Every MX8 survey instrument renders in both the chat and traditional styles, and the style is set once per survey in the Theme tab. This reference summarizes how each instrument behaves in each mode and which style we recommend when you have a choice.

For background on the two styles and when to choose between them, see Traditional or Chat based survey experience?.

How to read this table

• Chat and Traditional columns describe how the instrument is laid out in each style.
• Recommended flags the style we generally prefer for that instrument. "Either" means both styles work well.
• Grid-style instruments render row-by-row in chat. They are fully functional, but with many rows they can feel repetitive and add length to the survey — prefer the traditional style when grids dominate the questionnaire.

Compatibility Matrix

Instrument	Chat behavior	Traditional behavior	Recommended
Select	Single-question conversational prompt with tap-to-select options	Standard radio list or button group	Either
Multi-select	One prompt with tap-to-toggle options	Checkbox list	Either
Rating	Single rating prompt (slider, button, or star)	Inline rating control	Either
Net Promoter Score	0–10 scale as a single prompt	Standard 0–10 scale row	Either
Number	Inline numeric input in the chat thread	Numeric field on the page	Either
Text	Message-style open-end input	Text area on the page	Either
Ranking	Drag-and-drop or tap-to-rank with anchor labels	Grid or stacked ranking, optimized for larger viewports	Either — chat for short lists, traditional for long ones
This or That	Each pair shown as its own row with two side-by-side options	All pairs displayed simultaneously in a grid	Chat for engagement; Traditional when there are many pairs
This or That Rating	Each comparison shown one at a time	Multiple comparisons laid out together	Chat
MaxDiff	Two passes per set (one per label, e.g. "Least" then "Most")	Each set shown once with both labels side by side	Traditional — especially with many sets; chat works but doubles the prompts
Grid Rating	Each row presented individually with its rating control	Full grid: rows as items, columns as ratings	Traditional
Grid Select	Each row shown as a vertical block	Tabular grid with rows and columns visible at once	Traditional
Multi-select Grid	One row at a time, with checkboxes or buttons per row	Grid with rows and columns visible simultaneously	Traditional
Numeric Grid	Each row shown with numeric inputs for its columns	Full grid with scrollable columns and autosums	Traditional
Text Highlighter	Passage rendered in the chat bubble; respondent drags to highlight	Passage rendered on the page; respondent drags to highlight	Either
Consent	Consent message delivered conversationally with accept control	Consent screen with accept control	Either
Kid Picker	Guided multi-step prompts (count → ages → genders → availability)	Sequential screens and compact multi-field forms	Chat — the guided flow maps naturally to the conversational style
Get All Children	Three-step flow: count → ages → genders	Sequential screens with optional summary list	Chat
Location	Inline location prompt	Standard location prompt	Either
Secure Video Player	Video fills most of the screen; dial, if used, appears beneath or beside	Video plays centered or fullscreen; dial appears on the right or bottom	Either — both support DRM playback and dial testing
Show Message	Message shown as a chat bubble	Message shown on a page	Either
Show Image	Image shown in the chat thread	Image shown on the page	Either
Show Embedded Content	Embedded content rendered inline	Embedded content rendered on the page	Either

Practical guidance

• If your questionnaire is dominated by grids or a large MaxDiff, choose the traditional style. The grid instruments were designed to be seen as a whole, and running them row-by-row in chat adds length without adding insight.
• If your questionnaire is mostly single-answer, rating, and open-end questions, choose the chat style. You will typically see faster completion, higher completion rates on mobile, and lower rates of poor-quality responses — see the head-to-head numbers.
• If your questionnaire is mixed, pick the style that fits the majority of the critical questions. Because the style is set per survey, it is often worth splitting long studies with very different sections into separate surveys linked by a pass-through.