Tableau in 2020: Upcoming features and improvements
The 2019 Tableau Conference recently wrapped in Las Vegas, and I was lucky enough to attend. Now that the music has died down and Iron Viz has ended, I wanted to share what I learned.
State of visual analytics
The visual analytics space continues to grow with more vendors hitting the market each year selling their take on data visualization and analytics. According to Gartner, Tableau and Microsoft have established themselves in a higher tier of market leadership compared to their competitors.
Several of the visionaries and niche players are interesting to me. ThoughtSpot has an interesting approach, Qlik is well rounded, and I think Domo should be in the visionary quadrant.
When we have discussions with clients, we tend to start with Tableau and Microsoft BI unless there is an obvious business reason for another software (specific business domain, current licenses, etc.).
Coming out of the Tableau Partner Summit and Conference, I think Tableau will maintain its position and start to close the gap on Microsoft from an overall analytics tool while extending its lead as a pure data visualization tool.
Key themes from the Tableau conference
Three key themes from the conference are: enterprise ready, viz sizzle, and data culture. In this article, I’ll cover only the enterprise ready and viz sizzle updates. Data culture deserves its own article!
- • Data modeling
- • Tableau Prep improvements
- • Data catalog
- • Server administration
- • Usability improvements
- • Animation
- • Explain data
Tableau has made major investments in the last few years and they continue to invest appropriately.
In Tableau, we've always recommended feeding wide, flat tables created specifically for Tableau coming from a DataMart. It has always been challenge to model many to many relationships in Tableau, but that is changing.
Tableau will be releasing an incredibly flexible data model they refer to as “relationship noodles.” Essentially how this works is that Tableau will make the relationship type choice (1:M, M:1, M:M) at time of analysis by running all scenarios in memory ahead of time. This also means that snowflake data models can more easily be modeled in Tableau. Note that the new "relationships" are essentially logical tables; you can now create "views" with join logic within a "relationship."
I saw many demos I was impressed by. I've found many business analysts don't understand data modeling and their models end up creating development headaches.
I'm excited to see how robust the update is!
Tableau Prep improvements
In upcoming Tableau releases we’ll see:
- • Better R and Python support. This extends Prep’s capabilities to include R & Python libraries.
- • Simplified data pivots. New built in functionality for moving row data to columns.
- • Writeback to databases and TDEs. Tableau Prep is now a better-integrated environment.
- • Re-usable ETL components. Common ETL operations can be centralized to simplify development and change management.
- • Event based schedules. I’m most excited for these! You’ll be able to sync up when Tableau Prep finishes and a workbook refresh happens.
Tableau recently introduced their own data catalog and it feels like they will continue to invest. They are opening their API to their graph query, allowing you access to:
- • metadata inventory
- • lineage and impact analysis
- • enhanced data discovery
You’ll also be able to perform integrated analysis from end to end on the data flow. I’m curious to see how the developer network utilizes this to create some new 3rd party tools.
Every major data cataloging vendor was at the conference as well. I spoke to lots of them (Informatica, Collibra, Elation, ZenOptics, etc) and all of them are working closely with Tableau for tighter integration. Our team will be doing a deeper analysis to understand the pros and cons of each of these tools and defining appropriate use cases for each one.
Full disclosure, I haven't had a lot of experience administrating Tableau Server, so I was really happy to learn more at the conference.
- • Deployment is getting easier. Tableau is making it easier to deploy sites, project, and workbooks between environments. You no longer need to download the workbooks. Instead, you can simply point and shoot.
- • Product keys are going away. Login-based keys replace product keys for their Role Based Subscriptions! License management will be simplified.
- • Encryption at rest. There will now be an option for encryption at rest and full container deployment support. This is great for companies looking to easily deploy full test environments and ensure proper data security.
- • Built-in dashboards. It's easy to monitor both usage, server, and site-specific performance statistics with new built-in dashboards. Troubleshooting and improving performance will no longer require deep technical expertise in Tableau Server.
Tableau has been focused on best practices, and plans to increase usability in a handful of ways.
- • Basic users. We've seen PDF subscriptions and better Export to PowerPoint. Now we're going to see improvements on tables (scroll bars, greater number of columns, etc.). Tableau Server is also making their environment more modern with smart recommendations to guide people to new insights.
- • Advanced users. For more advanced users, Tableau is focusing on integration with data science platforms to bring all analytics together.
- • Developers. For developers, Tableau will be introducing dynamic parameters, which will make life easier for auto-filtering to the current date when opening a dashboard, etc.
Tableau continues investing in usability and new functions that push them ahead of their competitors' capabilities. In 2020, we can expect to see:
- • Animation. This will include both the ability to play a time series for a scatter plot (a la Hans Rosling), racing bar charts, and being able to more easily understand changes when you're working. See some examples from TC19 below.
- • Explain Data. Tableau purchased an AI firm, Empirical Solutions, a year ago and now we get to see their new product features. Explain Data runs statistical algorithms in the background to figure out correlation across an entire data set. This tool brings advanced data analysis to a wider audience and will help many companies identify the root causes of issues. Benefits include: o Understand outliers: Simply select an outlier on a chart, click Explain Data and see in a tooltip supporting charts and correlation analysis as to what may be driving these changes. o Identify poor data quality: Simple statistics attached to outliers make it easy to identify poor data quality. If you clean up the poor data, then the algorithm automatically recalculates the statistics!
- • Headline data for mobile. I know a lot of my executive clients will love this feature. Using a tablet, users can simply tap any metric they are interested in and it will be available in a mobile friendly dashboard with “headline data” and basic trending. This works across all workbooks, so executives can quickly create their own "Wall Street stock" headlines with their most important KPIs.
I'm excited to see Tableau continue to grow and push the capabilities of their technology. They are also improving the analytics community by providing guidelines for creating Data Cultures from their research and studies across their clients. The new Tableau Blueprint program provides both strategic and tactical advice for driving business success through analytics.