As a tool to help you operate more effectively amidst privacy checks in Ads Data Hub, the platform includes a synthetic dataset. Unlike normal data, this dataset isn’t subject to privacy checks, and as such should be used for experimentation and query development. The sandbox can improve your query development workflow by:
- Running jobs faster. Because queries run on sandbox data aren't subject to privacy checks, your jobs will run faster.
- Allowing you to inspect the underlying data. You can use
SELECT *on sandbox data, giving you a better understanding of how your results could look.
- Avoiding differential privacy checks. Differential privacy checks look at your query history, but aren't impacted by queries run on sandbox data. Learn more about differential privacy checks
These improvements allow you to iterate on draft queries faster and can reduce the frequency of errors due to privacy checks. As such, Ads Data Hub recommends that you use sandbox data to develop queries whenever possible.
Sandbox data and privacy restrictions
Because the sandbox data is synthetic, Ads Data Hub relaxes certain privacy restrictions. For example, the following actions are only possible when querying sandbox data:
- Running queries using
- Running a series of similar queries. This is a common practice for developing SQL queries, but doing so without using sandbox data makes it more likely that your jobs will run privacy-check issues. Jobs run on sandbox data won't impose data difference checks, allowing you to more rapidly iterate on SQL. Learn more about privacy checks in Ads Data Hub
Sandbox query workflow
The workflow for developing queries that use sandbox data is nearly identical to developing queries that use your own data. The only differences are:
- When querying sandbox data, you can write queries that use
- When running queries on sandbox data, you must:
- Select start and end dates between 2018-08-18 00:00:00 and 2018-09-17 23:59:59.
- Select ADH Sandbox Customer from the Ads data from dropdown.