05. Dimensions, metrics and aggregations

Dimensions, metrics and aggregations

Dimensions answer “according to who, what or when?”, metrics answer “how much?”. Aggregations combine both to produce meaningful results.

Open Dashboard Designer on the desktop Administration.

Your goal

You can classify fields as a dimension or measure, choose appropriate summaries, and review results at the right level of detail.

Why this is crucial in the dashboard designer

Every visualization is based on this decision. Sales by month requires month as dimension and sales as total; An average order value, on the other hand, requires a technically defined calculation.

Elements and settings at a glance

• Dimension: Groups non-aggregated values, such as month, partner, item group, or status.
• Key figure/measure: Is summarized, such as sum, number, average, minimum or maximum.
• Aggregation: Calculation rule for the key figure.
• Count: Counts lines; may contain multiple counts in joins.
• Count Distinction: Counts different values, for example unique receipts.
• Date grouping: Condenses dates by year, quarter, month, week or day.
• Discrete / Continuous: Affects axis, filters and temporal representation.
• Hidden Data Item: Supports sorting, filtering, or interaction without being visibly displayed.

Here's how to do it

1. Describe each required statement as a “measure by dimension.”
2. Bind dimensions and measures into the intended areas of the element.
3. Choose aggregation and date grouping consciously.
4. Check the result against a known X-ERP list or checksum.
5. Test detail level changes and empty groups.

Exercise

Create sales by month, number of unique receipts by receipt type, and average item value. Compare Count and Count Distinct on a data set with multiple locations.

Check the result

• Aggregation corresponds to the technical definition.
• Joins do not result in double counting.
• Date group and time period are unique.
• Units and signs are understandable.
• Total agrees with the X-ERP control evaluation.

Typical errors and their solutions

• Total price: First calculate the technical value for each process.
• Count for position data: Counts items instead of receipts; use Count Distinct if necessary.
• Date as text: Prevents clean time grouping and range filters.

Your result

Your key figures are not only visible, but also technically clear and mathematically reliable.

Further information / source

Related topics

1. For customizers – customize screens and forms › Dashboard designer › 04. Data sources, queries and data preview
2. For customizers – customize screens and forms › Dashboard designer › 07. Diagrams and time series
3. For Customizer – Customize screens and forms › Dashboard designer › 11. Calculated fields and expressions

Frequently asked questions