Start from the CSV
This guide uses a handful of the sample's columns: team, level, location, cohort, group, tenure_band.
Template examples
How to design your Diversity and Inclusion dashboard, widget by widget.
Representation dashboards are almost entirely counts, sliced by one category at a time. Here’s how those cards come together, built using the real Diversity and Inclusion template.
The building blocks
Five widget types cover almost every dashboard. Each one needs a specific shape of column.
KPIOne numeric column, summed, averaged, or counted
team → EmployeesChartA date or category column, plus a measure
team + group → by teamMatrixTwo category columns and one measure
team + location + groupRankingA category column and a measure, ranked
- not used in this templateDetail tableWhichever columns people need to inspect
all 9 columns, sortable and filterableTwo ways to build every widget
Point and click
Add widget → pick a type → fill in Title, Value column, Calculation, and Format → Review → Add widget. Three steps, and every field has a default to start from.
Describe it
Type what you want in plain English. This is the faster path for anything the point-and-click form doesn't have a field for yet - like a second, smaller comparison number on a KPI.
Worked example: Diversity and Inclusion
Count every row: Employees
Add widget → KPI. Value column: team. Calculation: count. Format: number. No filter - a bare count is the total row count.
Count a subset: Women
Add widget → KPI. Same column, same calculation - but add a filter: group is Women. The filter is the only thing that turns a total into a subset.
See headcount across two categories
Add widget → Matrix. Rows: team. Columns: location. Value: group. Calculation: count. Format: number. This dataset has no ranking widget - there’s no single "top N by one measure" question in the original template, so it isn’t forced in here either.
Common mistakes to avoid
- Distinct counts unique values, count counts rows."Teams" uses distinct on team to count how many different teams exist; "Employees" uses count to count every row. Mixing the two up gives a very different number.
- Count adds up rows, not a column."Women" isn’t a count of the group column - it’s a count of rows where group is Women.
- A stacked bar and a Matrix solve different questions."Representation by Team" splits one axis (team) by a second category (group) in a chart. "Headcount: Team × Location" cross-tabs two different dimensions in a grid. They can look similar but aren’t interchangeable.
See it built