Analysis
Secondary dimension
Also called breakdown dimension.
Secondary dimension is a second attribute added to a report so each row becomes a pair, such as page and country. It cross-tabulates without making a new report.
How it is measured
The report keeps its primary dimension and metrics and adds a second column of labels. Rows multiply: the row count is roughly primary values times secondary values that actually occur.
Keep the count of combinations manageable. A pair of two high-cardinality dimensions produces a table too sparse to read, and rows with tiny counts need caution.
Worked example
A tea importer opens landing pages by source and wants to know whether the Japan sencha page behaves differently for visitors from search and from email. Adding 'source' to the page report gives 4 rows for that page.
The sencha page gets 880 search sessions and 160 email sessions, but the email group has twice the checkout rate. The importer shifts newsletter space toward that page.
How it differs
A secondary dimension adds a second split to a report. A dimension is any single attribute. A primary dimension alone gives one row per value.
Common errors
Pairing two huge dimensions. Reading a row of five sessions as a pattern. Treating the order of dimensions as irrelevant. Forgetting the metric is totals across the pair. Exporting without the pairs.
In practice
When a total hides a difference, add one secondary dimension and check the biggest rows. Do not trust any row with fewer than a few dozen sessions.