Tableau Checklist: Key Principles for Dashboard Design

The Tableau best practices checklist is designed to provide developers with key questions to ask when designing a dashboard. The checklist serves as a resource for designers before, during, and after the dashboard is built. The components of the checklist ensure that the dashboard is following Tableau best practice standards and best-serves the desired audience. This checklist is intended to make you aware of the things to consider.

Purpose and Audience

When designing a Tableau dashboard, the audience and the purpose for the dashboard must always be front of mind. It is important to understand how much knowledge your audience has about the data prior to their viewing the dashboard. Understanding the level of knowledge your audience has will allow you to identify the cues your audience will need and what questions they want answered. The purpose of the dashboard should be conveyed to the audience through the title and surrounding text. A data visualization is only useful to an audience if they can interpret it.

Data Sources

Preparing the data properly before pulling it into Tableau to create a dashboard is essential to good performance. When connecting data sources and embedding data extracts it is key the data set is aggregated as much as possible. The smaller the data the better. To supplement data aggregation, filters and calculations should be applied directly in the database when possible. Performing calculations in the data set will aid in overall performance and centralize logic. Additionally, the data should be cleaned and setup structurally to match the analysis of the dashboard.

Dashboard Design

Arguably the most important aspect of a successful Tableau dashboard is using the proper charts to display your data. Pie charts should be avoided because they are more difficult for an audience to identify relative differences in data, have potential for misinterpretation and add complication to dimension with multiple categories. Chart types like bar or column charts are preferable in most cases.

The number of views included on a dashboard plays a key role in dashboard design. Tableau best practice is two-to-three views per dashboard. When a dashboard contains or exceeds five views, consider creating a separate dashboard. Too many views on one dashboard will slow down performance and potentially overwhelm the audience. For similar reasons, caution should be applied to using more than five filters. Over reliance on Tableau filters will not only take longer to query, but the results become less representative of the data.

The layout of the dashboard greatly effects impressionability. Views considered most important should be in the upper left corner of the dashboard. If the views are chained interactively then they should be constructed top to bottom and left to right. Visualizations that are related should be grouped close together, making it simpler for the audience to recognize the connection.

Dashboard titling must be intuitive for the highest success. The more intuitive the titles, the better understanding of the data the audience will have. Visualizations, tooltips, and axes should all have clear and specific titles that aid in how the audience interacts with the dashboard.

Lastly, the formatting of the dashboard plays a key role in the audience’s perception. When it comes to formatting, consistency is key. Consistent font and color must be used throughout the dashboard with proper branding. All formatting on the dashboard should be as intentional as possible. Consider removing default formatting for more intentionally tailored formatting for best results.

Performance

People are impatient and having a slow-loading dashboard is a quick way to lose your audience’s interest. Your dashboard needs to have an appropriate load time for the target audience. Below is a table detailing acceptable load times for various targeted audiences.

There are numerous steps you can take to increase performance. First, do not have unnecessary data connected to your dashboard. Smaller data sources can lead to significantly faster load times and dashboard performance. Next, certain functions perform better than others. Minimum, maximum, and attribute should always be used instead of average and sum. Additionally, the same can be said for using case instead of IF when applicable. It is best for calculations to be materialized and run before the user interacts with the dashboard. When calculations are not materialized within the data set, it is more efficient to use native Tableau features such as groups, sets, bins, custom date fields, combine fields, and aliases instead of calculations when applicable. 

Final Checks

Before taking the final jump and publishing your dashboard, give yourself a final walk through of your work. Confirm that your data is accurate, all filters are working as expected, and that there are no grammatical errors. Most importantly, ensure that your dashboard is telling a single story that can be understood in 30 seconds or less.

Download Avaap’s Tableau Checklist to begin building your dashboard.