How to start your first HR analytics project

Have you wondered how you can get started with your first HR or people analytics project? Applying descriptive and predictive analytics to an organization’s HR data is a relatively new area, and there is a growing need for HR professionals to build their competency in this area. Here’s how to get started. You may need the assistance of IT to get special software installed on your PC, but it is worth the effort to do so. With major time constraints and very little time to devote to your first project, I’m going to give you some suggestions on how you can get started.

Get your data.

Your HR data may be located in Excel files, or in your HRIS system. You may want to start by looking for basic Employee demographic data, salaries, names of employees and their titles, departments they work in, EEOE demographic data, and so forth. If you want sample data to play around with, Dr. Carla Patalano and I created a fictitious data set (based on real data though) and published it on Kaggle. You can get the data set by clicking here. Be sure to grab the file that says “core_Dataset.csv”. If you’re going to go this route first, I recommend downloading the .CSV file. CSV is short for comma-separated-value, and is essentially a text file with data separated by commas. Each line in the file is a separate row of data and is highly structured to make it easy to analyze with your data visualization software.  If you use your own data, it’s best if the data is highly structured either as a CSV file or Excel file. Perhaps the data may exist in a database, so you may need help determining which database, which server, and how to connect your visualization software to the database. That’s where you’ll need a little help from IT. Be sure that you’re working with cleaned data — data that is organized, complete, recent, and generally high quality. You won’t want to be working with stale or outdated data which has little meaning or utility. Once you have your data or have the information you need to connect to your data, it’s time for the next step.

Import the data or load the data into your data visualization software.

Be sure you get software installed on your PC like Tableau Desktop, Microsoft PowerBI, or MicroStrategy. These software applications all have trial versions that allow you to check out the software first before making any purchases. Tableau Desktop has a relatively low learning curve, and we teach this particular software application to HR students in New England College of Business’ graduate HR program. I highly recommend starting with an easy-to-use data visualization application like Tableau.

Explore the data by making some visualizations

Data geeks call this exploratory data analysis and descriptive analytics. We use data visualization techniques to explore the data and help us understand the data, and perhaps learn something new about the data. Once you import your data into Tableau (see how easy it is to do that HERE), you can simply start by playing around with the data. In Tableau, you can drag-and-drop the different fields (columns) onto the worksheet and the visualizations will start to appear. The visualizations that you make will help you uncover new insights about your data, and help you build a story about your employees and company. Look at some sample visualizations made with Tableau Desktop using the data set I mentioned above. I made these visualizations in Tableau Desktop and then uploaded them into Tableau Public.  Do NOT upload your own data to Tableau Public because your data is private and confidential. It’s fine to upload fictitious data there, but you wouldn’t want your private data becoming public information.

Pick low-hanging fruit and begin with relatively simple questions or problems

For your first real project, you should start small. There’s no need to start big in your first analytics project. In fact, it’s much better to choose something small, pick a smaller data set, and create small wins at the beginning especially if your organization is new to analytics.  These smaller projects can help you demonstrate the VALUE of developing analytic solutions and data visualizations. Once you see the visualizations, you should continue asking questions and digging deeper. By doing this, you may be able to justify the purchase of the software and start to take on slightly larger analytics projects and bigger questions. Explore the data in different ways perhaps by using vertical bar charts instead of horizontal. Heat maps and box plots can also be created which will help you gain more insight. Tableau lets you slice-and-dice the data in as many ways as you can think of. Our students at New England College of Business learn to make recommendations based on what they uncover in the data, and this skill is directly transferable to their jobs. Once you see your data, I believe it really comes alive and helps you make more sense out of it.

Ask questions, dig deeper

Some of the problems and questions you can explore can include:
  • What is the makeup of our organization in terms of diversity?
  • How are employees doing in terms of performance across different departments and perhaps for different managers?
  • What is the makeup of our best employees and best talent?
  • What are our best recruiting sources and where do we get the most value from which recruiting sources?
  • How has the organization changed over time in terms of pay equity across gender and race?
As you can see, many of these questions can be rather sensitive in nature, but are important questions to ask and have a solid understanding.

Present your interpretation of the data to leaders

By examining the data, and perhaps exporting the results into an Excel file or document, you can explain what you found in a few short sentences and include follow-up questions. Every data visualization project usually leads to additional insights and additional questions or problems. Ultimately, you are using the data visualizations to help you uncover truths about your organization and make informed decisions. You want to be making more decisions based on the data and evidence, rather than through hunches or “gut feelings”. My experience has been that once you show examples of data visualizations and things you’ve learned about the data, you can justify engaging in additional analytics and data visualization projects. Keep it simple and straightforward at first. Have fun with it and let me know how it goes!