Skip to main content
Dashboards & Charts let you turn Feedback Records into visual analytics. A Chart is a single visualization scoped to one Feedback Dataset. A Dashboard is a grid of charts you can share with your team. Charts read Feedback Records, not survey responses. A workspace therefore needs a Feedback Dataset and at least one Feedback Source before any chart has data - see the guide below for that path end to end.

Creating your first charts

From the NPS template to a shareable dashboard, with three hands-on CX use cases.

Charts

Charts live under Analyze → Analysis. Each chart is a query plus a visualization config.

Available chart types

You can build a chart in two ways:

Manual builder

Pick a Feedback Dataset, choose dimensions and measures (count of records, average NPS, …), apply filters, and select a chart type. Live preview updates as you tweak.

Filter by embedded data

Embedded data that a Feedback Source carries on its records (a plan, a region, a customer segment) shows up in the builder under Embedded data, next to the dataset’s other fields. You can group by an embedded-data field and filter on it like any other text field. The builder lists the 100 most frequent embedded-data keys of the dataset; a chart already saved on a rarer key keeps working.

Filter by answers to another question

Use Answer to question in the filters to keep only the records whose submission answered another question of the dataset a certain way, for example “comments from people who gave an NPS of 9 or 10” or “CSAT of customers who chose Enterprise as their plan”. Pick the question, then the answer it must have; for single-choice and rating questions you can also pick an answer it must not have. A multiple-choice question with several selections matches when any of the chosen options was selected, and has no “is not”. A submission that skipped the question never matches. A chart takes up to 5 answer filters, and you can combine them with other filters in AND and OR groups.

Matrix chart

To chart a matrix question:
1

Pick the Matrix type

In the chart builder, select Matrix in the chart type switch above the preview.
2

Pick your matrix question

Choose the question under Matrix question at the top of the left panel. Formbricks fills in the rows (the statements), the columns (the scale points) and the measure (Count) for you, in the order they appear in your survey.
3

Adjust how it reads

Under the preview, choose what each cell shows (Percent, Count or Both), and turn the Color scale, Totals and Swap rows and columns on or off.
Percentages are always taken per statement, so each statement adds up to 100%. Hover a cell to see its count and the number of answers behind it. To build any other matrix, turn on Group data and pick two fields — the first becomes the rows, the second the columns — plus exactly one measure. A matrix shows up to 30 rows and 15 columns; add a filter to narrow a larger one down.

AI builder

Describe what you want in natural language (“Average NPS by month over the last 90 days”) and Formbricks generates the Cube query for you. You can edit the result in the manual builder afterwards. The AI builder requires Smart functionality (AI) to be enabled at the organization level and a configured AI provider. See AI Features.

Dashboards

Dashboards live under Analyze → Analysis, alongside charts. Each dashboard is a grid you can resize and arrange. From a dashboard you can:
  • Create a new chart inline and have it added automatically.
  • Add existing charts.
  • Reorder and resize widgets.
  • Duplicate a chart for a quick variation; the copy lands on the same dashboard.
  • Delete a widget (the underlying chart stays in the workspace).
A chart can sit on several dashboards, and editing it changes it on all of them. To change it on one dashboard only, duplicate it there and remove the original from that dashboard. A chart that is still on a dashboard can’t be deleted: the delete dialog lists the dashboards that hold it, so you can remove it from them first. If someone else saves the dashboard while you are editing it, saving tells you so. Choose Reload to see their version and drop your changes, or Save mine anyway to keep yours; the dialog names any charts they added that saving yours would remove.

Dashboard filters

Dashboard filters narrow all of a dashboard’s charts at once, for example to one region, one product, or one plan. They use the same Filter button, Advanced filter and filter chips as the responses page, and sit at the top of the dashboard next to the date range.

Add your own filters

1

Open the Filter menu

Open the dashboard and click the Filter button. Every field you can filter by is listed there, grouped and searchable. Fields with a few fixed options, such as sentiment or value band, can be ticked right there.
2

Set the filter

Pick any other field and its filter opens on a chip: the operator is chosen for you; change it if you need (is, is not, contains, is greater than, is after, is set, is not set, …). Then set the value: pick one or several from a list, type it, or choose a day, depending on the field. Click a chip to change its filter, including a ticked one, whose chip offers the other operators too.
3

Done

Click Done. The filter narrows every chart right away. To write several rules at once, use Advanced filter at the bottom of the menu.
A filter can use any field the charts can be grouped by, including embedded-data fields. An embedded-data field named like another field reads, for example, “Channel · Embedded data”, so the two are told apart. Answer filters are not offered here: they belong in each chart, in the chart builder.

Values to pick from

  • Fields whose values are looked up, such as the source name, the field label, or an embedded-data field, list the values that occur in the dashboard’s charts for is and is not, narrowed by the date range and the other filters. Search the list to find more; it shows up to 100 values at a time. For contains you type the text. Value (Option), Source ID and Ending store ids, so their lists show each option’s label, the survey’s name or the ending’s headline instead, and you can search by that text.
  • Fields with a fixed set of options, such as sentiment, value band or a yes/no field, list all of their options, greying out the ones the dashboard’s charts don’t have. A field with none at all is shown disabled; sentiment then reads “No sentiment yet: turn on AI enrichment”.
  • Free-text fields, such as the response text or the user ID, and number fields take a value you type, and date fields a day you pick. Free-text fields never list values, so the text of individual responses isn’t shown to everyone who opens the dashboard.
Is and is not on a field with a list take one value or several: the chip reads “Product is Alpha”, “Product is any of Alpha, Beta” or “Product is none of Alpha, Beta”. Lists also offer (No value): picked with values, is keeps records with any of them or with no value; is not keeps records that have a value and it is none of the others. For Language this option reads Default: a response without a language code was answered in its survey’s default language. Every field can also be is set (has a value) or is not set (has none). Is not and contains none of also keep records that have no value for that field: “Value (Number) is not 10” keeps text answers, which have no number. Use the date range to pick the period every chart shows. It replaces the period each chart was saved with, so you can look further back or ahead than the chart’s author did; on a chart grouped by time, All time keeps the chart’s own period. Your filters and date range are your own: they are remembered for this dashboard in your browser, and they don’t change what anyone else sees. They are also part of the page address, so a link you share opens the dashboard with the same filters.

Lock filters for everyone

An owner, manager or member who can edit the dashboard can lock filters for everyone who opens it:
1

Set the filters

Add the filters you want, as above.
2

Lock them

Click Lock for everyone. The filters are saved on the dashboard and show with a lock, before each person’s own.
A locked filter applies to every reader, Dashboard viewers included, and they can’t remove it; their own filters narrow further. To unlock a filter, an editor clicks the × on its chip; Unlock filters takes every locked filter off at once, after you confirm. A dashboard holds up to 10 locked filters. The date range is never locked.

Filters only narrow a chart

A dashboard filter is added on top of each chart’s own filters and never replaces them; the date range is the one exception, as above. A chart saved for products A and B shows nothing when you filter for product C, rather than showing product C. A looked-up value list only offers values that at least one chart on the dashboard can show. If a chart’s Feedback Dataset doesn’t have the field a filter uses, the chart can’t be narrowed by it. The chart then says “This chart can’t be filtered by” and names the filter, instead of showing data the filter would have left out. The same goes for Sentiment, Sentiment Score and Emotions on a chart that only shows answers without text, such as an NPS, CSAT or rating chart: AI enrichment only reads text answers, so these answers never carry them. Such a chart says “This chart can’t be filtered by Sentiment” rather than showing no data. Filters that also keep records without a value, such as is not, contains none of or is not set, still apply to it.
A filter’s value list shows the values of that field to everyone who can open the dashboard. A Dashboard viewer can filter only by embedded-data fields that a chart on the dashboard groups by, so they never see the values of a field like an email address unless a chart already shows it. Locked filters still apply to them on any field.
Charts and dashboards are also available through the MCP server, so an agent can build them for you and lock a dashboard’s filters.

Permissions

  • Owners, Managers, and members with Manage or Read & Write access can create and edit dashboards and charts.
  • Members with Read access can view dashboards and charts but cannot edit them.
  • Dashboard viewers can open the dashboards of the Workspaces their teams have access to and use their filters, and see nothing else. See Dashboard viewer.
  • Everyone who can open a dashboard can add their own filters, which only they see; only those who can edit it can lock filters for everyone or unlock them.

Requirements

  • A Feedback Dataset linked to your workspace.
  • At least one Feedback Source writing into it - a dataset with no source stays empty.
  • Workspace access to that dataset.
If Analysis shows “No feedback dataset linked”, start with Creating your first charts.