Google began rolling out a feature on October 19, 2026 that allows BigQuery data to be analyzed through conversations with Gemini directly in Google Sheets. Users can ask questions about connected data and request reports containing charts and pivot tables. The gradual rollout takes up to 15 days and applies to both Rapid Release and Scheduled Release domains.
The change concerns spreadsheets connected to BigQuery through Connected Sheets. According to Google, Gemini queries run against the full BigQuery dataset available through the connection, rather than only the rows visible in the spreadsheet. The data is processed in BigQuery, while query results are delivered in the Google Sheets interface.
From sales questions to connected reports
In the examples presented by Google, a person can request total sales in a region for a quarter and continue the conversation with a comparison against the same period in the previous year. Further questions can explore which products contributed to a change in revenue. These are examples of using an organization's data, rather than financial results obtained or validated by Google for a particular business.
To build a dashboard, Gemini first presents a plan for the user to review. It then creates a new tab with native connected objects: database formulas, pivot tables and charts. These components can be refreshed, including through schedules. The announcement therefore describes a report linked to its data source, rather than just a static chart image or explanatory text separate from the spreadsheet.
Connected Sheets already provided access to large volumes of information through familiar Sheets tools. Google Cloud documentation describes using data for business planning, comparing products sold in different locations, revenue reporting and analyzing customer requests. The October change adds conversational interaction with this data through Gemini, inside the same interface in which teams view and share reports.
A Gemini subscription and data access are separate requirements
The published availability list includes Business Standard and Business Plus, Enterprise Standard and Enterprise Plus, Google AI Pro and Ultra, and the Google AI Pro for Education and AI Expanded Access add-ons. The feature is enabled by default for eligible users in spreadsheets connected to BigQuery. Google specifies that there is no new administrative control dedicated to this feature.
Being enabled by default does not automatically grant database access. The user must have permission to query the connected data, and the organization must allow the necessary services. Connected Sheets configuration requires access to Google Cloud and BigQuery, and a project with billing configured. Permission to use Gemini and permission to read or query a dataset are administered separately.
Administrator documentation ties operations to BigQuery IAM roles and spreadsheet permissions. Creating connected objects and refreshing data require appropriate rights in both systems. A person can receive access to results already saved in a spreadsheet without automatically receiving permission to execute new queries. For supported organizations and editions, access delegation is a separate, administratively controlled option.
The connector also separates analysis from modifying the source. Google specifies that BigQuery data cannot be changed from within Google Sheets. Spreadsheets retain results for analysis and sharing, while direct access to tables and datasets is controlled in BigQuery. For infrastructure protected by VPC Service Controls, the perimeter must be configured to allow queries and the copying of results into spreadsheets.
Report updates follow Connected Sheets rules
BigQuery data is not continuously and automatically synchronized with every object in Connected Sheets. Results are updated through a manual refresh or according to a schedule. A chart or pivot table, an individual data source or all sources in a document can be refreshed separately. A connected report can therefore request new data from its source under the conditions established for the connection.
A scheduled refresh runs under the identity of the person who configured it. If another user adds or changes an existing data source, the schedule automatically pauses until it is resumed or taken over. Objects in a preview or failed state are not updated by a schedule. These existing rules apply to the reporting workflow on which the new connected objects also rely.
For custom query analysis, Connected Sheets retains its SQL editor and access to saved queries. The connector uses Google Standard SQL and supports parameters linked to cell values. Its documentation also describes checking the volume of data scanned before inserting results. The conversational interface complements these tools, while executions continue to use the BigQuery project associated with the connection.
Operations also leave a technical record. BigQuery access from Connected Sheets is recorded in Cloud Audit Logs, and administrators can identify the document through its spreadsheet ID. Administration documentation also describes Drive log events, distinguishing manually initiated queries, scheduled queries and those executed through the API or Apps Script. This information makes it possible to track the user and how the operation was initiated.
For eligible Google Workspace editions, Google states that Gemini respects existing permissions and that organizational data is not used to train generative models without the customer's permission or instruction. This statement applies to organizational accounts covered by Workspace agreements. The new feature is still rolling out, with access determined by an eligible subscription, a BigQuery connection and the user's rights to the data.
