An allocation target sum question asks respondents to distribute a fixed amount, such as points, dollars, hours, or percentages, across a list of choices. It works differently from a numeric question where each answer stands on its own: because the total is capped, giving more to one choice leaves less for the others. That constraint is what makes the data valuable. It shows relative priority rather than absolute opinion.
The target can be the same for every respondent, such as 100 points, or different for each respondent by piping in a number they gave earlier in the chat, such as their weekly grocery spend.
Use allocation target sum when you need to understand:
How respondents would split a fixed budget, spend, or amount of time across options
Relative importance between items that all look important when asked about one at a time
What respondents are willing to trade off, and what they give up to fund something else
Creating an allocation target sum question
To add an allocation target sum question to your chat:
Open the question picker and select Allocation Target Sum under Open Ended Questions.
Enter your question text.
Add your choices. A minimum of two choices is required. You can populate choices dynamically from a previous multiple choice question or a previous allocation question, so only items relevant to that respondent appear. When masking from a previous allocation question, choose which choices to carry through: those with a value greater than zero, those with a value equal to zero, or all choices.
You can randomize the order in which choices are displayed, and anchor specific choices to always appear in a fixed position.
Open Settings and choose a validation mode, either Must Equal or Must Not Exceed.
Set your target sum. Choose Custom value to enter a fixed number that applies to everyone. Or pipe the target from a previous hidden variable or a previous numeric question, such as an open ended number or a numeric slider, to give each respondent a target based on their own answer.
Set a minimum if you are using Must Not Exceed. The minimum is the smallest total a respondent can submit, and it can also be a custom value or piped from a previous question.
Choose whether to allow decimals. Responses are whole numbers by default. You can enable decimals and allow up to two decimal places.
Set a step value. This controls how much the value moves each time a respondent moves the slider, so a step of 5 moves the value in increments of 5. Decimal step values are only available when decimals are enabled.
Turn on Unit label if you want a unit shown alongside each value. Enter the unit and choose its position: left as a prefix, up to 3 characters, or right as a suffix, up to 6 characters. The preview on the right shows how the value and unit will appear to respondents.
Save your settings.
All settings on the card are global and apply to every locale in a multilingual chat, with one exception. Unit labels are locale specific, so you need to turn the unit label on and enter the unit on each language page.
Choosing a validation mode
The validation mode decides what counts as a complete answer, so it is the setting that shapes your data the most.
Must Equal - the total must match the target exactly. Submit stays disabled until the respondent lands on the number. Use this when the whole amount has to be accounted for, such as splitting 100 points across brands or allocating an entire budget.
Must Not Exceed - respondents can allocate up to the target but do not have to reach it. Use this when leaving part of the amount unallocated is a real and meaningful answer, such as how much of a budget someone would actually spend. This is the only mode where you can set a minimum, which gives respondents room to stop short of the target while still requiring them to allocate something.
Customizing the appearance
Appearance settings let you change how the question looks and feels to respondents without changing how it works or what data you collect. Open Appearance on the question card to configure them. Everything here is optional, so you can leave the defaults in place if you want the simplest possible experience.
Interaction style - choose between Classic fill, where respondents allocate using a slider track under each choice, and Tactile liquid fill, where the whole choice row acts as the slider and fills as the respondent allocates. Tactile liquid fill is more engaging and more physical on mobile, but respondents have to work out that they can slide on the row itself, so consider your audience before turning it on.
Distribution bar - show a stacked bar in the header that breaks the allocated total down by choice, so respondents can see the shape of their allocation as they build it.
Media icons - add an image or emoji to each choice to make the list quicker to scan.
Colors - set the colors used for the fill and the distribution bar to match the look of your chat.
Whichever interaction style you choose, respondents can always type a value directly into a choice instead of sliding.
What participants see
When an allocation target sum question appears in the chat, respondents see each choice with its own value, all starting at zero, and a header showing the target along with how much they have allocated so far and how much is left. The remaining amount updates as they go, so they never have to add the numbers up themselves.
Respondents allocate by sliding on a choice, using the slider track in Classic fill or the choice row itself in Tactile liquid fill, or by typing a value into the choice directly. As the total gets close to the target, the sliders adjust to what is actually available: once nothing is left, a slider will stop at its current value rather than letting the respondent go over. To give more to one choice at that point, they can go back and lower another choice to free up the amount.
On mobile, the header stays pinned to the top of the screen while respondents scroll, so the target, the remaining amount, and the distribution bar are always visible no matter how many choices there are.
The Submit button enables once the allocation satisfies the validation mode: exactly the target for Must Equal, or at or below the target and at or above the minimum for Must Not Exceed.
Every choice starts at zero, so a choice a respondent chooses not to fund is recorded as a zero rather than a blank. That zero is a real answer and is treated as one in reporting.
Where your data appears
Reports
A top line report is created automatically as soon as responses come in. It shows the number of responses to the question and a table with a row for each choice, so you can see how the allocation is distributed across choices and sort by any column to find the highest and lowest allocated items.
Because a choice left at zero is a deliberate answer rather than a blank, the report gives you a mean that includes zeros and a mean that excludes them. Use the mean including zeros to understand allocation across everyone who saw the question, and the mean excluding zeros to understand allocation among only the respondents who funded that choice.
You can also add the question to an insight report or a custom report, where you have full control over how the results are presented and can build on the same figures.
Exports
Exports give you the response level data. Each respondent has one row, with a column for every choice holding the value they allocated to it, followed by a column for their total allocation.