pbPassingBI
/

Fields, types and data freshness

Dimensions versus metrics, aggregation defaults, and how caching works.

What you'll be able to do
  • Set field types and aggregation
  • Explain the difference between a dimension and a metric
  • Control data freshness and caching

Dimensions and metrics

Dimensions are green and describe things — country, campaign, date, page path. Metrics are blue and measure them — sessions, revenue, clicks.

A chart is essentially a choice of dimensions to break down by and metrics to display. Get that pairing right and the chart usually follows.

Types and aggregation

Each field has a type (text, number, date, currency, geo, URL) and metrics have a default aggregation (sum, average, count, count distinct, min, max).

Edit them in the data source, not per chart. Fixing the currency format once in the data source beats fixing it on nine charts.

Numeric IDs treated as metrics

An order ID or postcode arrives as a number, so Looker Studio makes it a metric and cheerfully sums it. Change the type to text and its role to dimension in the data source.

Changing a field is not a rename

Changing a field's aggregation in a chart affects only that chart. Changing it in the data source affects everything using it. When something looks wrong in several places, the fix belongs in the data source.

Data freshness

Looker Studio caches results. Resource → Manage data sources → Data freshness sets how long, and the options depend on the connector — 15 minutes for BigQuery, 12 hours for some others.

The Refresh data button in the toolbar clears the cache for the current report immediately, which is the answer when someone says the numbers look stale.

Shorter freshness means more queries. On BigQuery that is a direct cost, so match the interval to how often the data genuinely changes rather than setting it to the minimum.

Key points
  • Fix field types and aggregation in the data source, not per chart
  • Numeric identifiers get treated as metrics — change them to text dimensions
  • Data freshness controls caching; shorter intervals cost more queries
Check yourself