Analytics blocks let you create charts that turn model data into insights. You can choose chart types such as area, bar, line, or pie, and configure them to display counts, totals, averages, or other aggregations. This helps you monitor trends, compare values, and take action based on data.
Combining values in charts
You can aggregate your data in different ways. The table below shows the options:
| Aggregation type | What it does | Example use case |
|---|---|---|
| Count | Counts the number of values with the same property. | Show how many employees are in each job role. |
| Sum | Adds all the values for the selected property. | Show the total salary spend for each department. |
| Minimum | Shows only the lowest value for the property. | Find the lowest salary in each job role. |
| Maximum | Shows only the highest value for the property. | Find the highest salary in each job role. |
| Average | Calculates the average value for the property. | Compare the average salary across job roles. |
Detailed walkthroughs
Each aggregation can be selected in the Properties list, with the property name followed by - Count, - Sum, - minimum, - maximum, or - average.
Count
Counts how many values exist for each property.
Example: a pie chart showing the number of employees in each job role.
Sum
Adds together the total values for a property.
Example: total salaries for each job role.
Minimum
Displays only the lowest value for each group.
Example: lowest salary in each job role.
Maximum
Displays only the highest value for each group.
Example: the highest salary in each job role.
Average
Calculates the average value.
Example: average salary for each job role.
Best practices
- Use Count for distribution insights, for example, headcount by role.
- Use Sum for totals, for example, salary spend, hours worked.
- Use Min/Max to spot outliers or benchmarks.
- Use Average for comparisons, for example, pay across departments.
- Keep charts focused and straightforward, and avoid adding too many aggregations in one view.
Task you can perform here:
Create an analytic block with a chart