Test the survey people will actually answer.

The editor shows how a survey is configured. Preview shows what that configuration becomes for a respondent: the introduction, the question sequence, the wording produced by piping and the route selected by an answer. Seeing those elements together gives a different perspective on the draft.

Preview runs use temporary answers and do not create stored response records. You can complete one route, start over and try another without adding test submissions to analytics. That makes it a useful part of development and review, before the survey link is used for real collection.

The first complete run often follows the route the author knows best. Other answers reveal more: a skipped optional input, a value at a numeric limit or a changed answer after going Back. These cases matter when the next question depends on what came before.

Lensym lets reviewers try those alternatives within the respondent flow. A branch can be followed from both sides, a required question can be left empty and a piped sentence can be read without its expected source. The result is a concrete basis for revising wording, settings or routes.

• Check every intended route through to its ending.
• Read piped wording when its source is blank or bypassed.
• Try long answers and a narrow screen.
• Repeat affected checks after changing question order or rules.

Before publication, Lensym validates the survey and identifies configuration issues that need attention. Incomplete conditions, missing destinations, unsuitable references, invalid numeric settings and features outside the selected plan can be addressed before the questionnaire is opened to respondents.

These checks complement a human review. A reference can exist while its wording is confusing, and an accepted condition can still express the wrong eligibility criterion. Validation provides evidence about the configuration; Preview provides an experience to examine; the research team decides whether the instrument fits its purpose.

Published questions are locked, giving the draft stage a clear role: it is where wording, routes and calculations are reviewed. A colleague can take the survey without the author explaining each screen, and the team can repeat the affected checks after a revision.

Preview does not establish live response saving, export contents or the distribution of published Split assignments. Those need appropriate collection checks. It also does not establish the validity of a research measure. Its value is specific: making the respondent flow available for examination before the study moves into collection.