Strengths and Weaknesses of Tableau Desktop

Tableau Desktop is one of the most common tools used by analysts. We will explore some of the strengths and weaknesses of Tableau Desktop. I’ll provide some options to make working with large data sets and advanced calculations in Tableau Desktop more efficient and collaborative. Finally, we will also see how Neebo extends the value of Tableau Desktop and all your analytics assets.

Tableau is one of the best tools when it comes to visualization and interactivity with data for an analysis. Tableau performance will degrade if you have large data sets, or data spread across different systems and need the ability to connect to all of your data at once. Given the need to create extracts of data in Tableau, other methods used with Tableau are better suited for optimizing complex data landscapes.

In addition, Tableau Desktop, by itself, cannot be published online for viewing. This hinders valuable collaboration for analytics teams. This collaboration is an element that is even more important with the recent need for analytics teams to work remotely.

Strengths of Tableau Desktop

Intuitive Interface

  • Tableau uses a drag-and-drop interface, and it works great for newcomers. The “show me” menu allows you to build out graphs with a click of a button.
  • Tableau makes items easy to find when building a viz. Users also discover many features by simply right-clicking. For example, converting dimensions to measure and vica-versa and assigning Geographic roles to dimensions is a breeze and promotes discovery with the data.
  • There is no scripting or programming knowledge needed to use Tableau. This is great for non-technical analysts and business users who want to get to the insights. Calculated fields, which are lite coding, act similarly to Excel formulas. With a little trial-and-error, business users who write formulas in Excel pick this up quickly.
  • Because of the drag-and-drop interface and the show-me menu, users build visualizations in seconds while other BI tools take much longer. To update formatting, simply right click and hit format. The items are optimized for speed since there is no scripting layer needed to update.

Interactivity With “Canned” Filters

Tableau comes with three great interactivity options that are fairly easy to set up: Filters, Actions, and Parameters.

Think of it this way: filters in Tableau are specific to a data source, parameters are not. Filters are created on the worksheet level. Tableau’s parameters can be reused across the entire workbook. Generally, filters can get the job done but there is compromise. While filters are easy they can be slow, particularly when there are many data sources in the dashboard. In addition, filters cannot be used when there are different data sources in a dashboard. Parameters must be used instead of a filter when trying to filter across different data sources.

Creating a filter in Tableau:

Building a parameter in Tableau:

These are similar ways of filtering data in dashboards but are used for different reasons. Each has advantages and disadvantages.

  • Filters: Tableau gives users the power to filter sheets or data sources in one click. This is impossible or requires advanced setup in some other BI tools.
  • Actions: Users actions are communicated between sheets. This makes Interactive dashboards easy to set up, with no scripting needed.
  • Parameters: Parameters allow developers to build customized input boxes where a user can select from a list, a range, or type in their own value.

Weaknesses of Tableau Desktop

Cannot Store Variables

Programming languages such as R and Python work in a waterfall manner and allow you to store variables at different stages in your script. Since Tableau has no scripting layer, there’s no easy way to store variables to save an overall number. Tableau does provide level-of-detail (LOD) functions to help solve this issue, but it is not intuitive and requires some guidance from a Tableau power-user to use. Other BI tools that come with a scripting layer, such as Power Query Editor in Power BI Desktop, can provide more intuitive ways to set it up.

Not for Complex Calculations or Advanced Statistics

Also due to the lack of a scripting layer, Tableau is not optimized for building some of the most complex or advanced calculations. Users build relatively basic functions easily, but as layers are added to dashboards, they quickly run into issues including painfully slow response times or corrupting workbooks (tip: there is no autosave in Tableau, so save often!). In a common example, referencing the same calculated field multiple times within another calculation will also result in performance issues.

If you are using Tableau but need to include complicated logic, LOD calculations can help solve some issues. A better method (for technical users) is to use a scripting layer prior to sending data to Tableau. Another method is to use Neebo to perform advanced calculations prior to sending data to Tableau. Neebo provides that intuitive environment where any user can perform advanced calculations on the data. Neebo also has a scripting layer, for those who know how to write some code. In addition, Neebo is a live connection to Tableau, so the results of advance calculations appear immediately in workbooks and dashboards.

Breaks With Too Much Data

Although this has improved, Tableau is not meant for native connections to big datasets. Tableau users are familiar with blending challenges and limitations. I spent too many lunch breaks wondering if my blend (or query) would be complete when I returned to my desk.

But in addition, if you have billions of rows or terabytes of data, Tableau’s data engine (named Hyper) is not meant to connect to that raw data. Custom SQL features within Tableau Desktop can make these datasets smaller through extracts prior to sending and blending in Tableau. But this not only requires a great deal of data movement but also data replication. Add the need for coding knowledge on top and we see that would limit the data that can be explored and used to make the best business decisions.

As these unbridled data extracts grow, it creates not only additional storage but also compute requirements. More importantly, it means the data used in dashboards isn’t live – and it may not be current. This method requires analysts to manually refresh an extract to update data which can take time and resources away from more important work.

The Hyper Data Engine

As mentioned above, Hyper is Tableau’s in-memory data engine. The in-memory approach is great, as long as you have enough memory. But Tableau’s native in-memory approach degrades as the data increases – and it fails at 3TB. Although there are workarounds and performance tips, other BI tools are much better at connecting to big datasets. Both Looker and Amazon Quicksight are good examples. Although both of these BI tools are harder to learn, slower, and do not provide as much visualization or interactivity options compared to Tableau.

Extend Tableau’s Value With Neebo

Neebo solves this issue by letting Tableau analysts explore and refine live data without moving or copying it. This speeds the process of getting the right data for the analysis at hand and makes the most efficient use of resources. Neebo works as an analytics hub to discover and model data prior to opening it in Tableau.

Neebo connects virtually and securely to various data sources, both on-premises and in public clouds. Point-and-click modeling tools, for not only analysts but also business users, provide the ability to shape live datasets to their needs without writing any code, and then explore the data in Tableau to answer their business problem.

Neebo was built to provide self-service access to all the data assets across the organization. Users find subject matter experts, reuse prior analytics projects, and discover new data sources, all from Neebo’s analytics hub. They collaborate around the data to understand what it is and how it was used (and by who) in the past – adding pooled wisdom to analyses and general making data projects more robust. Finally, analysts gain inspiration from their colleagues inside Neebo, which creates this compounding effect for knowledge as teams share resources and best practices across the organization.

See how Neebo can streamline analytics workflows with any BI tool. Test drive Neebo yourself by registering for a free 14-day trial.

Sign up to our newsletter

Editors Pick

Analytics Knowledge Management Collaboration analytics knowledge management
Advanced Analytics Banking Customer Experience
Best Practices Collaboration General big-data-analytics

Want to see our Virtual Analytics Hub in action ? Request a demo

Request a Demo