choosing a tier and planning your sample size

Before you activate, you choose a tier for the study: Starter, Growth, or Pro, each with its own guaranteed number of respondents, its own price in credits, and its own set of unlocked features, such as Explore or the data quality and integrity audit. A comparison view lays out all three side by side if you’re not sure which fits. To help you decide how many respondents to aim for, a sample size calculator sits alongside this choice. Set the size of the population you’re trying to represent, a confidence level (90%, 95%, or 99%), and a margin of error (±3%, ±5%, or ±10%); it tells you how many responses you’d need, or that you should aim to reach everyone if your population is small enough to survey in full. Apply its recommendation directly to set your respondent count. This is a planning step you use once, before you launch. See Billing & subscription for how tiers and credits work.

​
Activating your study

Activating spends credits and creates your live link in one action; the button reads Activate (or Continue to payment if your balance is short) and shows exactly how many credits that will cost. Confirming shows you, before anything is charged: the study name, the tier and guaranteed responses you’re buying, and a reminder that credits are spent at launch and aren’t refunded. If the study is funded by another organisation rather than your own balance, you’re shown that here too. You also see a preview of exactly what respondents will be asked to consent to before you commit. A missing privacy policy link produces a warning at this point, not a block; you can launch anyway, though your organisation stays responsible for that decision as the data controller. An unaccepted Data Processing Agreement is a hard block instead: your workspace has to accept it first, and only an owner can do that. See Workspace settings.

​
Study status

A study’s status is set automatically by what’s actually happening with it, not by a manual pause button. To get a paused study collecting again, fix whatever’s blocking it (raise the limit, top up your balance, resolve a payment issue) and press Resume; resuming itself never charges anything extra. See Billing & subscription for how the underlying limits and balance work. The welcome screen tells respondents an AI system is conducting the interview, before the first question, and the default consent text repeats it. You can customise this text in the interview outline in Plan, but you stay responsible for it as the data controller, and the AI-processing and recording disclosures need to stay in place. EU transparency rules require respondents to be told. Where your study touches health or other sensitive topics, check what your own national law requires for explicit consent.

​
Extra respondents beyond your guaranteed limit

There’s no fixed start or end date for collection; a study simply collects until you close it or it pauses. What you do control is how many respondents beyond your tier’s guaranteed number you’re willing to allow, set as a cap on the Share tab. Whether extra respondents are admitted automatically from your balance, or the study pauses instead once the limit is reached, is a setting the paying account controls; see Billing & subscription. Preventing the same respondent from submitting twice is handled separately, back in Plan’s study settings, as the Allow repeat submissions toggle alongside the 48-hour resume window.

​
Pausing and resuming for a respondent

This is distinct from the study’s own status above: an individual respondent’s progress saves as they answer, and they can resume an unfinished interview themselves within 48 hours. After that, voi deletes their session automatically, along with any answers and recordings it held. Once responses start coming in, move to Review to track them.