Introduction to Tableau
Tableau is an innovative Business Intelligence company that headquartered in Seattle. Founded in 2013, Tableau is one of the fastest growing company in analytics. It has increased to a 2400-employee company with more than 200 million of annual profit. According to Forbes review, Tableau had around 4400 customers by the end of year 2010, and the number has increased to 26000 in 2014. The company has developed so fast that nowadays the term “Tableau” has become a buzz word in BI world.
In this article we want to get you started in tableau. First please download the 14-day trial version of this software. In this blog, we will share some of ATCG’s first-hand development on this software.
In general, most users express the interface is very “intuitive”. It looks quite friendly for especially non-technical end-users. You can select different types of connections from the left side of the screen, Tableau supports main databases such as SQL Server, Oracle, etc. and flat files as data sources. Or if your database is not listed in there, you can also use general ODBC connector for data connection.
- In this exercise, we used the superstore sample data spreadsheet provided by Tableau for exercise. Similar to eFashion from SAP, the superstore data is all virtual. In this worksheet, there are 3 different worksheets included in the file. After connecting to Tableau, you can see all these 3 worksheets on the left of the screen. On the right side, you can drag and drop your worksheets to the canvas and manipulate your tables.
- In the following screenshot, please put the orders worksheet into the canvas and check out the data from Orders sheet. You can do all the changes on names and properties in this screen.
- Date Split: you could split columns based on hyphen or other signals. For example, if you would like to split the date field to 3 columns---Month, Date and Year, you can split my date field based on “/” signal and get it done within several clicks.
- Joins: Another important function of this screen is that you can do all kinds of joins between your worksheets, you can simply drag and drop your worksheets (tables for databases) to the canvas, then connect related tables and define the type of the join. To add more data sources, simply clicking on the top of the page and you can define the extra data source.
Similar to Excel Pivot Table, Tableau supports simple “drag and drop” interface for users in data analysis. And there are also 2 types of fields in Tableau: Dimensions as non-numeric objects and Measures as numeric fields, there is a list of available objects on the left side. To start analyzing the data, the user simply drags and drops fields from the list to the Columns/Rows shelves at the middle top of the screen. Also there is filter/marks area at the middle of the page, to use filter/marks, one can still drag and drop targeted objects to those areas. Another main function of Tableau is that the software can make recommendations on the type of chart to be used in the analysis. As soon as the user selects the targeted fields he/she is interested in, the “show me” pop up will show recommended types of charts to the user.
- Drag and drop Field Product Category to the row shelf and profit as the measure
- Change it into pie chart
Before publishing your work, there are 2 ways to organize your analysis: story and dashboard. Either way, you will drag and drop all the analysis sheets you created and arrange them in the canvas. There is 1 main difference between story board and dashboard: dashboard is more like a combination of the analysis you made on 1 page, you could also make some interactivities as you like. In story board you can have multiple captions (pages) to consist the story, in each caption it can only contain 1 KPI that you want to show your audiences, also you can add comments to each page. After finishing the dashboard/story board, there are several ways you can publish your work. You could either publish it on the Tableau server or to local machine. To save it locally, you can either save your work as “.twb” file (Tableau Workbook) or as “.twbx” file (Packaged Tableau Workbook).
In this blog we provided a general picture of how does Tableau look like and how does it work. We also went through the general process to develop, summarize and publish our analysis dashboard/story board. In the future posts, we will dig deeper on the use of Tableau and find out more interesting and practical usages for Tableau.