The Lab Notes

Google

Gemini in Sheets now answers questions over full BigQuery tables

Google is rolling out natural-language analysis of BigQuery data in Gemini in Google Sheets, on by default with no admin control.

A brass two-drawer filing cabinet with a top handle, its lower drawer pulled open by a white cartoon hand while another hand grips the handle, with a wavy blue and purple ribbon looping around it, set against a blue background of pink, light blue, yellow and red quarter-circle shapes.

Gemini in Google Sheets can now answer plain-language questions about BigQuery data. Its queries run against the whole dataset rather than the rows on screen. The feature is on by default. Admins have no setting to turn it off.

Google began rolling it out on its Workspace Updates blog on Oct. 19, 2026. It works in spreadsheets connected to BigQuery through Connected Sheets. Users open the Gemini side panel and ask questions. Gemini answers conversationally.

Google Showed Follow-Up Questions

Google’s examples pair a first question with two follow-ups in the side panel. Google said users can also build views over very large datasets.

  • “What were total sales in the Northeast last quarter?”
  • “How does that compare to the same quarter last year?”
  • “Which products drove the increase?”

Gemini Builds Dashboards From a Description

Gemini shares a plan for the user to review, then builds a new spreadsheet tab of refreshable objects. Google’s example request was “Build a quarterly sales review with revenue by region, a monthly trend, and our top 10 products.” The tab holds database formulas, pivot tables and charts. These are native and connected. Google said they stay current with scheduled refreshes.

Google said data teams should see a shorter backlog of one-off requests for reports and data pulls from BigQuery.

Availability Follows the Query Permission

The feature is available by default in connected spreadsheets for users with permission to query the connected data. Rapid Release and Scheduled Release domains get a gradual rollout that began Oct. 19, 2026. Feature visibility can take up to 15 days.

It is rolling out to these plans:

  • Business Standard and Plus
  • Enterprise Standard and Plus
  • Google AI Pro and Ultra for consumers
  • Google AI Pro for Education
  • AI Expanded Access

Our View

We think the full-table reach matters most to analysts. A business user who asks Gemini about last quarter’s sales no longer needs someone to pull the rows first. Those pulls are the backlog Google expects to shrink.

Analysts who still get dashboard requests have one place to catch a misreading. It is the plan Gemini shares before it builds the tab. We would treat reading that plan as part of the work.

For admins the lever is who holds query permission on the connected data. We would review that list before the rollout reaches the domain.

The shorter backlog holds only if users read the plan before approving it and permissions match who should see which data.