When people talk about user experience in surveys, they usually mean the questionnaire itself. But respondents meet the survey earlier: in the message that reaches their phone, in the link inside it, and on the first screen that opens. At this stage they decide whether to open the survey at all, and whether to go on.
Respondents are wary, with good reason, of links that arrive by text message. A message whose sender is unclear, or a link that looks suspicious, lowers the response rate before a single question is asked. Low response also affects the composition of the sample.
The message and the link
- Who is asking and why. A line or two: who runs the survey, what it is about, and how long it takes.
- An address that earns trust. A short link on a recognizable website address, not an anonymous shortener that looks like spam.
- A tidy preview. In WhatsApp the link appears as a card with a title and an image, and they should belong to the organization that is asking.
Personal link or general link
A personal link, unique to each respondent, is almost always the right choice. It prevents duplicate entries, allows reminders to be sent only to those who have not answered yet, and brings respondents back to the point where they stopped. A general link suits open distribution, for example on social media, but then other safeguards against duplicates are needed, and usually screening questions at the start.
What travels inside the link
A good link carries information that is already known, so there is no need to ask for it again: the channel the respondent came from, their panel ID, or the quota group they belong to. That makes it possible to follow quotas in real time and keep the questionnaire short.
What does not travel in the link: identifying details. A name, phone number or email address should not appear in an address that can be forwarded.
Working with a panel
At the end of the survey, every respondent is redirected to the panel with a status: completed, screened out, or arrived after their quota was full. These redirects must work exactly, because panel payments and response-rate calculations depend on them. Each of the three routes is therefore tested before fieldwork opens.
The first screen
A short welcome: who runs the survey, how long it actually takes, and what will be done with the information. Then a large start button, with no long page of terms before the first question.
An accurate time estimate is part of the terms of participation. A respondent promised five minutes who is kept for twenty will not take part in the next survey.
Testing before fieldwork
Before every fieldwork we open the link from several apps, WhatsApp, text message, email and Facebook, and on several phones. We also test the less routine cases: a second click on a personal link, entry after a quota has filled, and entry after the survey has closed.
When this stage is planned properly, respondents immediately recognize who is asking and start answering.