Tableau excels at data visualization and exploration. Terminologies such as a "sheet" are similar to the concept of a worksheet in Excel. Connection to most databases are supported out of the box. These factors keeps the learning curve short for business users. Charts in Tableau are rendered using visualization grammar called VizQL. This enables creating unique and out-of-box charts such as lollipop, sankey, sunburst, etc. The charts you can build in Tableau is only limited by your creativity. In addition, it is relatively to easy build interactivity and pass parameters in your dashboards so you can drill in and progressively reveal more details as the user interacts with data.
Improvements to My Organization:
Tableau has empowered the business users to get answers to their data questions without relying on IT teams. Rather than following build, wait, and iterate cycle, business teams are able to analyze the data much more quickly and identify opportunities for efficiency and revenue generation. Prior to the adoption of Tableau, data was present in multiple spreadsheet; thus, there was a lack of "single source of truth". This problem was resolved by using "Published Data Sources" residing on Tableau Server.
Room for Improvement:
Many areas for improvement are in Enterprise features. Some of which are:
1. A private folder on the server for a user to store his/her files.
2. A built-in tool for deployment and migration between multiple Tableau Server environments.
3. Ability to share a "database connection" between multiple Tableau data sources so that one change to connection information will be reflected in multiple data sources which share that connection.
4. Email distribution: Delivering analytics via emails on enterprise scale is difficult.
Areas for improvement in Visual Analytics:
1. Blending improvements: Slow performance when two sources are blended together on a high cardinality dimension. Blending does not support full outer join and does not allow using non-additive measures ( e.g. count distinct, average) from secondary data sources.
Update for Tableau 10.1: Some of the above are getting addressed in version or in future version as per Tableau road-map declared at Tableau Conference 2016 in Austin, TX. Specifically, there may be features for private folder in future. Tableau migration pains are somewhat alleviated with TabMigrate, their open source tool ( no support) and email distribution is getting better with conditional subscriptions ( 10.1).
Use of Solution:
Four years. Eleven years in BI.
In my experiences, Tableau deployment has been straightforward when deploying self contained dashboards (called packaged workbooks). However, you need to follow multiple steps to deploy a dashboard to your production server from your development environments.
Tableau (both Desktop + Server) is one of most stable software I have worked with.
Scalability in Tableau depends on many factors such as server configuration, networking, and workbook design. In my experience, the single biggest factor that affects scalability is the dashboard design. Excessive use of quick filters, multiple data sources blended together, long conditional expressions impact scalability. Reading "Designing efficient workbooks" whitepaper is extremely helpful. If your dashboards follow the best practices, it is possible to accomplish near-linear scalability by scaling your Tableau Server horizontally or vertically for large number of concurrent users.
Tableau customer service is very helpful and responsive. Both the customer service and technical support are integrated into their customer portal.
Tableau technical support is very professional and responsive. Many questions are answered in 1-2 business days depending on the severity. However, specific questions such as improving performance fall outside their scope. Thankfully, the Tableau community is one of the most engaged, knowledgeable, and helpful community. If you have a question that is not already answered by the community (very rare), you are guaranteed to receive a response within a few hours.
Initial setup was very straightforward involving running the installer program. I did not run into any issues with multiple versions of Tableau Desktop or Tableau Server.
All of my implementation have been in-house working closely with IT team.
Cost and Licensing Advice:
Tableau pricing is competitive with the other options available in the market.
Other Solutions Considered:
Yes. We evaluated Microstrategy v10, Sisense, SAP Business Objects.
Before committing to Tableau, it is helpful to list the top problems you are trying to resolve. If most of your needs are in operational business intelligence, you might want to evaluate other tools in addition to Tableau. If you want to analyze data, discover insights, communicate your story with data, or impress your customers with great visualizations, Tableau is by far the best tool.
For a successful Tableau adoption it is very important to have strong business user support and understanding the importance of data-driven decisions. Without the business user engagement, Tableau is just another tool.
Disclosure: I am a real user, and this review is based on my own experience and opinions.
Nov 22 2016