Overall Satisfaction with Tableau Desktop
Tableau is being used to support data-driven decision making across the whole organization. Our use of Tableau is associated with providing support to various departments in answering specific questions they pose, examples include: demand forecasting, assessment of marketing campaigns, sales performance across various points of sale, demographic characterization of clients, and other performance indicators.
- Ease to connect to various types of data sources, hosted locally and in the cloud: almost "plug-and-play"
- Support to quickly create compelling dashboards that can be used by various analysts across the organization: I meet with the stakeholders who pose the questions and we brainstorm together so they lead me through the construction of the right set of assets. At any step, I am able to show them the current stage of the visualizations being built, until we reach a point at which what is left to be done is the finessing of the final assets.
- Ease to blend heterogeneous data sources: this "point-and- click" feature has been useful when we need to combine heterogeneous data sources in the creation of dashboards
- Easy sharing of visualization assets among peers: we create versions of the visualization assets under construction using extracts, which makes the sharing of the resources particularly easy since stakeholders focus on analyzing the data and not on issues related to the underlying data infrastructure
- Since I am a senior software engineer, I wish Tableau had a more robust underlying programming language with clearly defined semantics. To be fair, the ability to connect Tableau with R helps in this direction.
- I believe the quality and quantity of the official documentation maintained by the company could improve. Fortunately, there is a thriving community on which I have relied heavily. However, given the current level of maturity of Tableau, this should naturally prompt the company to document well-known solution patterns for recurrent visualization questions/problems.
- I would appreciate the ability to easily access data inherent to some "out-of-the-box" analyses available in Tableau. For instance, it would be nice to programmatically have access to parameters that are calculated in connection with linear trends such as the slope, the independent term, etc.
Tableau is best suited for scenarios where the data has been already prepared for visualization--in other words, where Tableau is used for actually building the visualization assets, and not for pre-processing the data. On the other hand, Tableau does not seem to shine in situations where complex data pre-processing is needed--think, for instance, of complex joins of large tables.