Here's everything new and improved in Golden Analytics this week.
You can now share any dashboard directly to a Slack channel. Connect your Slack workspace from Settings, then use the "Share to Slack" button on any dashboard to post a Block Kit message with the dashboard title, an optional note, and a live thumbnail. The feature respects existing share settings and creates a public link automatically.
Dashboard chat can now build out your dashboard layout directly. Ask it to pull an existing chart from one of your analyses onto the current dashboard, add or edit a text block, or place a shape (rectangle, circle, line, arrow) with fill color, border, opacity, and optional inside-text. New elements auto-place into empty space instead of stacking on top of each other.
Chat now sees the full canvas layout and can act on it intentionally. You can ask it to place elements at specific coordinates, position things relative to each other ("put a caption below the revenue chart", "add a title spanning the top"), bring items to the front or send them behind everything, resize the canvas, or switch to infinite-scroll mode. Previously, chat-placed elements always landed in the first available slot.
The Export dialog now has two tabs — Slides and Data. From the Data tab, export chart data from any dashboard as CSV (one file per chart), an Excel workbook (one sheet per chart), or directly to a new Google Sheets spreadsheet. All existing export-to-slides options remain in the Slides tab.
BigQuery is no longer gated behind a preview flag and now appears in the connector picker for all Golden Analytics accounts alongside Snowflake and Databricks.
⌘K → "View SQL" in the dashboard view opens a panel showing the fully-resolved SQL for every data-backed chart on the current sheet, including chart-level filters, cross-filters, and active filter bar selections. The dialog lists all data-backed charts in a sidebar, shows syntax-highlighted SQL, adapts to the database dialect, and includes a copy button.
Running total, % of total, moving average, rank, and difference calculations — previously only available in the Analyze canvas — now work on dashboard charts. Set them from the measure pill's modifier strip on any dashboard chart, and they survive copy-to-dashboard and sheet clones.
Fields can now be hidden from field list panes across Analyze, Dashboards, and Stories. Hidden fields show an eye-off indicator and a "Show hidden fields" toggle controls visibility in the Prep data table. The AI no longer sees hidden fields in its context, and formula dependencies are tracked so cascaded deletes work correctly.
Prep chat can now create, rename, edit formulas, and delete calculated fields end-to-end. It can also set display formats on any measure column (currency, percentage, duration), edit default aggregation, descriptions, date format, and geographic subtype in a single turn. Ask it to "enhance all descriptions" and it rewrites every field's description in one pass.