HR professionals who harness ChatGPT for people analytics gain a powerful tool for extracting meaningful insight from the vast amounts of data being collected. With its ability to comprehend language and generate text, ChatGPT can participate in a conversation that simplifies your efforts to leverage data effectively.
In this article, we’ll explain how to use ChatGPT to analyze and interpret people analytics data for improved HR strategies and decision-making. Let’s get into it!
ChatGPT for people analytics: 9 use cases
Data analysts are using ChatGPT to quickly process large quantities of unstructured data. It helps them generate code snippets in Python, R, and SQL, as well as analyze datasets, visualize data, and more.
HR and people analytics professionals can also benefit from the tool when working with data. In addition to using ChatGPT for general HR work or recruiting, it can also simplify the understanding and interpretation of complex datasets and other people analytics-related tasks.
Here’s a look at ways to unlock ChatGPT’s capabilities with examples of how to use it in the following people analytics situations:
1. Exploratory data analysis (EDA) and data visualization
Data analysis with ChatGPT makes it easier to gather information for decision-making and create visual representations of the data.
Instead of having to write code to analyze the data or build formulas in Excel, you can ask ChatGPT to summarize it and describe the insights it reveals. Data visualization, such as charts and graphs, can clarify the main patterns in the data. ChatGPT can generate ideas for the best methods of demonstrating the relationships within the data in a visual format.
We provide practical examples of this below in the How to use ChatGPT for people analytics section.
2. Data summarization
Summarizing content from certain data sources, such as performance reviews, can be a time-consuming task. However, ChatGPT can help make it much more efficient and provide summaries.
|Example prompt: “Name the top three core strengths that the highest-performing managers have in common according to the presented data.”
3. Employee survey creation
ChatGPT can generate employee survey questions for collecting data on job satisfaction, engagement, or workplace culture.
|Example prompt: “Create a set of questions for an employee engagement survey. Include some questions that require responses on a 1-10 rating scale and some open-ended questions. Ask about what hinders team members from feeling engaged with their work and the company.”
4. Employee feedback analysis
ChatGPT can also help you analyze survey responses, including the qualitative data from open-ended questions. You can perform quick analysis of feedback from sources such as pulse surveys to make responsive, data-driven decisions. It becomes easier and more productive to collect feedback frequently.
|Example prompt: “Organize the following survey responses into three main themes. Display them in a list format in order of priority from most common to least common.”
By taking action and making changes based on survey results, HR can create a more transparent and supportive work environment.
5. Hypothesis testing
You can save time articulating potential explanations for certain challenges with ChatGPT. It can help formulate hypothesis statements to test your assumptions against the data.
|Example prompt: “Based on our observation of high turnover among new hires, suggest some hypotheses that we could test using our HR data.”
6. Generation of sample data sets
If you’re learning how to analyze different types of HR data but you don’t have such data at your disposal at your organization, you can use ChatGPT to generate sample data sets for you. You can then put your people analytics skills to work with them.
|Example prompt: “Generate a sample data set of 100 employees of a fictional company with the following columns: Employee name, gender, role, department, start date, and salary.”
7. Collaboration analytics
Data is needed to assess the effectiveness of collaboration efforts within the workplace. For example, HR professionals can use ChatGPT to analyze patterns in the timestamps of messages on Slack or Teams.
When are most conversations happening? Are there any obvious lulls in communication that might indicate a lack of collaboration or engagement? For instance, Microsoft analyzed Microsoft Teams data and uncovered that many people’s activity increased outside of a regular workday.
|Example prompts: “Analyze the distribution of messages sent throughout each 24-hour day. Identify any peak times or lulls in communication.” and
“Analyze the distribution of messages sent throughout the week, categorizing by day. Identify the days with the most and least activity.”
8. Sentiment analysis
ChatGPT can help businesses gauge the general sentiment of employees towards the organization, management, or specific initiatives. It processes content from sources such as internal communication platforms, social media, and anonymous feedback systems and assigns it a sentiment score.
|Example prompt: “Evaluate the overall employee sentiment about the new self-service payroll and benefits system.”
9. Generating Excel formulas for analyses
When you want to get the most out of data contained in an Excel spreadsheet, ChatGPT will do part of the work for you. You just have to describe what you want to do and have it generate the right formula for you.
|Example prompt: “I’m trying to calculate the average tenure of employees in our company. I have an Excel spreadsheet where column A lists employee names, column B contains their start dates, and column C contains their end dates if they’ve left the company, or the cell is empty if they’re still with the company. Provide a formula I can use in Excel to calculate the average tenure in years.”
The dos and don’ts of ChatGPT data analysis
While ChatGPT certainly has many useful applications for people analytics, there are certain factors to keep in mind to ensure it works to your benefit.
Below, we have listed what you should and shouldn’t do when interacting with ChatGPT:
- Prioritize data privacy by understanding and complying with applicable laws and regulations.
- Use detailed prompts and provide context to get more specific and accurate information. ChatGPT is based on natural language, so avoid using elaborate phrases and technical jargon.
- Try different types of questions and refine your prompts to learn which wording results in the most relevant and insightful responses. Pose follow-up questions and ask for clarification when necessary.
- Fact-check the results against the data for any anomalies, especially when the results seem doubtful. ChatGPT is a great tool for gaining quick insights, but make sure to double-check the results before presenting them to the team or leadership.
- Regularly check ChatGPT to stay updated on any changes, advancements, and new features.
- Input personal or confidential employee information. Anonymize any personally identifiable information about your employees and sensitive details about your business.
- Accept interpretations as absolute truth. These are observations made from the available data without the ability to consider every possible contributing factor or sentiment.
- Rely solely on ChatGPT insights for complex decision-making. Be sure to enrich its recommendations with human judgment, critical thinking, and discretion.
Double-check what you upload. Before uploading anything into ChatGPT, ensure that all identifying and sensitive information is removed.
This includes employee and company names, contact information like phone numbers, email and physical addresses, and medical information.
Note: Because of the strides taking place in technology, it’s foreseeable that companies will soon have their own ChatGPT-like tools that can be tailored to ensure data privacy and security.
How to use ChatGPT for people analytics
Here’s a step-by-step guide on how to utilize ChatGPT in analyzing people-related data.
Step 1: Enable the right features
To get the most out of ChatGPT for people analytics, you need to use ChatGPT Plus. That will enable you to install plugins, upload files, and create data visualizations.
Enable experimental features:
- Go to Settings & Beta and click “Beta features”.
- Enable the features.
Now, you can upload, for example, Excel files with data you want to analyze.
- Explore ChatGPT’s Plugin Store. There are tens of plugins available to install that you can play around with. ChatGPT will automatically apply the suitable plugin for the task you’re giving it. Note that these are third-party plugins, and ChatGPT may share information with the external plugin providers. New plugins are being added frequently. You can find a list of plugins for data and research here.
Step 2: Upload data to ChatGPT
Let’s have a look at a practical example of using ChatGPT for people analytics. We’ll work with a sample data set in .csv format, which you can download here. This is what its first rows look like:
We will start by loading this file into ChatGPT. To do that, select “Advanced Data Analysis” in the ChatGPT-4 tab and click on the Upload (+) button.
Then, you can start giving instructions to ChatGPT.
Acknowledge ChatGPT’s limitations. You can only enter a certain number of queries within a certain time period.
The chats with experimental features are also timed out, so you might not be able to go back to the conversation after some time.
Furthermore, you need to verify the results ChatGPT generates, as there can be mistakes and wrong interpretations.
Step 3: Provide data analysis prompts and utilize visualizations
Let’s say that we want to see the average salary by department, using a prompt: “Create a graph showing average salary by department.” You’ll get a simple bar chart without much effort:
You can follow up by looking into contract distribution in a certain department. (Prompt: “Create a pie chart showing the distribution of contracts in the HR department.”)
You can also get a quick insight into the average salary by education level. (Prompt: “Create a box plot showing salary by education level.”)
If you’re unsure what type of visualization would be most helpful, you can ask ChatGPT for advice. (Prompt: I want to visualize salary distribution by performance score. What type of graph or visualization would be best for that?”)
With our sample data set, you can go more granular and compare the salaries of men and women on a 100% contract per seniority level. This would help you understand the state of pay equity in your organization. (Prompt: “Create a table showing average salaries by gender and seniority for employees on a 100% contract. Include a column with a percentage difference between genders.”)
Step 4: Interpret the results and get further recommendations
The table above shows that women earn less on all levels except for the Director level. In your analysis, you might want to dive into the details – Are there differences in performance, education, or tenure that could explain this disparity? Then, you can start making changes to rectify the pay differences within your organization.
As you can see, ChatGPT can be helpful in gaining quick insights into your people data. That gives you a good understanding of what you need to focus on in your further analysis.
If you’re not sure what you can do with your data, you can ask ChatGPT for ideas. (Prompt: “Are there any particular analyses that you would recommend doing on this data set?”)
Machine learning and AI are playing a crucial role in transforming HR into a force for data-driven strategizing and decision-making. ChatGPT provides HR professionals with an easy-to-use tool for understanding and utilizing people analytics data to its full potential.
With the insights compiled by ChatGPT, HR can efficiently access more detailed information, focus on the right metrics, and organize and justify new initiatives. This makes HR more relevant in achieving business goals and driving success for the organization.
Andrea Boatman is a former SHRM certified HR manager with a degree in English who now enjoys combining the two as an HR writer. Her previous positions were held with employers in the education, healthcare, and pension consulting industries.