PDE - Preparing and Using Data for Analysis (15% of the exam) - Section 4.3

Share data by defining sharing rules, publishing datasets and reports, and using Analytics Hub for BigQuery data sharing.

Define sharing rules to publish BigQuery datasets and reports to internal or external consumers using Analytics Hub listings, and recognise when data clean rooms provide privacy-preserving collaboration without exposing the underlying rows.

Analytics HubBigQuery datasetsData clean rooms

Practice question for this objective

Free samplePreparing and Using Data for Analysiseasy

A retail analytics team in London wants to publish a curated BigQuery dataset of weekly product sales aggregates so that partner suppliers in separate Google Cloud organisations can query the data from their own projects without the analytics team having to add each partner as a principal on the source dataset. The analytics team will continue to update the underlying tables, and partners should always see the latest data when they query. Which approach in BigQuery best supports this sharing pattern?

AExport the weekly aggregates to Cloud Storage as Parquet files in a partner-readable bucket and ask each partner to load the files into their own BigQuery project.
BGrant each partner principal the BigQuery Data Viewer role directly on the source dataset in the publisher's project so they can query the tables in place.
CPublish the curated dataset as a listing in Analytics Hub, then have each partner subscribe to the listing so a linked read-only dataset appears in their own project. Correct
DCreate authorised views over the aggregate tables in the publisher's project and grant each partner the BigQuery Job User role in that same project so they can run queries.

Recognise Analytics Hub listings as the BigQuery mechanism for sharing curated datasets with external subscribers who query live data from their own projects. Analytics Hub lets a publisher expose a BigQuery dataset as a listing in an exchange. Subscribers add the listing to their own project as a linked dataset, which is a read-only symbolic reference back to the source. Queries run in the subscriber's project against the latest source data, so the publisher avoids managing per-partner IAM on production tables and partners do not need their own copy of the data.

Why A is wrong: Tempting because Cloud Storage is a common interchange layer, but it forces every partner to manage ingestion, breaks the live-query expectation, and does not give the publisher a managed sharing surface in BigQuery.

Why B is wrong: This works for a small number of internal users but does not scale to many external partners across organisations and requires the publisher to maintain a long IAM policy on the production dataset.

Why C is correct: Analytics Hub is the BigQuery feature designed for publishing curated datasets as listings that subscribers consume as linked datasets, with the publisher controlling the source and subscribers querying live data in their own projects.

Why D is wrong: Authorised views can expose subsets of data, but giving external partners job-running permissions in the publisher's project mixes billing and access, and still lacks the managed listing model that Analytics Hub provides.

See more PDE practice questions, answers explained.

Exam traps in Preparing and Using Data for Analysis

Answers that look right on this material and are not. Each one is a distractor from a different question in the PDE bank for this domain.

BigQuery authorised views over a single combined dataset in the healthcare group's project, with the pharmaceutical team added as principals with the Data Viewer role.
Why it is wrong: Authorised views can hide columns or rows from one party, but they assume a single owner project and do not provide the multi-party governance model where each contributor controls its own contributed data.
They schedule when contributed tables are refreshed inside the clean room, so each party knows how stale the shared data may be at query time.
Why it is wrong: Refresh scheduling is handled by the contributing party's own pipelines and BigQuery features, not by analysis rules. Analysis rules are about governance, not freshness.
Each company publishes a standard Analytics Hub listing of its raw patient data so that the other side can read row-level records and write join queries directly.
Why it is wrong: A standard listing exposes the underlying rows to subscribers, which directly violates the rule that neither side may see the other's row-level records.

Examworthy is not affiliated with or endorsed by Google Cloud. Original, blueprint-aligned practice material only.