Questions Step

The Questions step is where you compose the flow that respondents see. Think of it as the palette that feeds the completion moment: each block or step contributes signals, variables, scores, and routing decisions before you land on the final card.

Blocks & steps

Editor modes

Question Blocks supports two editing modes:

• List: Traditional accordion editor for detailed block settings.
• Canvas (Beta): Visual map of Start, blocks, and completion outcomes for faster flow inspection.

In Canvas mode you can:

• Add blocks from the in-canvas Add block panel.
• Drag and reconnect routing edges for Jump Logic, step jumps, and agent exits/default routes.
• Double-click supported nodes to edit prompt text without leaving Canvas.

Canvas routing guardrails match builder guardrails:

• Destinations are forward-only.
• Lead capture can be a destination and can also use a step-level jump destination as a source step.
• Structural warnings/errors shown in Canvas are projection hints; publish validation remains authoritative.

Block options

• Question Blocks overview: Base your flow on text, multiple choice, dropdown, choice buttons, yes/no, scale, slider, card/carousel, product cards, and YouTube blocks. Visit the Question Blocks guide for the full palette.
• Multiple Choice block: Single- and multi-select with scoring, tags, and Jump Logic. Use the Settings cog on the block to open the Multiple Choice settings panel where selection mode, scoring, style, labels, and randomization live. See the dedicated guide for step-by-step usage.
• Dropdown block: Single-select only with compact UI, plus scoring, tags, and Jump Logic. Use this when you want a cleaner selector for long option lists.
• Yes / No block: Binary logic with separate tags and scoring that plays nicely with Jump Logic rules.
• Other blocks (Text, Choice Buttons, Scale, Cards, Product Cards, YouTube): Each exposes similar settings for prompts, validation, and CTAs; reuse tags and logic the same way the multiple choice and yes/no pages describe.

Steps

• API Step: Run a server-side HTTP request in the flow, map response values into quiz variables, and optionally route failures. The API Step guide covers setup, testing, and capture mappings.
• Agent Step: Slide in GPT to harvest structured variables, render buttons/cards/carousels, or route respondents before they hit completion. The Agent Step guide explains configuration and prompt controls.

Supporting functions

Several shared functions power the Questions step:

• Response Tags drive placeholders like {{tag:<questionId>}}, {{tags:<questionId>}}, and derived custom variables.
• Variables Library gives you a cross-quiz control panel for definitions, usages, renames, and archive state.
• Logic & Branching explains Jump Logic rules, operators, and destination mechanics.
• Scoring Logic covers scoring rules, {{total_score}}, and completion rule ordering.

Step-by-step

1. Open the Questions step in any quiz.
2. Add blocks and steps in the order respondents should experience them.
3. Configure prompts, variables/tags, scores, Jump Logic, or agent instructions as needed.
4. Reorder, preview, then move on to Completion when the flow is ready.

Tips / Gotchas

• Every block benefits from variables/tags: custom variables keep personalization consistent across logic, completion copy, and exports.
• Jump Logic is forward-only; route only to later questions or completion and reorder rules to keep the most important checks at the top.
• Scoring flows into completion rules; if you expect {{total_score}} in copy or use score-based routing, enable scoring on every applicable block.
• Agent Step output variables behave like any other variable, so mention their keys in instructions and reuse them in Jump Logic or completion copy.
• Canvas is best for flow shape; list mode is best for deep settings. Switch between modes as needed while building.