Skip to main content
Detrics fetches data live from each platform’s API, so a query reflects what the platform reports at that moment. The platforms themselves, however, keep adjusting recent numbers. Understanding when data is still settling saves a lot of “why don’t these numbers match” confusion.

​
Why Recent Numbers Change

  • Attribution windows. Ad platforms credit conversions to clicks and views that happened days earlier. A purchase today can be attributed to a click from last week, so last week’s conversion numbers grow after the fact.
  • Processing delays. Analytics platforms ingest and process events for hours before totals stabilize. The most recent day is almost always incomplete.
  • Restatements. Platforms remove invalid traffic and correct errors retroactively, which can also revise numbers downward.
The practical consequence: the same query over a recent date range can return different numbers today and tomorrow, and both are correct snapshots of what the platform reported.

​
When Data Becomes Final

Each platform has a window after which historical data stops changing, ranging from a couple of days for analytics sources to roughly a month for ad platforms with long attribution windows. Once a date range falls entirely outside that window, the numbers are stable and queries over it are fully reproducible. If you need exact period totals for reporting, query the period a few days after it closes, or schedule a refresh so the final numbers overwrite the early snapshot.

​
Comparing Detrics with the Platform’s UI

When a Detrics number differs from what you see in the platform’s own dashboard, check these in order:
  1. Date range: does the platform UI include today the same way your query does?
  2. Timezone: Detrics interprets dates in UTC; the platform UI uses the account’s timezone
  3. Attribution settings: the UI may display a different attribution window than the API default
  4. Metric definition: similarly named metrics can measure different things (for example, “views” vs. “plays” on video metrics)
The per-platform troubleshooting pages in the Fields Catalog cover known discrepancies for each source.

​
Keeping Reports Fresh

For dashboards and recurring reports, schedule automatic refreshes instead of pulling data manually. The scheduler re-runs your query, so settling numbers update on their own.

Automate Queries in Sheets

Schedule refreshes for spreadsheet reports

BigQuery Sync Scheduling

Refresh windows for warehouse syncs