Tableau 9.3 New Features

Tableau has speeded up its release cycles from one release per year to three releases 2015. Tableau also announced that there would be four releases in 2016. Tableau is going to spend more R&D $ this year than all the last 13 years of the company combined. I love the pace of innovation.

Tableau 9.3 is released on 3/24. I was able to demo some 9.3 new features at Tableau server upgrade and new feature demo webinar on the day when 9.3 was released, which is cool.

I am excited for Tableau 9.3 release, which features powerful upgrades to Self-Service Analytics environment. These include Workbook Revision History, union Excel or text-based data sources, passing parameters in initial SQL, Snowflake data connector, Map enhancements, Content Analytics, etc.

Workbook Revision History

This is the feature that many Tableau fans have been waiting for long time. In the past, publishers had to manage their own workbook versioning, which is a difficult task for many publishers. When changes did not work out and had to rollback to previous version, sometime publishers had challenges to remember what was the right version before. Unfortunately Tableau server team was helpless. Now 9.3 server keeps published workbook revision history so that publishers can go back to any of the their previous version if changes did not work out. This is huge!

Union & More Data Prep Features

Data prep is the area where most analysts spend a lot of their time unfortunately. Tableau continues enhancing data prep features so analysts can spend their valuable time on analysis and insights vs. copy & paste the data. 9.2 released feature of sub-table detection, data grid editing, data pan searching, etc. 9.3 added union feature that combines data that have been split across multiple files or tables into a single Tableau data source. Union works for Excel or text-based data sources only. I am sure that Tableau will make union work for database tables as well. You can also do more data grid editing now with 9.3: preview data extract or Web Data Connector, creating group or bin, etc.

Parameters in Initial SQL for Row-Level Security

This is huge feature for customers who are looking for better row-level security solution. Initial SQL is a set of commands that can run when you open the workbook, refresh an extract, sign in to Tableau Server, or publish to Tableau Server. Initial SQL can be used to set up temporary tables or a custom data environment during the sessions. Initial SQL is not new but was missing a critical feature – you could not dynamically pass parameters like username. Tableau 9.3 is able to pass parameters (TableauServerUser, etc) to some database. When TableauServerUser as parameter is passed to database for the duration of that user session, you can leverage database’s user security mapping (if you have implemented it) so database will render user specific data only to achieve the row-level security. 9.3 parameter in initial SQL supports Oracle, SQL Server, Sybase ASE, Redshift, and Greenplum only. Click here for details. For Teradata, you can use query band to pass parameters to achieve row-level security.

Project Leader Content Management

I have a blog about how to use Tableau site. I know that many Tableau customers avoid creating a new site unless they have to. How to make sure that site admins not becoming a bottleneck when you scale out Tableau but your deployment has only one or very few sites? If you struggle with this, you will love 9.3 new features that allow project leaders to change workbook owners, run refresh schedule and move contents that are tasks that can be done by site/server admin only in the past. This new feature together with 9.2’s project permission locking feature really empowers project leaders.

Server Management

9.3 added bunch of server management features. Like low disk-space alerts; ProtgreSQL improvement allows failing over from one repository to another much more quickly w/o server restarting; The REST API is underpinned by a completely new platform with significant performance and usability improvements for admins; Postgres connectivity monitoring allows server admin check the underlying PostgreSQL database for corruption with a new tabadmin command.

Publishing Workflow

Publishing data sources or workbooks become easier and faster in 9.3: Tableau Desktop remembers your Tableau Online or Tableau Server connection and signs you in to the last server you used. It is easier to publish, keep your data fresh, and stay connected with the new Publish Data Source flow.

Better Map

Map is enhanced with postal codes for 39 European countries, districts in India, and US demographic data layers for 2016. Postal codes for UK, France, Germany, and the US are also updated. Mapbox supports new Mapbox GL as well in additional to 9.2’s Mapbox Classic.

Progressive Dashboard Load

It is cool that Tableau has progressive dashboard load feature now, which means you can start analyzing your data sooner w/o having to wait for the entire dashboard to load.

First time being keynote speaker to Tableau West Coast Customer Advisory Summit

Last week I was customer keynote speaker at Tableau’s annual West Coast Customer Advisory Summit. My talk was about how to scale NetAApp’s Tableau enterprise deployment to 4,000+ users within one year. It was well received. A lot of people came to me and said that they were inspired by my presentation. It was actually similar talk that I gave at TC15 Las Vegas but I added some recent work around content certification framework. Since it is close door summit, there is no recording but I made my slides public that can be downloaded @

Tableau VP Dave Story shared Tableau product roadmap. Other presentations include Tableau alerting, Desktop license management, which are all very good. Of course, we also went through product customer feedback exercises that customers voted for top ask features.  It was fun one-day event. It was great of meeting other Tableau big customers and a lot of Tableau product managers.

First time hosting Webinar for the entire Tableau Server & Online Admin group

I love Tableau. My passion is about Tableau server deployment – how to create governed self-service model with Tableau, people and process.  My company’s Tableau server added 4,000+ users within one year ( I got a lot tips & helps from Tableau community during last 2 years. I also want to give back, which is why I created a Silicon Valley Enterprise TUG focusing on Tableau server deployment which got a lot positive feedback. Recently it is recommended to extend the Silicon Valley Enterprise TUG to nationwide, which is why I become the co-owner of Tableau Server & Online Admin group. This is the first webinar for this group.

This webinar went extremely well with about 200 audiences via Zoom. Zoom is cool with its video, chatting and Q/A features. Speaker Mike Roberts (Zen Master) did amazing job to keep audiences closely for about 50 minis while I was busy answering questions via Q/A messaging. Mike shared great insights on workbook performance:

  • Workbook MetaData : What’s actually in the workbook (filters, row shelf, column shelf, etc)? Where do we get all the metadata WITHOUT using tabcmd/rest api? PostgreSQL / psql
  • Desktop vs Server: Something that performs well on desktop *should* perform equally well on Server. But sometimes that’s not true, how to troubleshoot it?
  • Alerts: How to create performance alerts to your workbook as people often don’t know if their workbook performance gets slow or not.

According to Mike,  a good workbook isn’t just one which performs well in Tableau Desktop. A good workbook should have following characters:

  • Data – general rule: more data = potential for high latency and poor performance
  • Design – proper use of filters, action, mark types, etc
  • Delivery – Where it’s delivered has a large impact on how it performs

The slides used is @

The code used is @

The recording is @ here for Webinar recording