8 Data Formats Tableau Doesn’t Like

Sometimes you need to use data in Tableau that isn’t in a clean, denormalized format. It might have been exported from an application or prepared by a coworker in a way that Tableau doesn’t like. If possible, it is always best to get access to source data that is formatted in only rows and columns. However, sometimes that isn’t possible or practical. We’re going to cover what you can do in those scenarios.

Here are 8 data formats Tableau doesn’t like.

 

Multiple Values in the Same Cell

66-1.png

This is bad news. Tableau will concatenate the information as text so row 2, column C will show up as “137417%”. Each individual data value needs its own cell.

 

Multiple Dimensions in the Same Cell

66-2.png

When two dimensions share the same cell, it takes away flexibility in analysis. For instance, if we want to find out how many jobs were added in Ohio, we would first need to split “City and State” into two separate fields to be able to filter appropriately.

 

Blank Headers

66-3.png

Tableau needs data that is strictly in a rows and columns format. Row 1, Column A should be the name of the first data field. Row 1, Column B should be the name of the second data field. 

 

Total Rows

66-4.png

Totals should not be brought into Tableau as a distinct row in a crosstab. This kind of totaling can lead to accidental duplication and is better done inside Tableau. 

(Note that if a total is created as a separate column of data, that is generally okay.)

 

 Multiple Headers for the Same Column of Data

66-5.png

Notice that Column B has two headers broken out on different rows. The data is associated with “Store 100” but the measure name is “Sales”. Each column of data should only have one header.

 

Multiple Columns for the Same Data Value

66-6.png

Each unique measure should only be in a single column of data. Every value listed in the above example represents the number of enrolled people from different years and states. Each of the categorizations (“Year” and “State”) should be a column of data and “Enrolled” should be a single column of data as well.

 

Header Rows with Following Blanks

66-7.png

Every cell of data should be filled unless it was purposefully left blank because there was no data to report. In this case, Tableau will treat the blanks as nulls.

 

Duplicate Column Names

66-8.png

Tableau will begin naming fields with the same name [Field] 1, [Field] 2, etc. To avoid confusion, do not use the same column name twice.

 

Want to learn how to solve these problems and get your data ready for Tableau? Reserve your spot in our revamped Data & Tableau workshop today!