Salesforce to BigQuery: How to Move Your Data
Connect Salesforce to BigQuery with Airbyte. Sync standard and custom objects incrementally, work within daily API limits, and keep formula fields current.

Replicating Salesforce to BigQuery puts your CRM objects somewhere they can be joined to product usage, marketing spend and finance data, rather than queried one report at a time. The connector reads standard and custom objects, each as its own stream.
This guide covers the managed path with Airbyte. Two things shape the build: your Salesforce API allocation decides how much moves each day, and certain changes in Salesforce never reach the warehouse unless somebody intervenes.
Salesforce to BigQuery at a glance:
Why move data from Salesforce to BigQuery?
Because the questions worth asking about a CRM are rarely answerable inside it.
The good case is the join. Pipeline data beside product telemetry shows whether the accounts your team calls high intent actually use the thing. Opportunities beside marketing spend give you cost per closed deal rather than cost per lead. Salesforce reports cannot reach any of that, because the other half of each question lives somewhere else.
The poor case is reporting that stays entirely within Salesforce, where native reports and dashboards are already adequate and a pipeline adds moving parts for nothing. If you need an operational copy rather than an analytical one, Salesforce to PostgreSQL is less to run.
What do you need before you start?
Four things, and the last one is a conversation rather than a setting:
A dedicated read-only Salesforce user. That user's permissions decide exactly what the connector can see, so an object missing from your stream list is usually a permissions question rather than a connector limitation. Only queryable objects can be fetched.
Headroom in your daily API allocation. Salesforce caps daily requests by edition and user count, and this pipeline competes with every other integration you run. The Salesforce source documentation covers how the connector behaves when that cap is reached.
A shortlist of the objects you actually need. A mature org exposes hundreds of standard and custom objects, and every selected stream spends quota the objects you care about could have used. Start narrow and add.
An agreement with your Salesforce admins about formula fields. Editing a formula changes what the field means without changing any record, and the warehouse will not notice. Your admins have no reason to guess that, so tell them.
If your Salesforce org restricts traffic by IP, add the Airbyte Cloud IP addresses to the allow list before you begin.
How do you build a Salesforce to BigQuery pipeline in Airbyte?
Step 1: Prepare the Salesforce side
Create the read-only integration user and confirm it can see the objects you want. Check your current API usage in Setup under System Overview, so you know how much headroom the initial load has before it starts competing with your other integrations.
Step 2: Configure the Salesforce source
Click Sources in the left navigation, then New Source, and select Salesforce, following adding a source. Authenticate with OAuth and set a start date. Read that field carefully: the bound is exclusive, so a date-only value means midnight UTC and the named day itself is not included, and full refresh streams ignore it entirely.
Step 3: Configure the BigQuery destination
Click Destinations, then New Destination, and select BigQuery, following adding a destination. Supply a service account holding BigQuery User and BigQuery Data Editor. Batched Standard Inserts is the right loading method at CRM volumes. Set the dataset location to match your marketing and finance datasets, because BigQuery cannot query across physical locations and the location is fixed at creation.
Step 4: Create the connection and select streams
Click Connections, then New connection, select your shortlist of objects and a sync mode. Incremental append with deduplication on the record ID is right for most objects, so updates replace earlier versions rather than accumulating.
Then plan a downstream model, because the default partitioning will not serve the queries you run.
Why might a successful sync still be incomplete?
Because of how the connector handles your daily limit. It syncs until it reaches the cap, then stops, marks the job successful, and picks up where it left off next run. That is the right design, since the alternative is a failed job and a partial load you cannot resume.
It does mean a green sync is not a complete one. For an initial load of a large org, expect several days of partial syncs before BigQuery holds everything, with the sync history looking perfectly healthy throughout. A dashboard built on day two will be quietly short, and nothing will say so.
Monitor row counts against Salesforce rather than job status until the initial load settles. Watch which API is doing the work, too: the connector uses the Bulk API by default and falls back to REST for objects Bulk does not support or that carry unsupported column types, and REST-synced streams can cost more quota. An object quietly falling back is one reason an org runs out of requests sooner than expected.
What changes in Salesforce without changing your data?
Formula fields, and this is the one that surfaces months later. The connector syncs the output of a formula rather than the formula itself, which is correct, and editing that formula does not change any record's modification timestamp.
Incremental sync therefore never re-reads those records, and BigQuery goes on serving values calculated under the old formula indefinitely, with no error anywhere. The fix is to clear the affected stream and backfill it, which is why the agreement with your admins matters more than any setting in the connector.
A second, quieter transformation happens on the way out. Bulk API results arrive as CSV, which cannot distinguish an empty cell from the literal text inside it, so values such as NA, N/A, NULL, None and NaN become null by default. If a picklist in your org uses NA meaningfully, it disappears. The connector exposes a setting to preserve them, affecting Bulk streams only, since REST-synced streams keep them as they are.
Frequently asked questions
My sync succeeded but data is missing. Why?
The connector most likely hit your Salesforce daily API limit, ended early with a success status, and will resume next run. Check row counts rather than job status during an initial load.
A formula changed but BigQuery still shows old values.
Editing a formula does not update any record's modification timestamp, so incremental sync never re-reads those records. Clear the stream and backfill it.
Why did my NA values become nulls?
Bulk API results arrive as CSV, which cannot tell an empty cell from literal text, so NA, N/A, NULL, None and NaN become null by default. Enable the setting that preserves them. It affects Bulk streams only.
An object I need is not in the stream list.
Check the integration user's permissions first, since what Airbyte can read is decided by them. Only queryable objects can be fetched through the API.
Why are my queries scanning so much?
Tables are partitioned on the extraction timestamp, not on close date or created date, so filtering on a business date prunes nothing. Model a table partitioned on that date and point dashboards at it.
Get your Salesforce data into BigQuery
Name the join that justifies this, then select objects narrowly to protect your API quota. Monitor row counts rather than job status while the initial load settles, tell your Salesforce admins that a formula change needs a stream reset, and decide whether literal NA values should survive. Then model a table partitioned on the business date, because the default partitioning prunes nothing your analysts will filter on.
Airbyte's connector catalog includes 600+ pre-built connectors, so your CRM can sit beside the product and spend data that makes it answerable. For the same source into a warehouse with different economics, see Salesforce to Snowflake, and for a lakehouse, Salesforce to Databricks.
Integrate with 700+ apps using Airbyte
Move data from 700+ sources into warehouses, lakes, and beyond. Set up pipelines in minutes with pre-built connectors and the Connector Builder.
