Matrix, distribution, constant sum, and multi numeric questions handle grids, fixed totals, and multiple related numbers on one screen. They are common in product testing, budget allocation, and attribute batteries.
The Matrix group#
The Matrix group in the add-question list holds twelve interfaces. They all ask the same underlying thing, a set of statements rated on a shared set of answers, and they differ in how that is put in front of the respondent: a classic grid, a card stack, a swipe, a dial, a timed bar.
They share their authoring. Eleven of the twelve are built from a Rows list and a Columns list in the question editor, with Add row and Add column beneath each. Rows are the statements, brands, or concepts being rated. Columns are the answers available for each one. Rich text and images work on both lists, and display logic can be put on an individual row or column. The exception is Statement Agreements, which has no column axis at all and authors its statements as Answer Options.
Everything specific to an interface, its layout, timing, and gestures, lives in the Settings tab of the properties panel. Button and helper copy lives in the Messages tab.
| Type | What it is for | What the respondent does |
|---|---|---|
| Standard Matrix | Agreement, satisfaction, and attribute batteries on one screen | Picks a column for each row in a grid |
| Carousel Matrix | The same battery on a phone, one statement at a time | Swipes through statements and picks a column answer on each |
| Transposed Matrix | Brands down the side, one question per screen | Picks brands on each screen, one question at a time |
| Funnel | A purchase funnel, forward or reverse, over a brand list | Picks brands on each successive screen, with the list narrowing as they go |
| Swipe Matrix | Fast sorting of many statements into a few buckets | Swipes a card left or right, or up or down, into one of the columns |
| Brand Association | Which brands go with which concepts | Taps items from a grid to project them onto the concept shown |
| Time Based Rating | Rating where the effort of holding is part of the measure | Slides and holds along a bar to fill a rating, per statement |
| Fast Agreement Matrix | A quick agree / strongly agree battery | Taps a statement once to agree, twice to strongly agree |
| Fast Bipolar Agreement | Trade-offs between two opposed statements or images | Taps toward Side A or Side B, twice for the stronger position |
| Dial Gauge Grid | Percentages and shares rather than scale points | Presses and holds a dial to move each statement from 0 to 100 percent |
| Skin In The Game Rating Matrix | Intensity, positive or negative, on a concept | Swipes or taps a statement repeatedly, each tap moving the score further |
| Statement Agreements | Forced choice between statements, screen after screen | Picks the statement they agree with most on each screen |
Everything on this page is one question in the survey and one set of answers in your data. Logic, piping, carry forward, and section loops treat all of them as matrix questions.
Standard Matrix#
Full guide: Matrix / Likert →
A grid of rows and columns. Each row is a statement; each column is a scale point, for example Strongly Disagree through Strongly Agree.
When to use it:
- Agreement or satisfaction grids covering multiple items at once.
- Attribute ratings across several products or features.
Selection Mode in the Settings tab decides what a row accepts:
| Mode | What it means |
|---|---|
| Single Select | Respondents select one option per row |
| Multiple Select | Respondents can select multiple options per row |
| Text Entry | Respondents enter text in the grid |
| Rank Order | Respondents pick a rank position for each row |
Tip: You can add images and formatted text to both rows and columns for richer visual grids, great for product or concept testing.
Logic support#
Matrix questions have dedicated condition operators (for example "any row equals", "row count >=") that let you build logic based on how respondents fill in the grid. You can also use matrix questions as a source for section loops (loop over rows or filtered subsets of rows) and carry forward (carry only rows that match specific criteria). See Logic, Flow & Conditions for details.
Carousel Matrix#
The same rows and columns as a Standard Matrix, shown one statement at a time so a long battery stays readable on a phone.
Answer type in the Settings tab chooses the behaviour. Single select picks one column per statement and auto-advances. Multiple select allows several columns per statement and shows a Next button.
The rest of the Settings tab is layout. Layout heights splits the screen between Statement %, Choices %, and Instruction %, each a percentage of viewport height. Answer alignment and Statement alignment position the text, and Zoom thumbnail lets respondents tap a statement image to enlarge it.
Zoom thumbnails. When the question text includes an image, tapping the thumbnail opens it full screen. The ? beside the setting, labelled Zoom thumbnail info, explains this in the panel. The captions the respondent sees default to "Tap to view" and "Tap to close" and are editable on the Messages tab.
Transposed Matrix and Funnel#
Two palette entries for the same kind of question. Each column is one screen in the sequence, for example "bought in the past 2 weeks", then "past month", then "past 3 months", and the rows are the brands the respondent picks from on each screen. They differ only in the choices under Selection Mode in the Settings tab, the setting that matters most, because it changes what carries from one screen to the next:
| Question | Mode | What happens |
|---|---|---|
| Transposed Matrix | Single (the default) | Selected brands disappear, and the remaining ones carry to the next screen |
| Transposed Matrix | Multiple | Selected brands carry forward, and unselected ones disappear |
| Funnel | Forward Funnel | The same as Multiple: only the brands picked on a screen carry to the next one |
| Funnel | Reverse Funnel (the default) | Looks like Single, but once a brand is picked it stays recorded on every later screen |
In Single and Reverse Funnel a None of these column is added for you. It is marked as a system row in the Columns list, cannot be edited or deleted, and respondents never see it as a screen. It exists so brands nobody picked are still recorded. Multiple and Forward Funnel don't have it. A Reverse Funnel made before October 2026 has no such column until you switch it to Forward Funnel and back; its Settings say so. See Transposed Matrix and Funnel for the details.
Also in Settings: Number of Columns sets the grid width separately for Desktop and Mobile, Fit to Screen shrinks the cards to avoid scrolling, Keep Column Position holds cards in place as the list narrows, and Show Instructions in Popup can be set to Both, Mobile, Desktop, or Neither. Button labels are authored in the Messages tab.
Selection mode. The ? beside it, labelled Selection mode info, explains the modes in plain words for the question you are editing: Single and Multiple on a Transposed Matrix, Forward Funnel and Reverse Funnel on a Funnel.
Validation messages. Min-selection validation message and Max-selection validation
message are the texts a respondent sees when they have picked too few or too many. Each carries its own
placeholder, minValidateLimit and maxValidateLimit, replaced with the real limit when the
message is shown, so one message serves every limit you set.
Swipe Matrix#
Each row becomes a card, and each column becomes a direction to swipe it. Good for sorting a long list quickly.
The column list is not edited directly here. Swipe Directions in the Settings tab controls how many columns exist, up to four, and the Columns list says so where the Add column button would normally be.
Other Settings worth knowing: Show Buttons gives desktop respondents buttons instead of a drag, Back Button lets them undo the last card, Card Width, Stack Depth, and Stack Direction shape the deck, Rotate While Swiping tilts the card as it is dragged, and Swipe Overlay Opacity sets how strongly the direction tints the card. Questions (Rows) Display chooses whether a card shows its text, its image, or both.
Directions sets how many categories a card can be swiped to: 2-Way, 3-Way or 4-Way. The buttons are exclusive and the selected one is filled in.
Brand Association#
A concept is shown, and respondents tap items from a grid to associate with it.
Projection Variant in the Settings tab puts the concept either above the grid (Row Top) or below it (Row Bottom). Concept Area Height sets how much of the screen the concept takes. Under Validation, Min / Max Selections sets how many items must be picked, and Allow Deselect lets a respondent tap a chosen item again to remove it. Auto-advance moves on once the selection is complete.
Time Based Rating#
An animated rating bar. Respondents slide and hold, and the rating fills while they hold, so hesitation is visible in the data.
Variant picks between Multi-Statement, one concept rated on several statements, and Multi-Question, a separate concept per screen.
Because the rating is a single value, this type has no column list. In Multi-Statement the Columns section is replaced by Anchor Labels, where you write the Left Side Label and Right Side Label that sit at the ends of the bar.
Rating Display switches between Pips, a numbered dot scale, and Stars, where respondents slide and hold to fill stars. With Pips, Rating Scale offers a Type of Left from 0, Left from 1, or Centered; a centred scale needs an odd number of points so the middle pip exists, and the count is bumped for you when you switch to it. With Stars, set Number of stars under Star Scale. Animation Speed controls how fast the fill moves.
Animation settings. Dot fill speed (ms) is the auto-fill cadence while a respondent drags or holds the bar, and a lower number fills faster. Pip (s) sets the pip animation in seconds, from 0.1 to 2. Initial highlight is the point the bar starts on, and it accepts a negative number so a centred scale can start below the midpoint.
Input mode switches between Drag & Hold, where the rating fills while the respondent holds, and None (tap a dot), where they simply tap the point they want and nothing is timed.
Zoom thumbnails lets a respondent tap a row image to open a full-screen view of the concept. With it on, two captions become editable on the Messages tab: Tap-to-view caption (zoom mode), which defaults to "Tap to view full size concept", and Tap-to-close caption (zoom mode), which defaults to "Tap to close".
Fast Agreement Matrix#
An agreement battery reduced to two taps. One tap on a statement means agree, a second tap means strongly agree.
The Columns list is fixed at two columns, Agree and Strongly Agree, and cannot be added to. Author the statements as rows.
In the Settings tab, Slide Layout sets Items per slide, so a long list is broken into screens. Under Display, Randomize within slides shuffles the statements inside each group, and Show Instructions in Popup shows the question text in a popup before answering.
Fast Bipolar Agreement#
A statement or image is shown on each side and the respondent taps toward the one they agree with, twice for the stronger position.
Rows are authored as pairs rather than single statements: each row has a Side A and a Side B. The Columns list is fixed at five columns covering both strong positions, both moderate positions, and the neutral middle, and cannot be changed.
Under Display, Randomize A/B positions flips which side each statement appears on, so side does not bias the answer, and Show Instructions in Popup shows the question text first. Timing controls the pacing of the transitions.
Tap colours#
Both Fast Agreement Matrix and Fast Bipolar Agreement carry a Tap Colors pair on the Messages tab: First tap color and Second tap color. They are colour pickers, and they set what a statement looks like after one tap and after two, so the two states stay distinguishable in whatever palette your survey uses.
Dial Gauge Grid#
Each row is rated from 0 to 100 percent on a dial, driven by press-and-hold buttons. Use it when the answer is a share or a percentage rather than a scale point.
There is one column, which holds the percentage, and it cannot be added to or renamed. Everything you author is the row list.
In the Settings tab, Gauge style chooses Half-circle or Full-circle, Question text position puts the statement Above statement or Below statement relative to the dial, and Fit to screen scales the layout to the viewport. Sizing sets Gauge (%) plus Statement (desktop) and Statement (mobile) text sizes.
Skin In The Game Rating Matrix#
A concept is shown at the top and each statement below it is a bar. Swiping or tapping right raises the score, left lowers it, and repeating the gesture pushes it further, so intensity is captured as well as direction.
There is one column, which holds the score per statement, and it cannot be added to.
Scoring scale in the Settings tab sets how far the score can travel: 2×2 (max 4 per side) or 3×3 (max 9 per side). Under Layout, Statements per screen and Concept height (%) split the screen between the concept and the bars, Text Alignment positions the statement inside its bar, and Tap to zoom concept lets respondents enlarge the image.
Bar icons are the artwork the respondent taps, set as URLs and authored on the Messages tab beside the button copy rather than on Settings, because they read as part of the wording. The Tap-to-zoom hint is the caption inviting a respondent to tap an image for a closer look.
Statement Agreements#
A forced choice. A handful of statements is shown per screen and the respondent picks the one they agree with most, then the next screen appears.
This is the one Matrix type with no rows and columns. Write the statements in the Answer Options list, and the interface chunks them into screens for you.
Answer type in the Settings tab chooses Single, one pick per screen where the tap moves on, or Multiple, where several can be picked and a Next button appears.
Under Screens, Statements per screen sets the chunk size, Pause before next screen (ms) sets the beat before advancing, and Back arrow delay (ms) sets how long before the back arrow becomes usable again.
Under Randomization, Randomize screen order shuffles the screens. Statement order options is a master switch for the two below it: Shuffle within screens shuffles the statements on each screen, and Reverse within screens reverses them. Both are greyed out while the master switch is off.
Distribution (Constant Sum)#
Respondents allocate a fixed total (points, percentage, budget) across several options.
When to use it:
- Budget allocation questions.
- Time or share-of-wallet questions.
- Any scenario where the total must add up to a specific number.
Respondents see the total they need to reach (for example, 100%), and analytics show the average allocation per option.
Constant Sum#
Respondents allocate numeric values across items that must add up to a target total (not necessarily 100%). Use this when you need counts, points, or dollars rather than percentages.
A plugin variant also exists with enhanced controls (sliders, steppers) for mobile-friendly allocation studies. See Display Blocks & Plugins.
Constant Sum settings#
| Setting | What it does |
|---|---|
| Max total | The target every allocation must add up to. Defaults to 100, and goes far higher |
| Max digits (integer mode) | How many digits a respondent may type when values are whole numbers |
| Max digits (decimal mode) | The same limit when decimals are allowed |
| Max editable characters | How much of a box's text a respondent can edit |
| Box padding (px) | Spacing inside each allocation box, 0 to 32, for fitting longer numbers |
Two message fields sit on the Messages tab: Popup confirm text, the label on the confirm button of the allocation popup, and Tap/Hold instruction text, the line that tells a respondent how to enter a value.
Multi Numeric#
Multiple labeled numeric fields in one question, useful when you need several related numbers on the same screen (for example, units sold per product line).
Configuration tips#
- Label each field clearly in the properties panel.
- Use validation rules where min/max per field matters for data quality.
Jobs to be Done Multi-Tap#
Respondents tap each item once or more to mark it against a set of criteria, so one pass over the list records several judgements.
Tap mode is a pair of exclusive buttons, 2-tap (2 criteria + Clear) and **3-tap (3 criteria
- Clear)**, and it controls how many criteria columns are stored. Clear is always available in both modes, so a respondent can always take a mark back.
Max taps per criterion caps how far one criterion can be tapped up, from 1 to 8, defaulting to 4. Tap instruction text, on the Messages tab, is the line telling respondents how to use it.
Icon Rating Multi-Matrix#
Concepts rated on criteria using icons rather than a numeric scale.
Criteria rows per screen sets how many criteria a respondent sees at once, so a long criteria list can be split across several screens instead of scrolling.
Related guides#
- Question Types — overview and when-to-use table
- Logic, Flow & Conditions — matrix operators, section loops, carry forward
- Platform Feature List — plugin constant-sum UI and matrix-based plugins
Was this page helpful?