Summarize — grouped totals
Group by one column and aggregate others into a new table. Ten aggregate operations, and why the result is a live table rather than a snapshot.
The panel
Sumif in the toolbar (the panel is titled Summarize — aggregate table data and produce a new table), or Summarize Table on the Tools page:
- Choose or upload the table to summarize.
- Choose the column to group by — month, department, class, product. Several columns can be combined into one grouping.
- Summarize these columns into the new table — each row picks a column and an aggregate; Add another adds more. Ticking All columns applies the same aggregate to every numeric column.
- Confirm. A new table is created and opened.
Ask AI at the top reads the headers and proposes several plausible groupings and aggregates; pick one and the panel fills in.
The aggregates
| Operation | Gives | Function |
|---|---|---|
| Sum | The group total | sumif |
| Count | How many rows in the group | countif |
| Distinct count | How many different values (how many customers, how many products) | distinctcountif |
| Max / Min | The group's largest and smallest | maxif / minif |
| Average | The group mean | average |
| Median, lower quartile, upper quartile | Where the distribution sits | median / lowerquartile / upperquartile |
| Standard deviation | How spread out it is | stdevsif |
| Latest | The newest row's value in that group — works on text too, e.g. each customer's most recent contact | — |
The result is a live table
The summary's first column holds the grouping values; every column after it is an aggregate formula. Add rows or change numbers in the source and the summary updates; a new grouping value in the source (a new month, a new department) adds a row here. You can keep adding columns to it (Average order = Booked / Orders), reference it from a third table, put it on a dashboard, and export it. The mechanism is explained in growing a summary table.
Examples
| What you want | Group by | Aggregate |
|---|---|---|
| Revenue and order count per month | Month | Amount · Sum, Order number · Count |
| Headcount and average salary per department | Department | Employee id · Count, Salary · Average |
| Class average, top mark, number passing | Class | Score · Average, Score · Max, Passed? · Count (make a "Passed?" column with if first) |
| Each customer's most recent contact | Customer | Contact · Latest |
Against a pivot view
A pivot view is for looking; Summarize produces a table. When you need to add custom columns to the result, reference it, or export it, use Summarize. For a quick glance at a distribution, use a pivot view.