Skip to content
Open fromusers

Logic and randomization overview

SurveysUsability testsIn-product surveys

Logic and randomization control what each participant sees and in what order. Use them to personalize paths, balance experiment groups, and compare prototypes without building separate studies.

Study Logic and Variables tab for display logic, branching, and variables

ToolUse it whenTypical outcome
Display logicContent should appear only if a condition is trueShow or hide a question, page, or section
BranchingThe next step should change based on answersSkip ahead, jump to a section, or end the study
Variables and pipingYou need reusable values in text or rulesInsert answers, URL params, or sample fields with {{name}}
FormulasYou need calculated valuesScores, segment labels, combined checks
Block randomizationQuestions on a page should be shuffled or sampledReduce order bias inside a page
Section randomizationPages or subsections should be shuffled or sampledRotate modules or show a random subset
Randomization groupsYou need balanced A/B (or multi-arm) assignmentStable study groups for analysis
Compare assetsDifferent prototypes should map to different groupsSame tasks, different designs per arm
  1. Same page, conditional visibility? Use display logic.
  2. Different next destination? Use branching.
  3. Reuse answers or external data in text or rules? Use variables and piping.
  4. Need a calculated value? Use formulas.
  5. Reduce order bias inside a page? Use block randomization.
  6. Rotate whole pages or subsections? Use section randomization.
  7. Assign people to balanced experiment arms? Use randomization groups.
  8. Show different prototypes per arm? Use compare assets.

A common experiment flow looks like this:

  1. Assign a randomization group (for example Group A or Group B).
  2. Resolve variables (URL, participant sample, formulas).
  3. Show only the pages and questions that match display rules.
  4. Map the correct prototype with compare assets.
  5. Branch to follow-ups or an ending after key answers.

For recipes that combine these tools, see Experiment recipes. For the exact runtime order, see How rules are evaluated.

  • Questions: show/hide rules and skip/end branching after an answer.
  • Pages: page display logic, block randomization, page-level compare assets.
  • Sections: section display logic, section randomization, section branching, section compare assets.
  • Study logic: variables, formulas, and randomization groups used across the study.

Always preview after you change logic. Preview helps you verify paths. Live applies randomization and assignment for real participants.

Common questions

When should I use display logic instead of branching?

Use display logic to show or hide content on the current path. Use branching when the participant should jump to another page, section, or ending.

Can I combine randomization and display logic?

Yes. Randomization decides which pages or groups a participant sees. Display logic and branching can still filter questions inside that path.

Where do I configure logic in the builder?

Open the study editor and use the Logic tools on questions, pages, and sections. Variables and formulas live with the study logic settings.