Google Ads to Enforce 37 Month Data Retention Policy from June 2026

Starting June 1, 2026, Google Ads and related measurement APIs will enforce a 37-month data retention policy for granular performance statistics such as daily, hourly, and weekly data. However, high-level aggregated data (monthly, quarterly, yearly) will remain accessible for up to 11 years.

API Impact and Next Steps

Google Ads API and Google Ads scripts

• Queries requesting granular segments (e.g., segments.date, segments.week) for data older than 37 months will return errors like DateRangeError.INVALID_DATE or, in future versions, DateRangeError.REQUESTED_DATE_GRANULARITY_NOT_SUPPORTED.
• To access data beyond 37 months, queries must use coarser segments such as segments.month, segments.quarter, or segments.year. Unsegmented queries must align exactly with calendar month boundaries.
• Users should review and update their applications accordingly and export any needed granular historical data before the June 1, 2026 deadline.
• Support is available via Google Ads API and Google Ads scripts support channels.

Google Analytics Data API

• Metrics like Advertiser Ad Cost, Clicks, and Impressions will be truncated to the latest 36-month window if the report includes the "date" dimension or equivalent date dimensions (e.g., "Day of week", "day").
• Only reports with all affected metrics, older than 37 months, and including the "date" dimension are impacted.
• Users should adjust their data stitching logic to accommodate this truncation.

DV360 API and CM360 API

• These APIs maintain a 24-month retention period, which remains unchanged, so there is no impact.

BigQuery Data Transfer Service

• Starting June 1, 2026, the Google Ads connector will stop backfilling data older than 37 months, but existing data in BigQuery tables will remain unaffected.
• Similarly, the Google Analytics 4 connector will stop populating data for backfill runs older than 37 months. If a manual transfer is triggered for such dates, the data in BigQuery will be overwritten with empty values.
• To retain historical data beyond 37 months in BigQuery, users should start backfilling early to complete before the deadline.