Online tableau training, tableau tips, & video tutorials



Shaping Data for Use with Tableau



Ryan Sleeper

Start on the right foot by properly preparing data for Tableau

The single biggest barrier to Tableau adoption is properly shaping the data. Ryan shares his personal Tableau story and the role of migrating existing Excel reports to Tableau played in his journey.

This is Ryan with Playfair Data TV, and in this video I’m going to show you the ideal way to shape data for analyses in Tableau.

The reason that I like to share this is I view this as the number one barrier to Tableau adoption. This is a lot of people’s story, including my own, so I’ll just go ahead and share my take on it.

I started my career at a very small marketing agency. Only five people worked there. This was eight years ago. And one day the boss walked in and said, I’ve heard of this new tool, Tableau. Can you guys take a shot at migrating our existing Excel reports over to Tableau? There were three analysts, three of us analysts in the room. We had been doing all of our reporting in Excel, so we were going to attempt to migrate all of our Excel reporting over to Tableau.

Well, we opened Tableau. If you have ever used Tableau and connected to any data source, you’ll see that on the Welcome screen, the very first data connection is Microsoft Excel. So we’re thinking, great. This is going to be easy. It’s the first connection. It’s going to work great. Just find my other Excel file. Everything will work magically. Maybe we can go home early for the day.

You may know that is not the scenario. That is not what typically happens. What happens is if you click Excel, you find your old Excel report, Tableau misclassifies all the fields. Nothing’s working, everything seems broken. You don’t know how to get started, and people quickly revert back to the familiarity of Excel. That’s why I think most people give up on Tableau in those early stages.

But this is the reason why that happens. It has to do with the shape of the data. Most existing Excel reports, your data is probably laid out something like this, where you might have your KPIs going down the first column and then your dates going left to right in the column headers. As you can see in this scenario, the dates may or may not even be in an actual date format . You might be abbreviating them somehow or aggregating them at a certain level. In this case, mine say Q1, Q2, Q3, Q4.

This could even be worse if we had totals for the quarters or subtotals either below or on the right. All of these things are problematic for how Tableau interprets a data source. The ideal layout– and this isn’t quite a blanket statement, but it’s pretty close– I would say 95% of the time, especially for a beginner, you’re going to want your data to be laid out more like this.

So notice what I’ve done here. I’ve transposed that table in the top left corner so that in the bottom right corner, it’s in a more vertical orientation now instead of a horizontal orientation. When you make each field a column header, each subsequent row becomes a combination of those column headers. That’s what helps Tableau slice and dice these fields in different ways to provide insight.

It’s also going to look at that second row and do two big classifications. Well, first it’s going to determine the data type. But it’s also going to determine whether it’s discrete or continuous as well as if it’s, by default, a measure or a dimension. We’ll talk about both of those in subsequent videos.

This has been Ryan with Playfair Data TV – thanks for watching!