Video | Tableau | Data visualisation | Tool strategy

Tableau Workbook Optimiser: New in Tableau 2022.1

Tableau has finally codified all that workbook best-practice folklore into a built-in optimizer, and it's a huge win.

Part ofWhat's new in Tableau 2022.1
Watch on YouTube
  • The Workbook Optimizer appears at the bottom of the publish dialog and runs around 12 best-practice tests without altering your workbook, only offering guidance.
  • Failed tests flag issues like unused fields across data sources and calculations using multiple data sources, each with an explanation and a link to support guidance.
  • Removing or hiding unused fields cuts unnecessary load and compute time when a workbook is published to production.
  • You can rerun the optimizer after making changes and it retains a timestamped score from your last run, acting like a proactive to-do list.
  • The feature will also be available in WebEdit in the browser, and Tableau has committed to expanding the tests over time.

Tableau 2022.1 adds a built-in Workbook Optimizer that runs automated best-practice checks at publish time, turning informal performance folklore into a concrete, explained checklist.

When publishing a workbook to Tableau Server or Online, a new Workbook Optimizer option appears at the bottom of the publish dialog. It runs a set of tests against the workbook and reports pass/fail results without changing anything itself.

The Breakdown
  • Find the optimizer in the publish dialog 0:00

    When you publish a workbook, a Workbook Optimizer option sits at the bottom of the usual publish interface. Opening it runs around 12 best-practice tests and shows a pass/fail score, without altering the workbook itself.

  • Read the failed and amber results 1:41

    Each failed or flagged test is explained in plain terms, for example unused fields sitting across data sources or calculations that blend multiple data sources. Amber doesn't mean wrong, just a potential performance cost worth weighing, such as query slowdown from blending.

  • Act on unused fields to cut load 2:01

    Fields loaded but never used in the workbook still get pulled in at runtime, adding unnecessary compute time. The optimizer names the specific unused fields so you can remove or hide them from the extract before publishing to production.

  • Use the passed checklist as a build habit 3:54

    The list of passed tests covers things like number of data sources, visible and hidden sheets, calculation length, LOD count, and non-materialised filter columns. Even without deep optimisation knowledge, reviewing this list before every publish nudges you toward good habits as you build, not just at the end.

  • Rerun after fixes, score is timestamped 6:53

    You can make changes and rerun the optimizer from the same dialog; it keeps a timestamped score from the last run rather than recalculating automatically. This makes it usable as a proactive to-do list you work through and revisit.

  • Finish the publish cycle as normal 7:42

    Once you're satisfied with the results, the ordinary publish step still sits alongside the optimizer, so you close out and publish as you normally would. Tim notes the navigation between optimizer and publish dialog feels a bit clunky, but the workflow still completes fine.

  • Coming to WebEdit 8:41

    The same optimizer capability is planned for WebEdit in the browser, giving the same checks without needing Desktop.

Worth Knowing
  • The optimizer only advises; it never changes the workbook for you, so acting on the findings is manual.
  • Support-guidance links behind each test weren't live yet at recording, since this was captured ahead of the official release.
  • The red/amber/green colour scheme is inconsistent with Tableau's usual blue/orange styling and may be harder to read for colour-blind users.
  • Closing the optimizer dialog drops you out of the publish flow rather than back to the publish dialog, which feels like an extra hop.
Use It When

Reach for this every time you publish a workbook to production, especially before handing workbooks to server admins or standardising practice across a team, since it surfaces concrete, explained fixes rather than relying on remembered best-practice folklore.

How this Rollup was made provenance & method

A Rollup is drafted by AI from the video's transcript, then reviewed and edited by Tim. Everything used to produce this one is listed below — the model, the exact prompt, and the source video — so the process is transparent and reproducible.

Transcription
On-device — NVIDIA Parakeet v3 for recent videos, OpenAI Whisper large-v3 for earlier ones. The transcript never leaves the machine or gets published.
Drafting
Claude Sonnet 5 in the cloud, from that transcript.
Prompt
The exact Rollup prompt (v2) — the full system prompt, unedited.
Source video
Watch on YouTube
Drafted
5 July 2026 at 09:38
Reviewed & edited
5 July 2026 at 09:41 · by Tim Ngwena

Model + prompt + video is everything you'd need to recreate a Rollup like this yourself. The one thing we don't share is the transcript.

Rights. The video and its transcript are the property of TN Media Ltd. Unauthorised use or download is prohibited. © TN Media Ltd.