Building Out Interactive Dashboards in Tableau — Employee TurnoverRiley PredumBlockedUnblockFollowFollowingFeb 10Tableau is a graphical business intelligence and analytics platform, enabling people to quickly connect to a database and build plots, graphs, maps, and interactive dashboards.
It’s powerful, easy to use, and fun!In this article, I’ll share a project I created around a fabricated data set of demographic and employee turnover data throughout the US.
We know that employee turnover is one of the largest expenses for businesses: seriously.
The data for this project can be downloaded here.
If you want to see how the charts were constructed, the workbook is available for download.
You just need Tableau Public, the workbook, and the data set.
Now without further ado, I’ll dive into my analyses.
SetupI used the core_dataset.
csv file and, once I converted it to .
xlsx so that Tableau could load it, I got down to it.
Firstly, I examined the data to see what variables were available, what it looked like, and what I might want to look further into.
Glance at the data set (not all columns are visible here)Analysis and ChartsI knew I wanted to take a look at sex and race to see how those are distributed in relation to pay rate, department, and job.
I also wanted to look at these variables in relation to the turnover rate and the reason for employee termination.
So first, I needed to look at the general distribution of these.
Demographics of the employees data set: count by sex and raceThe majority of employees are White women, with the second largest group being White men.
The second most prevalent race are Black employees, and the third most prevalent are Asian employees.
To understand the pay rate differences, I started off by looking at how average pay rate differs across the country for the states in the data set.
Average pay rate by stateI then looked at a more granular level in terms of pay by race and department.
Pay by race and department, colored to the same scale as the map aboveFrom this table we can see a number of things:Only White employees are in the executive office, which is the highest paid position.
Hispanic employees have a moderately high range of pay rates, but are only represented in two departments.
They are the highest paid group in the IT/IS department.
This is average pay rate though, and I saw in the bar chart before that there were very few Hispanic employees in the sample.
As such, this is not a very accurate average.
American Indian or Alaskan Native employees are similarly under-represented.
This was visible in the bar chart prior as well.
Sales, Executive Office, Software Engineering, and IT/IS are the departments that pay the highest.
I then looked at turnover to try to get a sense of why employees were leaving their companies.
I focused in on voluntary reasons naturally.
Employment status and reason for voluntarily quitting by raceLooking at the charts, I saw that the majority of employees left for another position regardless of race.
The second most common reason was due to unhappiness in the workplace.
The third most common reason was to earn a higher wage.
To improve employee retention, I would recommend the following implementations based on these data:Perform an extensive exit survey/interview to understand how the company can improve to keep people from wanting to leave for other opportunities so much (retroactive approach).
Perform surveys before employees desire to leave, geared towards understanding happiness, satisfaction, and so on.
The areas to think about could include things like culture code, events and socials, performance reviews, and so on (preventative approach).
The cost of turnover is immense.
It should not be out of the question to consider offering employees more money to keep them around (preventative approach).
Diversity in companies is important as well to promote cultural understanding, as well as diversity of thought, and it has been shown to catalyze innovation.
With that in mind, I examined the employee source by race to understand where the companies were able to find diverse talent for their workforce.
The majority of diverse talent came from diversity job fairs.
The bin with the most number of employees sourced is referrals.
This makes sense and is common.
It’s also the cheapest way to hire.
Efforts can be focused based on past performance of different channels, much like how attribution models in marketing help companies decide where to invest ad spend.
Wrapping upAs you saw in this article, Tableau is a quick way to do some EDA on your data and build dashboards that colleagues can easily use and modify for their own analytical needs.
And this was all done in their free version!The full workbook is on my Tableau profile here.
See more projects on my GitHub.
See more of my writings.