Hello everyone! I’m trying to move data from GA4 to MotherDuck using the documentation on the marketplace connector on Airbyte Cloud. The connectors (both for GA4 and MotherDuck) authenticate accurately and the data loads, but majority of the columns show up as

Null

. Has anyone faced something like this before? What might be the issue? And how can I resolve this? Please let me know if you need any additional context or screenshots on my end to resolve this correctly.

A few different Airbyte issues and forum threads describe “columns all NULL” symptoms that sound similar to what you’re seeing, but for other connectors (AppsFlyer, BambooHR, GA4/GA, etc.). The root causes there fall into three main buckets: 1. Out-of-date or incorrect source schema - Example: the AppsFlyer source used

country_code

in its schema while the API actually returned

country

, so the destination columns stayed NULL even though data was present in the raw API payload.[\[Appsflyer schema issue\](https://github.com/airbytehq/airbyte/issues/60995)] - A similar pattern is reported for BambooHR’s

custom-reports-stream

, where only one field (employee ID) had values and the rest were NULL.[\[BambooHR issue\](https://github.com/airbytehq/airbyte/issues/55890)] For GA4 this could manifest if the connector’s schema doesn’t match what the API returns for the dimensions/metrics you’re using (especially if you’re using custom or newly added fields). Unfortunately, the provided GA4 docs and issues don’t list a specific schema bug like the AppsFlyer one, but they do show that outdated connector versions can cause data misalignment and NULL-like behavior in Google Analytics connectors, and that upgrading often resolves it.[\[GA4 data misalignment\](https://github.com/airbytehq/airbyte/issues/49946)] 2. Custom report configuration quirks - For the GA4 Data API source, users have run into problems where custom reports didn’t behave as expected until they adjusted the JSON syntax (e.g., wrapping the whole custom reports array in brackets

[...]

). That fixed a “not working like in 0.0.2” issue when moving to a newer custom-report version.[\[GA4 custom reports\](https://discuss.airbyte.io/t/2351)] - In Universal Analytics, one user saw many custom-report columns come through as NULL; after a full reset and rerun, it started working, suggesting either a transient issue or bad initial catalog/schema that was corrected when they reset.[\[GA custom report NULLs\](https://discuss.airbyte.io/t/1153)] 3. Destination behavior / new destination implementation (MotherDuck) - The MotherDuck destination is a Destinations V2 connector (a newer implementation of the DuckDB destination) with different final table structures and typing/deduping logic.[\[MotherDuck overview\](https://docs.airbyte.com/integrations/destinations/motherduck)] - Its changelog shows a lot of bug fixes around table creation, column naming, handling special characters, and primary key behavior; these types of issues can manifest as mismatched columns vs. incoming records if you’re on an older version.[\[MotherDuck changelog\](https://docs.airbyte.com/integrations/destinations/motherduck#changelog)] Given what we do have in the knowledge sources, here’s what you can concretely try: 1. Confirm connector versions and, if possible, upgrade both GA4 and MotherDuck to the latest versions - There is a prior GA4 issue where stale connector versions caused data misalignment and the recommendation was simply “please upgrade to the latest and, if it persists, open a new issue.”[\[GA4 data misalignment\](https://github.com/airbytehq/airbyte/issues/49946)] - MotherDuck has had several bugfix releases in late 2024–2025 addressing table/column alignment and write issues.[\[MotherDuck changelog\](https://docs.airbyte.com/integrations/destinations/motherduck#changelog)] If you’re not already on the latest GA4 source and MotherDuck destination in Airbyte Cloud, upgrading is the first step. 2. Check whether the NULLs are only in the final tables or also in the raw tables - The MotherDuck destination stores records with at least

_airbyte_raw_id

,

_airbyte_extracted_at

, and

_airbyte_meta

, plus any columns declared in the JSON schema.[\[MotherDuck schema\](https://docs.airbyte.com/integrations/destinations/motherduck)] - In MotherDuck, inspect both: - The “raw” tables (often prefixed with

_airbyte_raw_...

or containing

_airbyte_meta

) to see if the GA4 fields are present in the JSON/blob. - The typed/final tables (without

_raw_

in the name) where you see the NULLs. - If the raw blobs contain the data but the typed columns are NULL, this usually indicates a schema mismatch or typing/deduping problem (similar in spirit to the AppsFlyer and BambooHR issues).[\[Appsflyer schema issue\](https://github.com/airbytehq/airbyte/issues/60995);\[BambooHR issue\](https://github.com/airbytehq/airbyte/issues/55890)] 3. If you’re using GA4 custom reports, verify the custom report JSON format - Ensure your custom report definition is a valid JSON array, and if you followed older docs, make sure you kept the outer

[...]

brackets; users found missing brackets caused unexpected behavior in newer custom-report versions.[\[GA4 custom reports\](https://discuss.airbyte.io/t/2351)] 4. Reset the connection and run a full sync once - In at least one Google Analytics custom-report case, all-NULL columns started working correctly after a reset and fresh sync.[\[GA custom report NULLs\](https://discuss.airbyte.io/t/1153)] - This can regenerate the catalog and tables according to the current connector versions, which may correct mismatches that cause NULLs. 5. If the issue persists, capture details and open a ticket Since the knowledge sources don’t document a specific “GA4 → MotherDuck columns all NULL” bug, if you still see the problem after: - Updating GA4 and MotherDuck connectors - Verifying custom reports JSON (if applicable) - Confirming data exists in raw tables but not in typed tables - Resetting and re-running a sync then it’s likely a new bug in either the GA4 source or MotherDuck destination. At that point, I’d recommend opening a GitHub issue (or Airbyte Cloud support ticket) with: - Exact GA4 source & MotherDuck destination versions - Whether you’re using standard GA4 streams or custom reports - Screenshots or queries showing: - One affected field in the raw table (with data in

_airbyte_meta

or

_airbyte_data

) - The same field as NULL in the final table - Any relevant sync logs That evidence closely matches how similar issues (AppsFlyer schema mismatch, BambooHR custom reports) have been reported and fixed in the past.[\[Appsflyer schema issue\](https://github.com/airbytehq/airbyte/issues/60995);\[BambooHR issue\](https://github.com/airbytehq/airbyte/issues/55890)] If you’re comfortable sharing more detail here, the most useful next things would be: - The GA4 connector version and MotherDuck connector version from Airbyte Cloud - Whether the NULL columns are GA4 dimensions, metrics, or both - A quick sample query against one raw MotherDuck table showing that the underlying JSON/meta actually contains the values that appear as NULL in the final table.